Re: [Computer-go] [ANN] Imago - Go board optical recognition
This sounds like something we could add to the FireFly object recognition feature on the Fire Phone. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Huang Sent: Tuesday, August 12, 2014 5:15 AM To: computer-go@dvandva.org Cc: tomik.mu...@gmail.com Subject: Re: [Computer-go] [ANN] Imago - Go board optical recognition I'm dreaming about this scene in EGC 2015.. Several cameras are relaying the games of the best players. A smart optical recognition program automatically converts the streaming images to sgfs and sends them to a Go program. The Go program then shows rich analyses over the games, such as wining rate, best moves, principal variations, estimated territories, etc, even predicts the next moves. A spectator is watching a friend's game and wondering who is ahead. He doesn't understand Go very well. He uses his android phone takes a snapshot. After 3 seconds, the Go software oh his phone immediately tells him that his friend is ahead for 20 points and winning with 90% chance. It's more of just a dream though. :-) Aja 2014-08-12 12:35 GMT+01:00 Petr Baudis pa...@ucw.cz: Hi! Tomas Musil (a student of mine), has created a state-of-the-art open source Go board optical recognition software. We have focused on completely automatic runs, so it automatically detects the board corners and then the stones on the board, and the precision seems pretty good at least in reasonable lighting conditions. You can find it at http://tomasm.cz/imago together with a lot of pictures, documentation and bachelor thesis describing the algorithms in detail. In the thesis, Tomas also compares it against other similar apps, and it appears Imago shows the best performance of all these that were available to us. Unfortunately, we specifically couldn't easily compare it to Remi Coulom's Kifu-snap for multiple reasons - mainly because that is a mobile app. Hopefully, someone will be able to compare these two in the future. At any rate, I think Imago is a great starting point for anyone who would like to play with Go board recognition. My personal dream would be if we added video capability and further improved speed + reliability in time for EGC2015 (in Czech Republic) and were able to deploy it there to transfer large number of top boards. But this will depend on how much time Tomas will have after the summer (and we didn't actually check with EGC2015 organizers yet), so it's still more of just a dream. :-) Petr Baudis ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] What happened with the August bot tournament
I've been swamped with work the last 8 months (lab126/Amazon - I worked on Fire Phone and Fire TV). Now that they shipped I might have a little more time and start entering Many faces again. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Sunday, August 03, 2014 3:22 AM To: computer-go@dvandva.org Subject: [Computer-go] What happened with the August bot tournament Hi, I just looked into the website of the August bot tournament. Big surprise: no more 7 or more applicants, but only two unknown players. What has happened? Ingo. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Bitwise-parallel surround capture
I never used paths. I keep a linked list of each point in the chain, and each stone has the index of the head of the list it is in. The board has an array of 361 list head/tail pointers, and an array of 361 root-indexes. Merging linked lists is O(1) since I keep a pointer to the last item. Updating the list-head-index requires walking the shorter list. I don’t support undo in the playouts, so there is no need for the extra bookkeeping to support restoring the old state. Yes, I maintain bitmaps in parallel with the other structures, but I only use them for pattern matching. I didn’t add them to the playouts until the playout time was dominated by move generation, not updates. From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Peter Drake Sent: Saturday, May 24, 2014 11:57 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Bitwise-parallel surround capture On Sat, May 24, 2014 at 11:29 AM, David Fotland fotl...@smart-games.com wrote: No. Many Faces uses union find. I looked through some old literature, including Anders’ thesis. Even though we all used this algorithm it seems he didn’t mention it, probably because it seemed too obvious. Orego currently uses a variation on union-find with eager path compression; each point is given a direct link to the root of its chain at the time of merging. Is there a significant speed advantage to the lazy approach? I do use bitmaps for pattern matching arbitrary shaped patterns up to 8x8. Does this mean that you maintain the bitmaps in parallel with the other structures? -- Peter Drake https://sites.google.com/a/lclark.edu/drake/ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Bitwise-parallel surround capture
The referenced union-find algorithm has been known and used by computer go programs since at least the early 80's, and I'm sure it was published before 2011. I think your implementation of union-find is a little slow. Many Faces of Go uses this algorithm, and on 9x9 got about 55k playouts per second on a much slower machine in 2008. Here is a prior discussion of bitmap performance. https://groups.google.com/forum/#!searchin/computer-go-archive/fotland$20pla youts$20per$20second/computer-go-archive/IdZtmM1Q6Ow/Q4b1IxgdkZYJ A bitmap representation can be used for fast pattern matching, so there are other advantages than just performance to using it as the base representation. -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Cameron Browne Sent: Saturday, May 24, 2014 2:45 AM To: computer-go@dvandva.org Subject: [Computer-go] Bitwise-parallel surround capture Hi, This draft paper describes a simple bitwise-parallel method for performing surround capture: http://www.cameronius.com/research/go- bits-1.pdf Before I submit it, I just wanted to check with this list that the method is not known and already in use. Any general comments on the paper would also be appreciated. Regards, Cameron ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] A few questions from a beginner
On 9x9 you should be getting about 100k light playouts per second per core on a recent core i7. I was getting almost that many in 2008, and libego was faster. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Xavier Combelle Sent: Wednesday, March 19, 2014 11:44 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] A few questions from a beginner As other engine: - There is Łukasz Lew. libego https://github.com/lukaszlew/libego (I belived it will give you what you want) - There is oakfoam http://oakfoam.com/ - There is fuego http://fuego.sourceforge.net/ On Wed, Mar 19, 2014 at 5:28 PM, Mikko Aarnos mikko.aar...@kolumbus.fi wrote: Thanks for the info all. Does a fast open-source program exist which uses only light playouts? If there would be one I could get some reliable info on the amount of playouts per second I _should_ be getting. The only open-source I've found is Pachi and that isn't suitable for that purpose. I considered the playout method before asking but I rejected it because I didn't think it was any good at figuring out life/death. Apparently this is not the case so seems like that's what I'm going to be doing. Regards, Mikko Aarnos ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Japanese rule in MC
Many Faces chooses E5 with 67% with 5 seconds of thinking (about 150K
playouts). I don't use never fill sure territory in playouts. I tried
something like this (not filling in two point eyes), but it didn't help. I
also don't use the one point margin idea. I found a way to get the
correct result with Japanese rules instead.
David
From: computer-go-boun...@dvandva.org
[mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Huang
Sent: Sunday, March 16, 2014 1:43 AM
To: computer-go@dvandva.org
Subject: Re: [Computer-go] Japanese rule in MC
Hmm.. maybe it should be never fill own sure territory in the playouts,
except to defend a threat or reply to opponent's move inside the territory.
This is a hard problem because it involves wasting no move or cancelling out
both side's moves correctly in the playouts.
Aja
2014-03-16 8:28 GMT+00:00 Aja Huang ajahu...@gmail.com:
Hi Hiroshi,
Maybe it's worth trying never fill own sure territory in the playouts
except replying to opponent's move. That is to say, after certain number
(say 64) of playouts if the territory belongs to one side exclusively (say
99% of the playouts) then from the next playout don't fill the territory in
the playouts actively, unless the opponent plays inside the territory first.
Aja
2014-03-16 1:25 GMT+00:00 Hiroshi Yamashita y...@bd.mbn.or.jp:
Hi,
Recently I have tried to use Japanese rule without one point safe margin.
I use Erik and Aja's method. It works pretty well, but this position is
still diffcult.
9X.O..
8X.O..
7X.O.. Black(X) to play
6...XXOO.. komi is 6.5
5...X.O...
4...X.O...
3...X.O...
2...X.O...
1...X.O...
ABCDEFGHJ
(;GM[1]SZ[9]RE[B+0.5]KM[6.5]RU[Japanese]
;B[di];W[fi];B[dh];W[fh];B[dg];W[fg];B[df];W[ff];B[de];W[fe]
;B[dd];W[fd];B[ed];W[gd];B[ec];W[gc];B[eb];W[gb];B[ea];W[ga])
Best move is PASS or filling share liberties(F9,F8,F7,E5,E4,E3,E2,E1).
Aya can play it, but its winrate is around 0.50, far from understanding
this position is easy win for B.
Tentyo no Igo(commercial version of Zen) also returns around 0.50.
Is there good method that can return over 0.80 winrate?
I feel Do playout and modify score by number of pass is hard for B +0.5
win position. Because B has more territory about 7 pt, and W has bigger
chance to pass in playout. As a result, B winrate is lower than 0.50.
My code is like this.
if ( last_two_move_is_pass_in_tree ) { // confirmation phase
pass_w_playout = pass_b_playout = 0;
playout_length = 0;
}
game_length = record_length + tree_legth + playout_legnth;
pass_w = pass_w_record + pass_w_tree + pass_w_playout;
pass_b = pass_b_record + pass_b_tree + pass_b_playout;
Margin = 0;
if ( (game_length1) ) {
if ( first_player_is_white ) Margin = -1;
else Margin = +1;
}
minus = (Margin + pass_w - pass_b);
japanese_score = chinese_score - minus;
double final_score = japanese_score - komi - handicaps;
Without confirmation phase, winrate drops from 0.50 to 0.45.
Aya does not grow tree after consecutive pass in tree. Maybe I need to grow?
But to do that, I need to have two root positions that have same hash value,
one is root, and the other is after onconsecutive pass.
Steenvreter's solution
http://dvandva.org/pipermail/computer-go/2010-April/000233.html
Erica's solution
http://computer-go.org/pipermail/computer-go/2013-February/005757.html
Many Faces' solution
http://dvandva.org/pipermail/computer-go/2013-February/005748.html
Useful link by Fuego team
http://sourceforge.net/apps/trac/fuego/wiki/JapaneseRules
Regards,
Hiroshi Yamashita
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] commercial go program
I sell Many Faces as a 32-bit app, which will run on every machine. It can’t use more than 2 MB of memory though. Unless you are letting it think for several minutes per move, the memory limit won't affect the playing strength. I can make the 64-bit version available to current owners, which would lift this restriction. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Tuesday, March 11, 2014 12:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] commercial go program Helly Ray, I have experience with CrazyStone, running under Windows XP and Windows Vista. In both cases the program uses all cores very extensively. On the one side this is very nice, on the other side it has side effects: For instance, when you want to start text processing (doc-files or odt- files), it takes veeery long until the corresponding window opens - when CrazyStone is running simultaneously. ManyFaces is less greedy for the resources. Concerning maximum size of ram use I can not help you. Ingo. Gesendet: Dienstag, 11. M rz 2014 um 00:19 Uhr Von: Ray Tayek rta...@ca.rr.com An: computer-go@dvandva.org Betreff: [Computer-go] commercial go program hi, i was thinking of buying a go program. i have the latest many faces, but it can't use my 16gb of ram on windows 8. are any of the go programs able to use that much ram and take advantage of my 8 cores? thanks --- co-chair http://ocjug.org/ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] [O-T] Arimaa
I barely have enough time to work on computer go, so arimaa will be untouched probably until I retire or at least switch to working part time. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Peter McKenzie Sent: Wednesday, February 26, 2014 8:54 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] [O-T] Arimaa On Fri, Feb 21, 2014 at 11:22 PM, David Fotland fotland@smart- games.com wrote: Both are hard. I worked hard on Bomb the first year or two since I thought the best chance to beat a person was before people developed expertise. I published the algorithms I used for Arimaa in Bomb. The paper should be on-line somewhere. Oh yes, I had already read the paper. Thanks for contributing it - it's definitely one of the main references for anyone interested in writing an Arimaa program. Do you think you'll come back to working on your program at some point? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Peter McKenzie Sent: Friday, February 21, 2014 9:25 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] [O-T] Arimaa Hi Joshua, I learnt to play Arimaa a few days ago and started by playing some games against the Android program. I then joined up ariima.com (handle is 'petermck') and have been working my way through the bot ladder. It's quite an interesting and addictive game! It certainly looks like a fascinating programming challenge. I've started tinkering with some Arimaa code but who knows if I'll have time to turn it into a fully fledged bot... I wonder whether computers will overtake humans in Arimaa or Go first? cheers, Peter On Tue, Dec 10, 2013 at 11:52 PM, Joshua Shriver jshri...@gmail.com wrote: Curious has anyone else here been infected with the Arimaa bug? I'm an avid chess guy, as well as Go. Hence why I'm here :) But from a playing point of view, I find myself playing Arimaa so much more, and anxious to develop for it. -Josh ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] [O-T] Arimaa
Both are hard. I worked hard on Bomb the first year or two since I thought the best chance to beat a person was before people developed expertise. I published the algorithms I used for Arimaa in Bomb. The paper should be on-line somewhere. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Peter McKenzie Sent: Friday, February 21, 2014 9:25 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] [O-T] Arimaa Hi Joshua, I learnt to play Arimaa a few days ago and started by playing some games against the Android program. I then joined up ariima.com (handle is 'petermck') and have been working my way through the bot ladder. It's quite an interesting and addictive game! It certainly looks like a fascinating programming challenge. I've started tinkering with some Arimaa code but who knows if I'll have time to turn it into a fully fledged bot... I wonder whether computers will overtake humans in Arimaa or Go first? cheers, Peter On Tue, Dec 10, 2013 at 11:52 PM, Joshua Shriver jshri...@gmail.com wrote: Curious has anyone else here been infected with the Arimaa bug? I'm an avid chess guy, as well as Go. Hence why I'm here :) But from a playing point of view, I find myself playing Arimaa so much more, and anxious to develop for it. -Josh ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Arimaa
I wrote the first strong arimaa bot (bot_bomb), and wrote a paper on how I did it, If you want a starting point. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Dave Dyer Sent: Wednesday, December 11, 2013 12:21 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Arimaa At 11:52 PM 12/10/2013, Joshua Shriver wrote: Curious has anyone else here been infected with the Arimaa bug? I'm an avid chess guy, as well as Go. Hence why I'm here :) But from a playing point of view, I find myself playing Arimaa so much more, and anxious to develop for it. Arimaa is designed to be hard for computers. There's a community of 'bot builders at arimaa.com. I'd also like to have a better robot for Boardspace.net. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] How many probes down the tree are necessary for a good bot?
Bot strength is still progressing, but now it is more a matter of engineering than grand new ideas. It's not so interesting to say something like, I fixed 5 bugs this weekend and got 30 ELO, or I used CLOP to tune 30 parameters for a week and got 50 ELO. I think all the strong programs use some form of dynamic Komi. There are a few published papers on it, but I suspect the strong implementations are rather different from the papers. Mine is. Nothing secret, just more complex. On-line learning in the playout policy is another thing that is worth doing, but needs considerable engineering to a good strength boost. Several papers talk about this, so the new ideas are not secret, just the details of the implementations. Since each program has different go knowledge, any sharing of tuning parameters or small bits of knowledge would not be very interesting to the larger community. I think many of us are happy to answer questions, so feel free to ask. Lately I've been working on making better use of Many Face's patterns in the tree. I do not use large harvested diamond-shaped patterns in the tree like most programs. I use the old hand-entered pattern knowledge I put into Many faces over the last 20 years. These patterns are irregular shapes, so more general, but there are many missing patterns, so the knowledge has more holes. The old many Faces engine with all the knowledge is single threaded, which is a pretty severe bottleneck on modern machines. I pulled the pattern matcher out of the old engine and made it thread safe so I can use it as an independent module in the tree. Right now it is still too slow, but once it is working well, I plan to do automatic tuning of the pattern weights to replace the values I created by hand over the years. Hopefully that will give a jump in strength, since I can do large pattern matching at all nodes in the tree. Today I only run the old engine (which includes the patterns) on nodes with at least 30 visits. Is this the kind of discussion you were looking for? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Stefan Kaitschick Sent: Friday, November 15, 2013 1:13 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] How many probes down the tree are necessary for a good bot? When you do only a single playout, you dont even need a playout, just generate the first move candidate - voilá. :-) Closed sources are indeed regretable at this point. Before, it sparked a kind of bot war, and the greatest technical advances are always made at war time. But now that the top bots are kind of treading water, looking for the next breakthrough, it's truly harmful. And my feeling is, that the attitude has also changed on the academic, open source side. I only know that this forum used to present really interesting ideas, but now that is seldom the case. Possibly the problems have become too technical for general discussion. But more plausible to me is, that programmers are trying to avoid a onesided disadvantage, and are discussing their ideas in private. Well, no use crying over spilt milk.(As I, admittedly, just have) Stefan On Fri, Nov 15, 2013 at 12:59 AM, Darren Cook dar...@dcook.org wrote: ... I'm going by the Measuring program strength thread (Aug 3013) where people are getting 50% against GnuGo with 1K to 10K playouts.) Actually, in that thread Hiroshi was saying he only needs 350 playouts, and Detlef was at 700. So it seems you're off by roughly a factor 10. Yes, I was surprised just how heavy the Aya (and Zen) playouts must be. We've almost come full loop, and soon the programs will be traditional computer go programs doing a single playout ;-) It is a shame the developer's skill is all dead knowledge (not open source, no papers). :-( Darren ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] September KGS bot tournament: 19x19, slow
It's not easy to code an engine that scales well to a cluster. In 2008 I ahd access to 1024 cores, but the cluster code I wrote at the time only scaled to 32 cores (4 machines, 8 cores each). That code no longer exists, and I have access to a single 16-core machine, so I can't currently run on clusters. With limited development time, I'd rather work on making the core algorithm stronger, than writing cluster code. I think it would interesting to have a slow tournament with a fixed maximum number of cores (4 or 8, since they are readily available). David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Detlef Schmicker Sent: Sunday, September 01, 2013 1:44 PM To: n...@maproom.co.uk; computer-go@dvandva.org Subject: Re: [Computer-go] September KGS bot tournament: 19x19, slow Thanks Nick, I love the slow bot tournaments. Two reasons: 1) For me computer go is a hobby and to hire 6 cluster instances on EC2 is about 150$ for a tournament. 3 tournaments one i7-4770k:) 2) Not all programs can handle clusters. It is an additional problem for authors, which are trying to get into this business. And I can tell you, it is difficult enough to get into it:) If you look at the last slow bot tournaments only few programs (gomorra, orego and zen) used clusters, maybe partly because of this reasons. Detlef Am Sonntag, den 01.09.2013, 20:00 +0100 schrieb Nick Wedd: The September KGS bot tournament will start at 22:00 UTC on Sunday September 8th, and end by 22:00 UTC on Tuesday August 10th. It will have 6 rounds, Swiss, with 19x19 boards. The time limits will be three hours each, sudden death. The komi will be 7.5. There are details at http://www.gokgs.com/tournInfo.jsp?id=835 . Please register by emailing me, with the words KGS Tournament Registration in the email title, at mapr...@gmail.com . This may be the last slow tournament, with time limits of over an hour each, that I run. Now that cloud computing is easily available, I believe that there is little purpose in setting such slow time limits. If you want to see how a bot does given a lot of thinking time, it makes more sense to hire multiple processors than to let it run for a long time. (If you think I am wrong, you can probably convince me of it.) Nick ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] algorithm quality assessment
For quality assessment, play many games against one or more reference opponents. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Oleg Barmin Sent: Wednesday, June 12, 2013 8:02 AM To: computer-go@dvandva.org Subject: [Computer-go] algorithm quality assessment Hi, everybody, I am working at the development of a cards game algorithm using MCTS. Technically, the game model is expect minmax tree search, where direct search takes up too much time, that is why I decided to use MCTS. The issue of using MCST, like any other approximation algorithm is its quality assessment. I am developing an algorithm for a game where no recognized masters exist. How do you think, guys, if for instance Go (or Amazons) provided no way to assess an algorithm playing with professional gamers (or other programs), how would you assets its quality? My second question: I have not yet learned Go in and out, however in my opinion, any search of a next step should identify a number of options with similar or even the same assessment. How do you resolve this issue? Best regards, Oleg Barmin. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Narrow wins
Sorry, secret. It took me quite a while to find something that worked well. I think you are right that most of the benefit comes from keeping the win rate close to 50%. I think Pachi found it was ideal to keep the win rate a little below 50%. I try to keep it in the range of 45% to 55%. I think Pachi's approach has been published. Mine is rather different. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Monday, June 10, 2013 9:02 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins On Mon, Jun 10, 2013 at 11:18 AM, David Fotland fotl...@smart-games.com wrote: The dynamic komi can be pretty large. I think I limit it to 30 points or so. If the komi were only a point or two you would see a lot of games won by 2.5 points or so. You can see from the results that the margins are often pretty large. A big issue which dynamic komi tries to solve is the difficulty in resolving close scores. For example in a won position the playouts might return 99.8% wins and a move which wins by larger margins might still only return 99.8 or even 99.7 or something just due to statistical noise. In computer chess testing we have a similar issue - especially as our program has gotten so strong.If you test against a weak program, it's almost impossible to resolve small differences in strength.Imagine a weak club player playing Go against the world champion in order to resolve the relative ELO difference (assuming no handicapping.)The pro is likely to win thousands of games before losing a single one and there would be no basis for determining where this player stands with any precision. So I assume that by far the primary benefit of dynamic komi is to bring the statistics back into the sweet spot so that it's easier to resolve the quality of move from one to the other. I have not kept up with this, but is your dynamic komi implementation a carefully guarded secret? I would like to know more. It seems to me that there would be some serious efficiency losses in fishing for the right value to use - or is this arrived at pretty quickly? Don David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Stefan Kaitschick Sent: Monday, June 10, 2013 1:24 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins On Sun, Jun 9, 2013 at 9:23 PM, David Fotland fotland@smart- games.com wrote: Dynamic komi and some other tricks work quite well. Thanks to Ingo for pushing dynamic komi until I figured out how to make it work well. Often the playout have some bias due to a misread in a fight, so it's important for the bot to keep its lead. Is that a small, fixed bonus of 1 or 2 points? A large dynamic komi in the endgame would really surprise me. But to actually compensate a misread fight, that komi would have to be substantial ... Like Don, I don't think that much good could come of it. Isn't it superior to award a point bonus in the playouts? Here's another idea: find a smart way to average the playout results into an expected result. If the bot is leading, try to maintain that lead in the next move. Stefan ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Narrow wins
The primary benefit is more natural play against people, which is more important to me than it is to some other programs. In testing against gnugo on 19x19 with a 3 stone handicap, dynamic komi increases the win rate by about 3%. Most of the variants I tried made the program a few percent weaker, but I found one that is slightly stronger. ManyFaces on kgs is stronger since I added dynamic komi, but it has many other changes as well. Perhaps people are more likely to resign when they are far behind than when the game is close. If you look at the game results, ManyFaces wins many games by resignation. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Darren Cook Sent: Tuesday, June 11, 2013 1:47 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins The dynamic komi can be pretty large. I think I limit it to 30 points or so. If the komi were only a point or two you would see a lot of games won by 2.5 points or so. You can see from the results that the margins are often pretty large. Hello David, Interesting! Is the benefit just from more natural-looking play (from the point of view of humans), or is it actually increasing the win rate do you think? (And if so, just the win rate against humans, or against other programs too?) Darren ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Narrow wins
I have a linearly decreasing handicap komi like Pachi. I have a different solution to the flapping problem. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Tuesday, June 11, 2013 6:17 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins Hi! On Mon, Jun 10, 2013 at 11:48:01PM -0700, David Fotland wrote: Sorry, secret. It took me quite a while to find something that worked well. I think you are right that most of the benefit comes from keeping the win rate close to 50%. I think Pachi found it was ideal to keep the win rate a little below 50%. I try to keep it in the range of 45% to 55%. I think Pachi's approach has been published. Mine is rather different. Note that you can read my findings in http://pasky.or.cz/go/dynkomi.pdf (I recommend e.g. Fig.2 to quickly grasp the picture). Petr Pasky Baudis ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Narrow wins
The dynamic komi can be pretty large. I think I limit it to 30 points or so. If the komi were only a point or two you would see a lot of games won by 2.5 points or so. You can see from the results that the margins are often pretty large. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Stefan Kaitschick Sent: Monday, June 10, 2013 1:24 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins On Sun, Jun 9, 2013 at 9:23 PM, David Fotland fotland@smart- games.com wrote: Dynamic komi and some other tricks work quite well. Thanks to Ingo for pushing dynamic komi until I figured out how to make it work well. Often the playout have some bias due to a misread in a fight, so it's important for the bot to keep its lead. Is that a small, fixed bonus of 1 or 2 points? A large dynamic komi in the endgame would really surprise me. But to actually compensate a misread fight, that komi would have to be substantial ... Like Don, I don't think that much good could come of it. Isn't it superior to award a point bonus in the playouts? Here's another idea: find a smart way to average the playout results into an expected result. If the bot is leading, try to maintain that lead in the next move. Stefan ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Narrow wins
Dynamic komi and some other tricks work quite well. Thanks to Ingo for pushing dynamic komi until I figured out how to make it work well. Often the playout have some bias due to a misread in a fight, so it's important for the bot to keep its lead. If you look at kgs games with strong bots, 0.5 wins are now very rare. Example: manyfaces: http://www.gokgs.com/gameArchives.jsp?user=manyfaces David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Saturday, June 08, 2013 12:29 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Narrow wins Stefan, To a bot this always comes down to risk analysis and it's no different than what we do. In fact you just did your own risk analysis here of the computers play and have decided that it does risk analysis wrong so your thinking is no different from the bots. To me the biggest issue isn't winning by 0.5 but not fighting when it's losing.I have a feeling that most of this is not about improving the play of the bot although it's always cast that way, but in reality it's about not offending our own sensibilities.We just hate to see a bot win by 0.5 when it probably could have won the entire board or most of it. And of course not fighting when losing is not much fun to watch when it may still have a realistic chance of winning due to an opponents mistake. The standard proposed solution is komi-manipulation in one way or another. But what we are probably really after is something to do with opponent modeling, giving the computer more of a fighting spirit instead of playing overly logical.I sense there must be a better solution as komi-manipulation just seems to me to be really illogical.It's like fixing the books because we don't like the numbers. I have failed to keep up much with computer go since getting involved with my chess program Komodo.But I see that this problem is still being talked about. Has there been any progress on this front?If this were computer chess we would modify the evaluation function, but the evaluation function in GO is really based on the play-outs. At the end of every playout we have a final board positions that returns some information. Some positions will return bigger wins than others (or smaller losses) but we know scoring this way is terrible.But maybe that information can be effectively used to shape the tree towards positions that win bigger? Sometimes it's a matter of which node to explore first - so it is quite likely that at least in some cases the computer does not make any distinction between a big or small win and thus we could possibly push it in that direction in the tree part? Maybe that is naive, I don't know. So what is the state of the art on this now? Don On Sat, Jun 8, 2013 at 2:21 PM, Stefan Kaitschick stefan.kaitsch...@hamburg.de wrote: Humans may be predisposed to the fallacy of greed. But bots have there own fallacies. Happily winning by 0.5, when a higher win at almost the same perceived risk is available, is a kind of full knowledge fallacy. A bot can lose an easily won game by merrily giving away everything but the last half point, then have the last point stolen from him by a single bot misread. So if you feel that the bot you're playing against is misjudging a particular situation, don't spring in on him unless that will put you in the lead. Otherwise, it's better to let him hand over other points that you need first. The bot might fix his problem, but that risk is preferable to having the bot go into business again when it is still ahead. Stefan ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Hardware pattern matching acceleration
Many Faces of Go uses 8x8 patterns with all types of wildcards (any, white-or-empty, black-or-empty, etc). I have about 2000 patterns, which would expand to over 100,000 without wildcards. It uses about 5% of total run time if I match patterns whenever I create a new node in the UCT tree (for initial bias). David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Mark Boon Sent: Wednesday, May 29, 2013 4:34 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Hardware pattern matching acceleration When using a Trie type data structure, there's no hard limit to the size of patterns in a pattern-matcher. It may be that some people only use small pattern sizes because they use fixed-size patterns with no wild-cards. Some years ago (late 2008 or early 2009) I outlined an incremental pattern-matcher on this list using the 'bread-crumb' strategy. I think you'll find it is quite competitive in speed. But since I never had anyone ask me about any details I assume the need for a pattern-matcher you're proposing is not all that great. Maybe one of the people with a competitive bot can tell you how much time they spend in a pattern-matcher. Mark On Wed, May 29, 2013 at 10:52 AM, Рождественский Дмитрий divx4...@yandex.ru wrote: I'd probably need less than 64 of it because it scales like a square. Making a chip twice larger (and twice more expensive) doubles the pattern library size and at the same time doubles the number of moves that may be matched in parallel. And when you tell that it is not an order a magnitude faster, do you mean that CPU can match that much 7*7 or 9*9 patterns with wildcards? I was told that current programs use relatively small patterns, and could benefit much from large ones. Dmitry 29.05.2013, 23:13, Mark Boon tesujisoftw...@gmail.com: Assuming the 2 microseconds mentioned is for the whole board (you don't mention what board-size, I assume 19x19) it sounds quite fast. But probably not an order of magnitude faster than a software solution. The question then becomes how well it scales. Would you need 64 of these to service a 64-core computer? Note: at some point you should decide for yourself what project you think is worth taking on. You'll always find that other people have 100 reasons why not to try something, but they're often wrong. Mark On Wed, May 29, 2013 at 8:02 AM, Detlef Schmicker d...@physik.de wrote: Hi, sounds interesting. I am not to experienced with large patterns at the moment, but we (oakfoam) are using 100 (circular) patterns up to 15 (as the size is counted in several papers), so 3 might be not enough for a strong program. Usually we need all pattern matches at the board at the same time, is your 2ms this time? I am very interested in GPU acceleration, as this might be very interesting in mobile devices. They usually have a strong GPU and I was thinking about accelerating exactly what you are talking about. It is somewhat difficult to calculate the performance this approach would offer. Another interesting acceleration might be real liberties (instead of pseudo liberties) in the playout moves. This might help to use more heavy playouts. But I do not have data, if this would help, as we do only have pseudo liberties. Detlef Am Mittwoch, den 29.05.2013, 21:38 +0400 schrieb Рождественский Дмитрий: Hi, after thinking over your advices in the previous thread and making some investigation I have figured out two options of working on the hardware accelerator. One is to develop a totaly new algorithm that fits for a hardware acceleration better than current ones. Or to find what can be improved in current algorithms moreorless revolutionary, because just acceleration of an algorithm part is not a solution, I thought that maybe it will be interesting to improve pattern matching. Current programs can massively match relatively small patterns. Hardware may have the following parametres: - pattern size up to 9x9 with wildcards (does not matter field state to eleminate influence of insignificant peripheral stones' positions); - additional attributes as usual (liberties, ko, distance to an edge) - internal position calculator with pattern extractor (just send a cell position and receive its belonging to a pattren back) - several patterns' evaluation at a time (should further specify how much) - about 3 7*7 patterns in an $200 device - about 2 microseconds time Does anyone have an idea will it be valuable? Dmitry ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go , ___ Computer-go mailing list Computer-go@dvandva.org
Re: [Computer-go] Hardware pattern matching acceleration
In the tree a node represents a position, so each visit after the first will have the same pattern matches. Your hardware would be quite a bit faster, but the overall program speedup is pretty small. My patterns are hand generated, so the size is limited by my ability to say interested in adding new patterns. For a program with automated pattern learning and no don't care your hardware might be interesting. Since I sell software, no hardware will work for me. I can't ask my customers to open up their computers and install a PCI card. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of ?? ??? Sent: Thursday, May 30, 2013 3:27 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Hardware pattern matching acceleration So, the cheapest hardware might have several times larger pattern database and give an ability to check for patterns not only for initial node bias but at every first node visit after each move (as long as position changes). Is it seems to be valuable? Will it give an ability to build tree deeper? Dmitry 30.05.2013, 10:32, David Fotland fotl...@smart-games.com: Many Faces of Go uses 8x8 patterns with all types of wildcards (any, white-or-empty, black-or-empty, etc). I have about 2000 patterns, which would expand to over 100,000 without wildcards. It uses about 5% of total run time if I match patterns whenever I create a new node in the UCT tree (for initial bias). David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Mark Boon Sent: Wednesday, May 29, 2013 4:34 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Hardware pattern matching acceleration When using a Trie type data structure, there's no hard limit to the size of patterns in a pattern-matcher. It may be that some people only use small pattern sizes because they use fixed-size patterns with no wild-cards. Some years ago (late 2008 or early 2009) I outlined an incremental pattern-matcher on this list using the 'bread-crumb' strategy. I think you'll find it is quite competitive in speed. But since I never had anyone ask me about any details I assume the need for a pattern-matcher you're proposing is not all that great. Maybe one of the people with a competitive bot can tell you how much time they spend in a pattern-matcher. Mark On Wed, May 29, 2013 at 10:52 AM, Рождественский Дмитрий divx4...@yandex.ru wrote: I'd probably need less than 64 of it because it scales like a square. Making a chip twice larger (and twice more expensive) doubles the pattern library size and at the same time doubles the number of moves that may be matched in parallel. And when you tell that it is not an order a magnitude faster, do you mean that CPU can match that much 7*7 or 9*9 patterns with wildcards? I was told that current programs use relatively small patterns, and could benefit much from large ones. Dmitry 29.05.2013, 23:13, Mark Boon tesujisoftw...@gmail.com: Assuming the 2 microseconds mentioned is for the whole board (you don't mention what board-size, I assume 19x19) it sounds quite fast. But probably not an order of magnitude faster than a software solution. The question then becomes how well it scales. Would you need 64 of these to service a 64-core computer? Note: at some point you should decide for yourself what project you think is worth taking on. You'll always find that other people have 100 reasons why not to try something, but they're often wrong. Mark On Wed, May 29, 2013 at 8:02 AM, Detlef Schmicker d...@physik.de wrote: Hi, sounds interesting. I am not to experienced with large patterns at the moment, but we (oakfoam) are using 100 (circular) patterns up to 15 (as the size is counted in several papers), so 3 might be not enough for a strong program. Usually we need all pattern matches at the board at the same time, is your 2ms this time? I am very interested in GPU acceleration, as this might be very interesting in mobile devices. They usually have a strong GPU and I was thinking about accelerating exactly what you are talking about. It is somewhat difficult to calculate the performance this approach would offer. Another interesting acceleration might be real liberties (instead of pseudo liberties) in the playout moves. This might help to use more heavy playouts. But I do not have data, if this would help, as we do only have pseudo liberties. Detlef Am Mittwoch, den 29.05.2013, 21:38 +0400 schrieb Рождественский Дмитрий: Hi, after thinking over your advices in the previous thread and making some investigation I have figured out two options of working on the hardware accelerator. One is to develop a totaly new algorithm that fits for a hardware acceleration better than current ones. Or to find what can be improved in current algorithms moreorless revolutionary, because just acceleration of an algorithm part
Re: [Computer-go] Go playing software accelerator development
I don't know how you are calculating 300 instructions in 100 ns. It seems too low. A modern CPU executes a peak of about 4 instructions per clock. At 3.4 GHz, in 100 ns, the peak number of instructions is 1360. Actual will be lower than peak, but it should be quite a bit higher than 300. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of ?? ??? Sent: Tuesday, May 21, 2013 11:36 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Go playing software accelerator development Incredible, 100 nanoseconds is only about 300 instructions of a CPU. Are you talking about 19x19? And 1 microsecond for my design will probably be a worst-case (as I calculate freedom and capture iteratively). When almost all stones have free places around it will be down to ~100 nanoseconds. As to the number of possible accelerators on-chip - it varies upon price. I think it can be 5-250, for the price $250-$5000. So the cost of a single simple accelerator will be $20-$50. Dmitry 21.05.2013, 23:13, Mark Boon tesujisoftw...@gmail.com: Sounds interesting. But 1 microsecond for a move is not particularly fast. There are already implementations that do that in the 100-300 nanoseconds range on one core. 1 microsecond is probably considered as 'semi-light' playout. I suppose the question then becomes, how many of these could your accelerator do in parallel? Mark On Tue, May 21, 2013 at 8:06 AM, Alexander Kozlovsky alexander.kozlov...@gmail.com wrote: Я тоже кстати из ЛИАПа, с четвертого факультета, может и пересекались :) On Tue, May 21, 2013 at 7:02 PM, Рождественский Дмитрий divx4...@yandex.ru wrote: Hi all, I have got an idea to create a hardware accelerator for Go playing software. It will probably be a USB (or maybe PCI-Express) device that will be able to do some basic, but very time-consuming for general-purpose CPU calculations very fast. For example load a goban layout, make a number of random moves (as used in Monte-Carlo algorithm) and unload result back to a computer. As long as it will be a hardware, it will be able to do specified calculations only, but the speed will be very high. For example, making just a copy of the particular goban layout will require typically about 10 nanoseconds only (one internal clock cycle). Calculation of the validity and results of a particular move (including a check for ko and captured stones) will probably take 1 microsecond. This as usual may vary during debugging, but the current move calculation engine draft I've started to develop is about this figures. My nearest aims here are: - to understand a demand from go playing software developers, and - to understand what particular calculation chains are most demanded for hardware acceleration. Dmitry ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go , ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Go playing software accelerator development
I measured the speed using only one core. 133 ns is the time is takes one core to make one move in the playouts. In Many Faces there is no small piece of code that uses a lot of time. The top functions and times are: 12% Generate local moves 8%: count liberties for ladder search in playouts 8% match big patterns on the full board during the tree search 8%: make a move in playouts6% generate moves to remove an eye in the playout move generation 5%: count liberties of adjacent enemy group during playout move generation 4%: calculate gamma for a move during move generation for tree search 3%: extract an 8x8 bitmap for large pattern matching during tree search 2%: check a generated move for legality 2%: make a move during ladder search in the playouts Etc. Many of these functions are complex, and accelerating any one of them will have negligible impact on performance. You may be able to find some simple algorithm that can play well and that can be accelerated. The top programs are too complex to gain by this kind of hardware. Accelerating a few key functions can’t make the program much faster. Amdahl's law severely limits the gains that are possible. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of ?? ??? Sent: Wednesday, May 22, 2013 12:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Go playing software accelerator development i7 is four-core, so probably it is closer to the upper bound. But if you say that only 8% of time spent on calculating board positon it changes nothing: it becomes obvious that some other things than board positoin calculations should be accelerated first. And I will highly appreciate any hints about what it could be. Dmitry 22.05.2013, 11:05, David Fotland fotl...@smart-games.com: I just checked a profile for 19x19 on one 3.4 GHz i7-3770 core. 8% of the time is spent in making moves in play outs.� So the maximum possible performance benefit of hardware accelerated move making is only 8% higher performance. On one core Many Faces is doing about 1200 games per second, of about 500 moves each, so it is making 600,000 moves in 8% of one second.� That�s 133 ns per move made, near the lower end of Mark�s range.� It seems that this hardware won�t make the program any faster. Many Faces� play outs are quite heavy, but the additional time is mostly in move generation, not making moves. The code to update the board state is quite efficient.� It includes things that are probably not included in the hardware below (like updating the index of the local 3x3 pattern at each empty point), keeping a list of liberties for each group, and updating local feature information that will be used by the move generator.� The �make move� function is 400 lines of C, so it�s doing much more than just simple board state update. David From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of ?? ??? Sent: Tuesday, May 21, 2013 11:36 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Go playing software accelerator development Incredible, 100 nanoseconds is only about 300 instructions of a CPU. Are you talking about 19x19? And 1 microsecond for my design will probably be a worst-case (as I calculate freedom and capture iteratively). When almost all stones have free places around it will be down to�~100 nanoseconds. As to the number of possible accelerators on-chip - it varies upon price. I think it can be 5-250, for the price $250-$5000. So the cost of a single simple accelerator will be $20-$50. Dmitry 21.05.2013, 23:13, Mark Boon tesujisoftw...@gmail.com: Sounds interesting. But 1 microsecond for a move is not particularly fast. There are already implementations that do that in the 100-300 nanoseconds range on one core. 1 microsecond is probably considered as 'semi-light' playout. I suppose the question then becomes, how many of these could your accelerator do in parallel? Mark On Tue, May 21, 2013 at 8:06 AM, Alexander Kozlovsky alexander.kozlov...@gmail.com wrote: ? ?? ?? ?, ? ?? ??, ? ? :) On Tue, May 21, 2013 at 7:02 PM, ?? ??? divx4...@yandex.ru wrote: Hi all, I have got an idea to create a hardware accelerator for Go playing software. It will probably be a USB (or maybe PCI-Express) device that will be able to do some basic, but very time-consuming for general- purpose CPU calculations very fast. For example load a goban layout, make a number of random moves (as used in Monte-Carlo algorithm) and unload result back to a computer. As long as it will be a hardware, it will be able to do specified calculations only, but the speed will be very high. For example, making just a copy of the particular goban layout will require
Re: [Computer-go] ladder
I'm not sure I understand exactly what you are asking, but I think neither is true. Many MCTS programs read ladders as part of move generation or bias. For example ladder status may be part of the pattern data used for move bias in the UCT tree. During play outs, some programs use ladder status to select the move to play. For example a local response to the last move might be If the last move put a stone in atari, play in the liberty of the stone in atari, provided it is not captured in a ladder. Many Faces uses ladder status in these ways. I don't think a program should play out the ladder in the play out, since half the moves are very bad. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Chun Sun Sent: Tuesday, May 14, 2013 7:22 AM To: computer-go@dvandva.org Subject: [Computer-go] ladder Hi... do MCTS programs usually put a specific ladder branch, or this branch should be automatically generated given the right implementation? This question has been hoovering in my head for a while and I wanted to take the shortcut to ask here :) Thanks, Chun ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] ladder
Many Faces reads ladders in both the tree and the play outs. In play-outs we already had a rule don't play self atari (with many exceptions). A similar rule is don't play a move that is captured in a ladder (with many exceptions). The moves that play out the ladder are not made in the play-out itself, but ladders are evaluated as mart of the play-out move generation policy. Once you have heavy play-outs, the overhead of reading ladders is not very high. In the tree portion, I use the Many faces of Go engine to bias moves, and it does tactical local search for up to 3 liberties, not just simple ladders. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Tuesday, May 14, 2013 8:51 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] ladder I was curious about this myself. I am so heavily involved in computer chess that I'm having trouble keeping up with Go. But I am amazed at how strong the 19x19 go programs when I know that ladders are a big part of go and that they are not easily dealt with in playouts. So if I understand this correctly, you do significant analysis in the tree portion of MFGO but not in the playouts?Is that pretty typical of the best programs? Don On Tue, May 14, 2013 at 11:41 AM, David Fotland fotl...@smart-games.com wrote: I'm not sure I understand exactly what you are asking, but I think neither is true. Many MCTS programs read ladders as part of move generation or bias. For example ladder status may be part of the pattern data used for move bias in the UCT tree. During play outs, some programs use ladder status to select the move to play. For example a local response to the last move might be If the last move put a stone in atari, play in the liberty of the stone in atari, provided it is not captured in a ladder. Many Faces uses ladder status in these ways. I don't think a program should play out the ladder in the play out, since half the moves are very bad. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Chun Sun Sent: Tuesday, May 14, 2013 7:22 AM To: computer-go@dvandva.org Subject: [Computer-go] ladder Hi... do MCTS programs usually put a specific ladder branch, or this branch should be automatically generated given the right implementation? This question has been hoovering in my head for a while and I wanted to take the shortcut to ask here :) Thanks, Chun ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Computer Go at the European Go Congress
I will enter if someone can run the program for me. Many faces only runs on Windows, but it would not be too hard to make a Linux version (no GUI, GTP control, for KGS). Have they said how the computers will communicate? KGS would be ideal, sine the operator would just run a script before the tournament then watch all the games. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Thursday, May 02, 2013 3:54 AM To: computer-go@dvandva.org Subject: [Computer-go] Computer Go at the European Go Congress The 2013 European Go Congress will be held in Olsztyn, Poland, from July 27th to August 11th. On Friday August 2nd, the EGC Science Conference will be held there, see http://egc2013.go.art.pl/page/egc-science- conference . It will include three computer Go tournaments, see the final paragraph of http://egc2013.go.art.pl/page/go-tournaments I learned today, from the organiser Pawel Noga, that entrants for these tournaments will be required to use the computers provided. I do not yet know the specification of these computers, though I guess they will be running Windows. I also do not yet know whether programmers will be allowed to appoint someone who is at the Congress, to operate their programs for them. I have asked Pawel about both. I would like to hear from the owners of strong programs whether they are planning to enter. If no strong program will be competing, I will not want to travel to Poland for the event. If there will be at least two strong programs competing, I will travel to Poland, and will offer to operate competing programs for those who cannot be there themselves, if this is to be permitted. If I am to be in Poland for the Congress, the August KGS bot tournament will probably be on August 25th; if I am not going there, it will be on August 4th. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] A Regression test set for exploring some limitations of current MCTS programs in Go
Interesting that Aya and ManyFaces scored the same. -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Hiroshi Yamashita Sent: Tuesday, March 26, 2013 4:28 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] A Regression test set for exploring some limitations of current MCTS programs in Go Hi Aja, Thanks for posting this result. It seems seki is easy for MC. But semeai and Life and Death are big problem. Almost all programs don't understand except Zen. Zen's result is awesome. I think it is a reason Zen is 5d or 6d and others are 2d or 3d. About test, this post is helpful. A Regression test set for exploring some limitations of current MCTS programs in Go http://www.mail-archive.com/computer- g...@dvandva.org/msg04954.html Seki128k playouts Aya 32/33 96% Crazy Stone 29/33 87% GNU Go L10 29/33 87% GNU Go MC 24/33 72% Gomorra 26/33 78% Many Faces 32/33 96% Pachi 26/33 78% StoneGrid 31/33 93% Steenvreter 33/33 100% Fuego 17/33 51% Zen 33/33 100% two_safe_groups_0.3 128k playouts (W two groups are alive.) Aya 2/15 13% Gomorra 0/15 0% Many Faces 2/15 13% Pachi0/15 0% Steenvreter 4/15 26% Fuego0/15 0% Zen 13/15 86% My anti-semeai version Aya gets Aya 7/15 46% (anti-semeai) Aya 1/15 6% (normal) But I could not get good result on KGS and selfplay from anti-semeai. Its strength is almost same. Maybe side-effects? My anti-semeai strategy is 1. Each node searchs its first 100 playouts. Then 2. Semeai analysis runs. 3. Recognize two adjacent groups that has similer living percentage, like 25% and 28%. 4. Count their true liberties. If one has 2 libs and one eye, 1 share lib, 5 nakade in the corner include 2 dead stones, etc... 5. In playout, if one reduces his own true libs, undo and play killing another group move. Regards, Hiroshi Yamashita - Original Message - From: Aja Huang ajahu...@gmail.com To: computer-go@dvandva.org Sent: Wednesday, March 20, 2013 9:39 PM Subject: [Computer-go] A Regression test set for exploring some limitations of current MCTS programs in Go Dear all, If you are interested, you can download the newest version of our regression test set (seki and two-safe-groups) at http://webdocs.cs.ualberta.ca/~shihchie/seki-and-two-safe-groups- regression-test.zip or in Fuego svn http://fuego.svn.sourceforge.net/viewvc/fuego/trunk/regression/name which contains the results of all participating programs including Crazy Stone, Zen, Steenvreter, pachi, ManyFaces, Gomorra and Fuego, etc. Kind regards, Aja ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
[Computer-go] UEC Cup Crazystone vs Ishida Game
I recorded this myself (and the Zen game), so any errors are mine. David ishida-crazystone.sgf Description: Binary data ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Some CrazyAnalysis of Zen game
All mcts programs have trouble with the positions near the end. The group in the center has miai for two eyes. Same for the group at the top. The upper left side group has one big eye shape. For all three groups the playouts sometimes kill them. The black stones are pretty solid, so the playouts let them survivie. SO even at the end, zen has 50% win rate, MFGO has 60%, and pache has 70% win rate for blasck. The CS game doesnt have this problem. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Mark Boon Sent: Wednesday, March 20, 2013 12:50 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Some CrazyAnalysis of Zen game On Wed, Mar 20, 2013 at 9:14 AM, Ingo Althöfer 3-hirn-ver...@gmx.de wrote: Question to stronger go players: Where do you think started Zen's problems? Cheers, Ingo. I hardly play anymore, but I'm maybe still strong enough to comment on computer games :) I think trouble started by taking on the white group at the top too strongly, culminating in a terrible move at B 50. I think B should have extended at R7 instead. Next, B desperately tries to surround the center, but plays elsewhere on crucial moments. B 78 should have been at N12, B 86 at M4. When B plays 92 I don't know what to say. Anywhere else seems better, but maybe the game was already lost by then. When W captures at 103, the game looks totally over to me. Still, four handicap is a lot, so there might have been a small chance had B played 104 at R3 or so, but it seems unlikely. Mark ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] UEC Cup will be broadcast on Ustream
They didn’t broadcast the preliminary. They plan to broadcast the final, today, which starts in about 3 hours. The preliminary 7 round Swiss was won by Zen: Zen Crazystone Aya Many Faces of Go Pachi Nomitan RGO Fuego (running on a backup machine – the big cluster had some issues) Etc. – 24 programs total. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Saturday, March 16, 2013 12:15 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] UEC Cup will be broadcast on Ustream Can someone please post the sgf of the games (or a link) here in the mailing list? Thanks in advance, Ingo. Gesendet: Mittwoch, 13. März 2013 um 22:24 Uhr Von: Hideki Kato hideki_ka...@ybb.ne.jp An: computer...@computer-go.org Betreff: [Computer-go] UEC Cup will be broadcast on Ustream Dear readers, The 6th UEC Cup (http://jsb.cs.uec.ac.jp/~igo/eng/index.html) will be broadcast on Ustream (http://www.ustream.tv/channel/%E7%AC%AC6%E5%9B%9Euec%E6%9D%AF%E3%82%B3%E3%83%B3%E3%83%94%E3%83%A5%E3%83%BC%E3%82%BF%E5%9B%B2%E7%A2%81%E5%A4%A7%E4%BC%9A) What will be broadcast are semi-final, final and exhibition match, and will start at 13:00, 17th March JST (UTC+0900). The exhibition will be the winner of the tournament vs. a strong Japanese amateur player (ex-trainee of Nihon-kiin) with no (!) handicap. Hideki -- Hideki Kato mailto:hideki_ka...@ybb.ne.jp ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Anomalies in MCTS
Many Faces of Go gets this one right (my current unreleased engine, not the 12.022 download). It requires some seki knowledge in the playouts. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Monday, February 04, 2013 5:38 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Anomalies in MCTS Hi Cameron, hi Jonas, thanks for your feedback. You might try some of the RAVE counter-examples, such as the following position from Martin Mueller (White to move): http://www.cameronius.com/research/go-rave-blunder-1.pdf It's a seki position in which FUEGO chooses the correct move B2 without RAVE, but the incorrect move D9 with RAVE enabled. I translated this position in sgf. Programmers can find it at www.althofer.de/mueller-rave-blunder-1.sgf and may try with their bots. Regards, Ingo. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Progressive Bias, progressive widening
Will you publish or share your progressive widening implementation? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of ds Sent: Sunday, January 20, 2013 7:51 AM To: computer-go@dvandva.org Subject: [Computer-go] Progressive Bias, progressive widening Hi, I know, this topic was in the list a while ago. My problem is, as in all science, nobody publishes negative results:) Oakfoam uses both, progressive bias and progressive widening. My understanding is, this is state of the art in many mc bots, at least the theses I read used both. Both is working well in oakfoam. Now I turned off progressive widening and tuned progressive bias carefully (good scaling of the bias and improved decay functions). I got the same playing strength as with both (bias and widening) before on 9x9 against gnugo, but I can not improve anymore with progressive widening turned on again. My interpretation is: Progressive bias is the superior concept, but it is easier to use progressive widening. Progressive widening is not sensitive to the ratio of the pre knowledge value of two moves, only the better move must be unpruned first, but progressive bias is sensitive to the ratio between the pre knowledge values. It may be even more difficult to improve the progressive bias on 19x19, so there might be a reason to use widening, but at the moment I feel I should try without? Am I wrong? Detlef ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Handicap 29 prize
I just put it back up. It crashes sometimes, so if it's not there, send me an email and I'll restart it. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Mark Goldfain Sent: Friday, January 18, 2013 10:32 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Handicap 29 prize Ingo and David -- Very nice offer! Some questions : 1. Right now on KGS, I do not see a user active with mfgo1998 as its name. 2. I do not even see a user account listed on the system with that name. 3. Assuming that does get set up, can a person play the mfgo1998 bot themselves, to get an idea of what they're up against? 4. Is there any other way to get a copy of the 1998 version of Many Faces of Go, to spar against on my own PC, for example. Or would that be against the rules? Thanks! -- Mark Goldfain ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Reminder: Handicap 29 Prize
If anyone wants to practice (with Ingo's permission), I can put the version up on KGS if you email me. David (fotl...@smart-games.com) -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Wednesday, January 16, 2013 4:03 AM To: computer-go@dvandva.org Subject: [Computer-go] Reminder: Handicap 29 Prize Reminder: My 1,000 Euro prize for beating the old Many Faces of Go at handicap 29 is still open (and runs until 2020). http://www.althofer.de/handicap-29-prize.html Ingo. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Hardware advice
I like AWS. 16 cores for $3 per hour is pretty good. That's as many as you get with shared memory though. To go higher you would need your program to scale with a network. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Erik van der Werf Sent: Friday, January 11, 2013 1:44 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Hardware advice Personally I'd go for the 6344, or perhaps the 6238, but in general I agree; comparable Intel systems are ridiculously over-priced. Erik On Fri, Jan 11, 2013 at 9:07 PM, Rémi Coulom remi.cou...@free.fr wrote: Hi, We are considering investing into some high-end tournament hardware for Crazy Stone. It has to be a single machine. Right now, I am considering a server such as this one: http://www.thinkmate.com/System/RAX_QS5-4410 With 4 x Sixteen-Core AMD Opteron Model 6386 SE (2.8 GHz) It seems that Intel alternatives with a similar price are much less powerful. Is there anything I am missing? I checked that the clock efficiency of our Intel server is 1.4 better than our AMD bulldozer. I expect that hyperthreading does not help very much because of scalability problems, but I am not so sure. Any recommendation would be welcome. Thanks, Rémi ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] GPU Go engine (re-sent with fixed quotes!!!)
-Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Tuesday, January 08, 2013 8:48 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] GPU Go engine (re-sent with fixed quotes!!!) Hi! On Tue, Jan 08, 2013 at 04:34:15PM +0400, Alexander Kozlovsky wrote: Is it possible to completely ignore superko during playouts and detect it during UTC tree descent only, or maybe, detect only relatively short superko cycles during playouts? I think that probability of long superko cycle are very small and statistically its effect during long series of random playout will be insignificant. Many faces only checks for simple ko in the playouts, and superko in the tree. Short superko cycles might be enough, but this will need careful debugging. (Hidden) superkos may matter surprisingly often in endgame positions, at least on 9x9 in such a way that guarding against superko in some way is essential for good performance. 2) What is a typical data size of structures used in modern Monte-Carlo Go engines? - size of UCT tree node; You will need at least UCT and RAVE values (#of playouts and value or #of wins), move coordinates and say another extra byte for some flags, per child, and pointer to its children list. I have a plan to completely avoid pointers to child nodes. Maybe I don't fully understand something, but if each UCT tree node have unique Zobrist hash and all nodes reside in gigantic hash table, then address of hash table bucket with specific child node can be calculated on the fly by calculating Zobrist hash of child position from parent Zobrist hash. So, it is enough for each node to keep relatively small header of node-specific information and array of UTCValues of chid positions, Many Faces does this (hash table only, no tree). I think also Aya. This is usually called transposition tables. It has its pros and cons; you will save on child pointers and allow merging transposed positions, but you will get all the hash table worries, mainly fixed capacity. There's no clear answer here which is better, I think. But you must have good entropy in your hashes as full collisions are catastrophic. where index in the array corresponds to coordinates on the board. This seems somewhat wasteful, though; by middlegame, you will have a significant overhead, and you will want to have a way to skip unused positions during iteration, which will mean some more overhead. So, I think is is possible to keep size of UTC node less then 1KB for 19x19 board assuming each UTC value in array of child UTC values fit in 2 bytes (I don't sure is 2 bytes is enough for UTC value, but why not). Many Faces is about 5 KB per node. You need to count wins and losses, and rave wins and rave losses at least. I prefer to keep integer counts rather than floating point win rates, to avoid accumulated rounding error. With longer thinking times, even single precision floats aren't enough. Careful here, typically many tree branches may have very similar values. This assume some calculations during tree descent, but I think it is not critical, because I think most processor time will be spend not in tree descent, but in random playouts. Most time is spent in move generation, since the heavy playouts are very nonuniform, and the tree has a prior per move. I find that surprising amount of CPU time is spent in the tree descent (due to cache misses and mathematic operations) in Pachi, but with efficient representation and superscalar evaluation you should be able to push this down. Also, I want to do not not one, but several playouts after each tree descent (I do not know whether this is a common practice), so tree walking time seems even more minor compared to playouts. Not common practice, since it is not a good use of CPU resources. This (or rather a variant of this) is sometimes called leaf parallelisation. It's not been found to work well so far, but I encourage you to do your own experiments. I want to store Zobrist hash values not in 6 KB of shared memory used for playout, but in 64 KB of GPU constant memory, which have 8 KB cache per SM and to me looks very suitable for Zobrist hash storage, because it seems optimized for both of Huh. Isn't the constant memory space shared with the shared memory? (Or worse, offloaded at will to main memory.) By the way, I have read several time already about keeping edges in array together with positions to avoid analyze of position index to determine is it lie near board, but I just don't understand this. Is it really so slow to check if position lie near board? As it seems to me, many board representation (such as bit masks) imply much more performance hit comparing to edge check. At the least, I can have a small shared lookup
Re: [Computer-go] TAAI details?
Many Faces was running on a 16 core Xeon (AWS Cluster Compute 8x extra large), using 16 threads, 48 GB of memory. It started well in both handicap games, then fell apart in the endgame. In the main 19x19 tournament, ManyFaces beat Zen once. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Tuesday, November 20, 2012 10:43 AM To: computer-go@dvandva.org Subject: [Computer-go] TAAI details? Hello, during the TAAI 2012 conference several exhibition games between pro players and top bots were played. Here seem to be the sgf of the games on 19x19: Zen - two games won with handicap 4 (against 9p Chun-Hsun Chou) http://www.gokgs.com/gameArchives.jsp?user=taai1 Perhaps Zen should get some shot with handicap 3 in next year. ManyFaces - one game lost with handicap 4 and one lost with handicap 5 http://www.gokgs.com/gameArchives.jsp?user=taai2 Aya - one game lost with handicap 4, one game one (on time) with handicap 5 http://www.gokgs.com/gameArchives.jsp?user=taai3 Can the programmers please give some detail information? Thanks in advance, Ingo. PS-1. It seems that the time setting (45 min sudden death) can lead to desasters ... Perhaps for future events some (mini) byoyomi might be a good solution. PS-2. Nick Wedds list also shows some forthcoming events. http://www.computer-go.info/h-c/index.html ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Develop a client for KGS?
KGS does not allow client development. The protocol is proprietary. KGS already has a fine, working web client, so there is no need for another one. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Thursday, November 08, 2012 5:09 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Develop a client for KGS? I thought KGS had their own proprietary protocol? GTP is the protocol for communicating with a Go engine, but not KGS as far as I know. KGS supports many features so if they used that protocol it would have to be significantly extended. Don On Thu, Nov 8, 2012 at 5:01 AM, Lukas van de Wiel lukas.drinkt.t...@gmail.com wrote: Hi Henry, you are probably interested in reading about the GTP. It is the protocol used to connect to KGS and it is well documented here: http://www.lysator.liu.se/~gunnar/gtp/ A lot of work already done with it can freely be downloaded and viewed Good luck with programming the client! Lukas On Thu, Nov 8, 2012 at 10:49 AM, Henry Hu henryhu...@gmail.com wrote: Hi folks, I'm planning to develop a web client for KGS. Surely it is necessary to connect to it firstly. Searching online, it seems lack of info related. How can my client connect to KGS and communicate with it. Are there any tools or introduction available for it? Thanks in advance. Regards, Henry ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] An unusual seki
At 305 Many Faces plays T9 with 96% win rate. The old traditional Many Faces of Go engine statically recognizes this seki, and the MCTS playouts understand it. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Ken Friedenbach Sent: Wednesday, August 15, 2012 8:45 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] An unusual seki For Black, the losing move was 305: B T7. Black at T9 would give atari on 4 stones and break the seki in Black's favor. But after T7, Black is short of liberties, and Seki is the best that can be achieved. (Bad move counting after Pass occurs. The count at top is only non-pass moves, I guess...) On Aug 15, 2012, at 10:22 AM, David Ongaro wrote: Am 15.08.2012 um 18:35 schrieb Michael Williams: First time using this. Not sure how long the link will live. http://eidogo.com/#sk815dh Go to the end of the game. It unfortunate that it is so small, ugly and that you can't link directly to a certain move. It seems we are in need of a good web SGF viewer. Hm, it seems to me that http://eidogo.com/#sk815dh:0,404 works fine as a direct link. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go Ken Friedenbach ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
Just curious... I understand how you update the counters lock-free, but surely you must have a lock to protect adding a new node to the tree? Do this impact scaling at some point? Are there any other infrequent locks like this? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Friday, August 10, 2012 12:47 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes On Fri, Aug 10, 2012 at 09:26:31AM -0700, David Fotland wrote: Because my current approach seems to work just as well (or maybe better), and I haven't had time to code up a shared try and tune it up to validate that assumption. Chaslot's paper indicates perhaps that not having a shared tree is stronger. My guess is that they are about the same, so it's not worth the effort to change. In Pachi, having a shared tree makes all the difference when scaling up to more threads. See the graph (really awful one, sorry, it's old!) at http://pachi.or.cz/root-vs-shared.png If you have some information sharing near the root, I imagine it might be similar to Pachi's distributed engine performance (or just slightly better). But that is still far behind in scaling compared to the shared tree in our experience. P.S.: There are two important things, virtual loss (not necessarily 1 simulation but possibly more) and mainly lockless updates. The latter also means that sane code should be really easy to modify to use single shared tree instead of multiple trees. Petr Pasky Baudis ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
I've been using this abort early to save time idea almost since the beginning. It works fine with root parallelization. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Mark Boon Sent: Sunday, August 12, 2012 9:07 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes Interesting, I'd have thought it would matter quite bit, especially with higher numbers of threads. One thing I found (quite a few years back now already) is that you can optimize a lot by doing the following: when one node has so many more wins than the second best that it can't be overtaken even if the second best wins all of the remaining playouts, abort thinking. With a couple of extensions to this general idea (aborting not just when it's impossible, just very unlikely to be overtaken) I found that a player that does 64K lightweight simulations using this method spends the same time and plays the same level as one that does a fixed 32K simulations. Roughly. The higher the number of simulations, the bigger the savings. This type of optimization must be much harder with root-level parallelization, so you'd have to factor that in when comparing methods. Mark On Aug 10, 2012, at 9:55 PM, David Fotland fotl...@smart-games.com wrote: Not much memory overhead. If you look at your tree you will find that most nodes are only visited one or two times. There is a lot of noise in the fringes of the tree, so there are few duplicates. This also means that not sharing most of the tree has no impact on strength. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, August 10, 2012 9:42 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I imagine you can get around the lack of implicit information sharing that you get with a shared tree by explicitly sharing information near the root. But doesn't having separate trees mean a large memory overhead due to duplicate nodes? On Fri, Aug 10, 2012 at 9:26 AM, David Fotland fotl...@smart-games.com wrote: Because my current approach seems to work just as well (or maybe better), and I haven't had time to code up a shared try and tune it up to validate that assumption. Chaslot's paper indicates perhaps that not having a shared tree is stronger. My guess is that they are about the same, so it's not worth the effort to change. david -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, August 10, 2012 12:06 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes Why don't you use a shared tree? On Thu, Aug 9, 2012 at 11:49 PM, David Fotland fotl...@smart-games.com wrote: On an i7-2600 Many Faces does 11.4K pps with 4 threads, and 18.7k with 8 threads, a 64% increase, so the 2600 scales a little better than the 3770, but the 3770 is still a litte bit faster. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Erik van der Werf Sent: Thursday, August 09, 2012 4:41 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I don't have an i7-2600, but I could run oakfoam on the 3930. I just downloaded it and it does compile. If you give me a list of gtp commands to run the benchmark, then I will send you the output back. Erik On Thu, Aug 9, 2012 at 12:38 PM, ds d...@physik.de wrote: This is very interesting, I have not more than 10% with oakfoam on i7-2600K. Would be interesting if it is the processor or if you e.g. access more often memory instead of cache due to your code... Do you have the chance to run your program on a i7-2600? or do you have to much time and try https://bitbucket.org/francoisvn/oakfoam/wiki/Home on your i7-3930. If so, I would be very much interested in the number you get in the beginning of a 19x19 game without book:) Detlef Am Donnerstag, den 09.08.2012, 12:16 +0200 schrieb Erik van der Werf: On Thu, Aug 9, 2012 at 11:14 AM, Petr Baudis pa...@ucw.cz wrote: On Wed, Aug 08, 2012 at 09:08:47PM +0200, ds wrote: Hyperthreading does the trick, I have the experience it increases the performance by about 10%. I think this is due to waiting for RAM I/O or things like that Yes. With hyperthreading, performance per thread goes down significantly, but total performance goes up by about 15%. In the Pentium 4 era, hyperthreading did not usually pay off, but with i7, its performance is much better. The basic idea
Re: [Computer-go] Kas Cup - results and prizes
I'm happy with MFGO's scaling. I'm running a scaling test now, 4 threads vs 8 threads, fixed 32K total playouts per move, 19x19, no pondering. Ideally the win rate should be 50%, since the total playouts are the same. Has anyone tried this kind of scaling experiment, and is willing to share results? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Friday, August 10, 2012 12:47 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes On Fri, Aug 10, 2012 at 09:26:31AM -0700, David Fotland wrote: Because my current approach seems to work just as well (or maybe better), and I haven't had time to code up a shared try and tune it up to validate that assumption. Chaslot's paper indicates perhaps that not having a shared tree is stronger. My guess is that they are about the same, so it's not worth the effort to change. In Pachi, having a shared tree makes all the difference when scaling up to more threads. See the graph (really awful one, sorry, it's old!) at http://pachi.or.cz/root-vs-shared.png If you have some information sharing near the root, I imagine it might be similar to Pachi's distributed engine performance (or just slightly better). But that is still far behind in scaling compared to the shared tree in our experience. P.S.: There are two important things, virtual loss (not necessarily 1 simulation but possibly more) and mainly lockless updates. The latter also means that sane code should be really easy to modify to use single shared tree instead of multiple trees. Petr Pasky Baudis ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
Yes, root parallelization with some sharing. http://www.personeel.unimaas.nl/G-Chaslot/papers/parallelMCTS.pdf said it was good and I tried it and it works well. Hardware is really important. But so are really smart playouts. The slower I make my playouts the stronger the program gets. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Peter Drake Sent: Friday, August 10, 2012 10:45 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes On Thu, Aug 9, 2012 at 11:42 PM, David Fotland fotl...@smart-games.com wrote: Or it might be an artifact of the way I do search, since I think I might be the only engine that doesn't use a single shared tree, and the old Many Faces of Go engine is single threaded. If not a single shared tree, what are you doing? Root parallelism? I've been wondering why other programs are pulling ahead of Orego, and now I'm starting to suspect the answer may be (in part) hardware. -- Peter Drake https://sites.google.com/a/lclark.edu/drake/ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
Not much memory overhead. If you look at your tree you will find that most nodes are only visited one or two times. There is a lot of noise in the fringes of the tree, so there are few duplicates. This also means that not sharing most of the tree has no impact on strength. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, August 10, 2012 9:42 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I imagine you can get around the lack of implicit information sharing that you get with a shared tree by explicitly sharing information near the root. But doesn't having separate trees mean a large memory overhead due to duplicate nodes? On Fri, Aug 10, 2012 at 9:26 AM, David Fotland fotl...@smart-games.com wrote: Because my current approach seems to work just as well (or maybe better), and I haven't had time to code up a shared try and tune it up to validate that assumption. Chaslot's paper indicates perhaps that not having a shared tree is stronger. My guess is that they are about the same, so it's not worth the effort to change. david -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, August 10, 2012 12:06 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes Why don't you use a shared tree? On Thu, Aug 9, 2012 at 11:49 PM, David Fotland fotl...@smart-games.com wrote: On an i7-2600 Many Faces does 11.4K pps with 4 threads, and 18.7k with 8 threads, a 64% increase, so the 2600 scales a little better than the 3770, but the 3770 is still a litte bit faster. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Erik van der Werf Sent: Thursday, August 09, 2012 4:41 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I don't have an i7-2600, but I could run oakfoam on the 3930. I just downloaded it and it does compile. If you give me a list of gtp commands to run the benchmark, then I will send you the output back. Erik On Thu, Aug 9, 2012 at 12:38 PM, ds d...@physik.de wrote: This is very interesting, I have not more than 10% with oakfoam on i7-2600K. Would be interesting if it is the processor or if you e.g. access more often memory instead of cache due to your code... Do you have the chance to run your program on a i7-2600? or do you have to much time and try https://bitbucket.org/francoisvn/oakfoam/wiki/Home on your i7-3930. If so, I would be very much interested in the number you get in the beginning of a 19x19 game without book:) Detlef Am Donnerstag, den 09.08.2012, 12:16 +0200 schrieb Erik van der Werf: On Thu, Aug 9, 2012 at 11:14 AM, Petr Baudis pa...@ucw.cz wrote: On Wed, Aug 08, 2012 at 09:08:47PM +0200, ds wrote: Hyperthreading does the trick, I have the experience it increases the performance by about 10%. I think this is due to waiting for RAM I/O or things like that Yes. With hyperthreading, performance per thread goes down significantly, but total performance goes up by about 15%. In the Pentium 4 era, hyperthreading did not usually pay off, but with i7, its performance is much better. The basic idea is that there are two instruction pipelines that share the same ALU and other processor units; if one of the pipelines stalls (usually due to memory fetch), the other can use the ALU in the meantime, or the two threads may use different parts of the CPU altogether based on what the instructions do. 10-15%, really, that low? For my program (on an i7-3930K, going from 6 to 12 threads) it is more in the order of 40% extra simulations per second. Erik ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
On my core i7-3770, 4 threads is 12.5K playouts/sec (19x19, average of first four moves by white), and 8 threads is 19.8K, a 58% increase. This is much higher than I expected. It seems Intel has improved hyperthreading since the last time I tried it. Or it might be an artifact of the way I do search, since I think I might be the only engine that doesn't use a single shared tree, and the old Many Faces of Go engine is single threaded. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Thursday, August 09, 2012 6:23 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes On Thu, Aug 09, 2012 at 08:09:55PM +0900, Hideki Kato wrote: Erik van der Werf: CAKkgGrM83_HsQ5Z2HJupkj=gDeh3+4GM- jmlvevtjroufqn...@mail.gmail.com: 10-15%, really, that low? For my program (on an i7-3930K, going from 6 to 12 threads) it is more in the order of 40% extra simulations per second. In general that number highly depends on the code, architecture of the processor (Intel's are usually better than AMD's), memory speed, cache size, use of ALUs, etc. For Zen, the number is also about 40% on both an i7 3930K (6 to 12 threads) and an i7 920 (4 to 8 threads). For Zen, I'm not surprised, since I assume that in simulations, you are matching some larger patterns which involves a lot of time-consuming hash table lookups which is ideal for hyperthreading. Not sure about stv. I think it matters a lot on whether you are matching patterns by explicit test code snippets or by a hash table. I measured the hyperthreading effect about 2 years ago with a lot older Pachi version. I think today, the hyperthreading effect would also be higher, but I cannot test it right now. Pasky, modern processors are much more complicated :). There are more than two sets of general registers, which are used not only for hyperthreading but also register renaming, for example. Sure, I just tried to sketch a rough explanation. I did not know that hyperthreading could reduce opportunity for register renaming, though. Petr Pasky Baudis ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
On an i7-2600 Many Faces does 11.4K pps with 4 threads, and 18.7k with 8 threads, a 64% increase, so the 2600 scales a little better than the 3770, but the 3770 is still a litte bit faster. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Erik van der Werf Sent: Thursday, August 09, 2012 4:41 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I don't have an i7-2600, but I could run oakfoam on the 3930. I just downloaded it and it does compile. If you give me a list of gtp commands to run the benchmark, then I will send you the output back. Erik On Thu, Aug 9, 2012 at 12:38 PM, ds d...@physik.de wrote: This is very interesting, I have not more than 10% with oakfoam on i7-2600K. Would be interesting if it is the processor or if you e.g. access more often memory instead of cache due to your code... Do you have the chance to run your program on a i7-2600? or do you have to much time and try https://bitbucket.org/francoisvn/oakfoam/wiki/Home on your i7-3930. If so, I would be very much interested in the number you get in the beginning of a 19x19 game without book:) Detlef Am Donnerstag, den 09.08.2012, 12:16 +0200 schrieb Erik van der Werf: On Thu, Aug 9, 2012 at 11:14 AM, Petr Baudis pa...@ucw.cz wrote: On Wed, Aug 08, 2012 at 09:08:47PM +0200, ds wrote: Hyperthreading does the trick, I have the experience it increases the performance by about 10%. I think this is due to waiting for RAM I/O or things like that Yes. With hyperthreading, performance per thread goes down significantly, but total performance goes up by about 15%. In the Pentium 4 era, hyperthreading did not usually pay off, but with i7, its performance is much better. The basic idea is that there are two instruction pipelines that share the same ALU and other processor units; if one of the pipelines stalls (usually due to memory fetch), the other can use the ALU in the meantime, or the two threads may use different parts of the CPU altogether based on what the instructions do. 10-15%, really, that low? For my program (on an i7-3930K, going from 6 to 12 threads) it is more in the order of 40% extra simulations per second. Erik ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kaœ Cup - results and prizes
It is far less hardware to add a thread than to add a whole extra core. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Richard Lorentz Sent: Wednesday, August 08, 2012 10:22 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kaœ Cup - results and prizes This surprises me. Isn't it inefficient to have one core run multiple- threads? Can you explain why you did it this way to a hardware challenged person like me. Thanks. On 08/07/2012 03:35 AM, Petr Baudis wrote: Program authors / operators, please state in this thread the hardware used. Intel(R) Core(TM) i7-2600 CPU @ 3.40GHz. It seems it is auto-overlocked to 3.7GHz. Four cores, eight threads. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup - results and prizes
Because my current approach seems to work just as well (or maybe better), and I haven't had time to code up a shared try and tune it up to validate that assumption. Chaslot's paper indicates perhaps that not having a shared tree is stronger. My guess is that they are about the same, so it's not worth the effort to change. david -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, August 10, 2012 12:06 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes Why don't you use a shared tree? On Thu, Aug 9, 2012 at 11:49 PM, David Fotland fotl...@smart-games.com wrote: On an i7-2600 Many Faces does 11.4K pps with 4 threads, and 18.7k with 8 threads, a 64% increase, so the 2600 scales a little better than the 3770, but the 3770 is still a litte bit faster. david From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Erik van der Werf Sent: Thursday, August 09, 2012 4:41 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup - results and prizes I don't have an i7-2600, but I could run oakfoam on the 3930. I just downloaded it and it does compile. If you give me a list of gtp commands to run the benchmark, then I will send you the output back. Erik On Thu, Aug 9, 2012 at 12:38 PM, ds d...@physik.de wrote: This is very interesting, I have not more than 10% with oakfoam on i7-2600K. Would be interesting if it is the processor or if you e.g. access more often memory instead of cache due to your code... Do you have the chance to run your program on a i7-2600? or do you have to much time and try https://bitbucket.org/francoisvn/oakfoam/wiki/Home on your i7-3930. If so, I would be very much interested in the number you get in the beginning of a 19x19 game without book:) Detlef Am Donnerstag, den 09.08.2012, 12:16 +0200 schrieb Erik van der Werf: On Thu, Aug 9, 2012 at 11:14 AM, Petr Baudis pa...@ucw.cz wrote: On Wed, Aug 08, 2012 at 09:08:47PM +0200, ds wrote: Hyperthreading does the trick, I have the experience it increases the performance by about 10%. I think this is due to waiting for RAM I/O or things like that Yes. With hyperthreading, performance per thread goes down significantly, but total performance goes up by about 15%. In the Pentium 4 era, hyperthreading did not usually pay off, but with i7, its performance is much better. The basic idea is that there are two instruction pipelines that share the same ALU and other processor units; if one of the pipelines stalls (usually due to memory fetch), the other can use the ALU in the meantime, or the two threads may use different parts of the CPU altogether based on what the instructions do. 10-15%, really, that low? For my program (on an i7-3930K, going from 6 to 12 threads) it is more in the order of 40% extra simulations per second. Erik ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kaœ Cup - results and prizes
Thank you for an interesting tournament. ManyFaces was running on an Amazon m2.4xlarge, a Xeon x5550 2.67 GHz. My tests show this to be the same speed as my core i7-3770, but the EC2 has much more memory installed. I worked on a no-resign preserve score version, but didn’t finish it in time, so I entered the same program as last month's KGS tournament. It has dynamic komi, but only to preserve extra points up to 20 points when it is winning. It resigns when the win rate is 5% or less and the score estimate is sufficiently negative and no large threats are available. You can use paypal to send money to fotl...@smart-games.com -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Lukasz Lew Sent: Monday, August 06, 2012 5:09 PM To: computer-go@dvandva.org Subject: [Computer-go] Kaś Cup - results and prizes Kaś Cup has finished with only one game ending with scoring.non-resign. Here are the results: Round 1. Aya : Zen -50 : 50 Many Faces : Pachi 2.5 : -2.5 Round 2. Zen : Many Faces 50 : -50 Aya : Pachi 50 : -50 Round 3. Zen : Pachi 50 : -50 Aya : Many Faces 50 : -50 Total scores: Zen 150 100$ Aya 50 60$ Many Faces -97.5 40$ Pachi -102.5 Prizes Zen 100$ Aya 60$ Many Faces 40$ Notably at least one game (Round 2 Aya : Pachi) ended with premature resignation at final position. Program authors / operators, please state in this thread the hardware used. If you wish, tell us whether and how you modified your programs. Winners, please write me how can I award you the prizes. (PayPal?) Congratulations to the winners! Łukasz ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup
I misread the prize pool. Thought it was $1,000 :) Should be fun. Count me in. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Lukasz Lew Sent: Wednesday, July 11, 2012 9:43 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Kas Cup On Wed, Jul 11, 2012 at 6:08 PM, David Fotland fotland@smart- games.com wrote: I didn t get this post either. There needs to be a way to score a resign, perhaps give 50 points to the winner. MCTS bots that resign when they are only a few points behind will skew the results in a random way. It might be better to penalize the resigner more than the gain given to the winner. Good point. Resignation will be treated as 50/-50 result. So is timeout and forfeit. With money involved, there needs to be a way to verify the number of cores used. I don t know any reasonable way to do this. The money isn't huge and I don't know any author of the program that would attempt cheating here. I'm assuming that hyperthreading is OK, so a core i7 would count as 4 cores, but use 8 threads. Yes. If you hesitate whether your program is worth to be modified, let me tell you this: If this tournament will be popular and considered fun / useful , I am considering sponsoring another edition. Also tunning for score is much quicker than for winning rate :) Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Wednesday, July 11, 2012 3:00 AM To: computer-go@dvandva.org Subject: [Computer-go] Kas Cup Lukasz Lew has today posted a message to this list, which I have not received. I know he has posted it, I can see his posting in the list archive, at http://dvandva.org/pipermail/computer-go/2012- July/005168.html I encourage you to read it, it is an invitation to a bot tournament with bangneki-type scoring. This is the first time I have been aware of not receiving a message which has definitely been posted. I have checked my spam bucket. And I have read roveg's statement that nothing about the list has changed. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go -- Lukasz ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Kas Cup
I didn’t get this post either. There needs to be a way to score a resign, perhaps give 50 points to the winner. MCTS bots that resign when they are only a few points behind will skew the results in a random way. It might be better to penalize the resigner more than the gain given to the winner. With money involved, there needs to be a way to verify the number of cores used. I don’t know any reasonable way to do this. I'm assuming that hyperthreading is OK, so a core i7 would count as 4 cores, but use 8 threads. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Wednesday, July 11, 2012 3:00 AM To: computer-go@dvandva.org Subject: [Computer-go] Kas Cup Lukasz Lew has today posted a message to this list, which I have not received. I know he has posted it, I can see his posting in the list archive, at http://dvandva.org/pipermail/computer-go/2012- July/005168.html I encourage you to read it, it is an invitation to a bot tournament with bangneki-type scoring. This is the first time I have been aware of not receiving a message which has definitely been posted. I have checked my spam bucket. And I have read roveg's statement that nothing about the list has changed. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] TCGA 19x19 Computer Go Tournament, error report
It looks like in rounds 10-15 every match timed out, which put the tournament back on schedule. These matches should not be counted in the official totals. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Saturday, June 30, 2012 4:19 PM To: computer-go@dvandva.org Subject: [Computer-go] TCGA 19x19 Computer Go Tournament, error report This report is about incidents in the ongoing TCGA 19x19 Computer Go Tournament, now running on KGS. I am writing this for the record, so that the players and organisers will be aware of all the information that I have. These statements are made for the record. I hope I can be regarded as impartial. I am only making statements of fact anyway. __Incident 1. During round 2, I removed 'ntnu' from the tournament, at the request of its operator. I hoped that this would reduce the length of the tournament from 36 rounds to a more manageable 30, which would allow it to end before the start of the July KGS tournament. However, it seems that that is not how the scheduler for a multiple- round-robin works. Round 10, with eight players, has three games and two byes, rather than the four games that one might expect. I assume that the entire schedule was drawn up in advance, and one of the two byes was scheduled as a game against the now-absent 'ntnu'. I do not know how to avoid a clash with the schedule of the KGS July bot tournament. I am now too tired to think sensibly about it, and will go to bed soon. I will leave 'guxxan' to think about it. __Incident 2. During round 3, 'ajahuang' saw that Fuego19, which he was operating, had a hopeless position against nomitan440, and proposed logging in and resigning for it. If I had been more alert, I would have realised what was likely to happen, and told him not to; but I said nothing. He logged in to Fuego19's account, and resigned for it. The KGS tournament system observed a human using a bot's account in a bot tournament, regarded this as an attempt to cheat, and marked the game as lost by Forfeit. Thus this round 2 game has two different results: KGS sees it as lost by Fuego19 by resignation, while the KGS tournament system sees it as lost by Fuego19 by Forfeit. There is no significant problem so far, this outcome was what I had expected. However, the tournament was running behind the optimistic 25-minutes-a- round specified when it was set up. So when this round 2 game ended (all the other round 2 games had already ended), round 3 began immediately. And ajahuang was still logged in to Fuego19's account. So Fuego19's round 3 game (against AyaMC5) was also treated, by the tournament system, as lost by Fuego19, by Forfeit. But the game continued to be played. When the other round 3 games were finished, round 4 started (the tournament system regarding the Fuego19/AyaMC5 game as already decided). I realised that Fuego19 and AyaMC we still playing in their round 3 game and had not joined their round 4 games, so, to free them up and let them play in round 4, I used my admin power to kill this game. I believe that Fuego19 had been winning it at the time. I killed it rather too slowly, so that when AyaMC5 joined its round 4 game against Zen19S, it has already lost one of its 15-second overtime periods. I apologise to AyaMC5's operator for this. Fuego19 did not suffer loss of an overtime period, as it had a bye in round 4. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] TCGA 19x19 Computer Go Tournament, error report
Perhaps the best thing to do is to cancel the 19x19 and schedule it again later? regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, June 30, 2012 6:15 PM To: n...@maproom.co.uk; computer-go@dvandva.org Subject: Re: [Computer-go] TCGA 19x19 Computer Go Tournament, error report It looks like in rounds 10-15 every match timed out, which put the tournament back on schedule. These matches should not be counted in the official totals. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Saturday, June 30, 2012 4:19 PM To: computer-go@dvandva.org Subject: [Computer-go] TCGA 19x19 Computer Go Tournament, error report This report is about incidents in the ongoing TCGA 19x19 Computer Go Tournament, now running on KGS. I am writing this for the record, so that the players and organisers will be aware of all the information that I have. These statements are made for the record. I hope I can be regarded as impartial. I am only making statements of fact anyway. __Incident 1. During round 2, I removed 'ntnu' from the tournament, at the request of its operator. I hoped that this would reduce the length of the tournament from 36 rounds to a more manageable 30, which would allow it to end before the start of the July KGS tournament. However, it seems that that is not how the scheduler for a multiple- round-robin works. Round 10, with eight players, has three games and two byes, rather than the four games that one might expect. I assume that the entire schedule was drawn up in advance, and one of the two byes was scheduled as a game against the now-absent 'ntnu'. I do not know how to avoid a clash with the schedule of the KGS July bot tournament. I am now too tired to think sensibly about it, and will go to bed soon. I will leave 'guxxan' to think about it. __Incident 2. During round 3, 'ajahuang' saw that Fuego19, which he was operating, had a hopeless position against nomitan440, and proposed logging in and resigning for it. If I had been more alert, I would have realised what was likely to happen, and told him not to; but I said nothing. He logged in to Fuego19's account, and resigned for it. The KGS tournament system observed a human using a bot's account in a bot tournament, regarded this as an attempt to cheat, and marked the game as lost by Forfeit. Thus this round 2 game has two different results: KGS sees it as lost by Fuego19 by resignation, while the KGS tournament system sees it as lost by Fuego19 by Forfeit. There is no significant problem so far, this outcome was what I had expected. However, the tournament was running behind the optimistic 25-minutes-a- round specified when it was set up. So when this round 2 game ended (all the other round 2 games had already ended), round 3 began immediately. And ajahuang was still logged in to Fuego19's account. So Fuego19's round 3 game (against AyaMC5) was also treated, by the tournament system, as lost by Fuego19, by Forfeit. But the game continued to be played. When the other round 3 games were finished, round 4 started (the tournament system regarding the Fuego19/AyaMC5 game as already decided). I realised that Fuego19 and AyaMC we still playing in their round 3 game and had not joined their round 4 games, so, to free them up and let them play in round 4, I used my admin power to kill this game. I believe that Fuego19 had been winning it at the time. I killed it rather too slowly, so that when AyaMC5 joined its round 4 game against Zen19S, it has already lost one of its 15-second overtime periods. I apologise to AyaMC5's operator for this. Fuego19 did not suffer loss of an overtime period, as it had a bye in round 4. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] New programmer
Correct scoring of arbitrary positions is very difficult because of the life and death analysis required. You don't say if you want to score final positions from human games, or any arbitrary positions during a game, or only positions that are finished under CGOS or similar rules, where it is safe to assume that stones on the board are alive. The last one is trivial to score. The others are very difficult. That's why Monte carlo works so well, since it only tries to score trivial positions. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nicolas FRANCOIS Sent: Sunday, May 27, 2012 10:36 AM To: computer-go@dvandva.org Subject: [Computer-go] New programmer Hi. I've been studying computer go for a while now, and would like to experiment on some ideas. I have one (well, in fact, two) big problem though : I can't figure out how to write a correct scoring procedure, which, I think, is linked to the problem of life and death. Could you give me some advices on readings on those subjects (especially deciding life and death), or some examples of well written codes on the same subject ? Thank you. \bye -- Nicolas FRANCOIS | /\ http://nicolas.francois.free.fr | |__| X--/\\ We are the Micro$oft. _\_V Resistance is futile. You will be assimilated. darthvader penguin ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] A Regression test set for exploring some limitations of current MCTS programs in Go
I used gogui-adapter too because many faces doesn't have loadsgf, but gogui doesn't send the komi, so I had to adjust it by hand. From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Huang Sent: Wednesday, May 16, 2012 8:53 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] A Regression test set for exploring some limitations of current MCTS programs in Go By the way, to use gogui-adapter to translate 'loadsgf' the command is something like ./run.sh -p java -jar gogui-adapter.jar \PATH_TO_PROGRAM \ -t g_seki_moves.tst (use backslash character (\) to escape the quotes in the string) I used gogui-adapter to run pachi and Mogo as well because they both don't support 'loadsgf'. Please don't hesitate to let me know if it doesn't work for you. Best regards, Aja ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] cgos
I have some free cores on my kgs machine. I'll put up at lease 10k and 50k playout Many Faces on 19x19 this evening. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Monday, April 23, 2012 10:53 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] cgos I have always wished that we could keep a range of programs with different strengths always playing, but that requires computer resources. If I was wealthy I would donate a 6 core machine solely for this purpose - but I am not :-) A single machine with a few cores could keep a few programs always on-line ready to play. Don On Mon, Apr 23, 2012 at 1:48 PM, ds d...@physik.de wrote: Hi, would be nice, if one could arrange a typical date, where one can hope to find more programs on cgos. E.g. once a week or once every two weeks on monday? Depending on the opponents one could let the programs run for some hours or some days. For the last weeks it was quite empty with only the standard programs running. You can only get rated around 1800ELO with some accuracy... Detlef ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] cgos
I meant 2K and 10K J (per move, not per game). David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of David Fotland Sent: Tuesday, April 24, 2012 8:57 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] cgos I have some free cores on my kgs machine. I'll put up at lease 10k and 50k playout Many Faces on 19x19 this evening. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Monday, April 23, 2012 10:53 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] cgos I have always wished that we could keep a range of programs with different strengths always playing, but that requires computer resources. If I was wealthy I would donate a 6 core machine solely for this purpose - but I am not :-) A single machine with a few cores could keep a few programs always on-line ready to play. Don On Mon, Apr 23, 2012 at 1:48 PM, ds d...@physik.de wrote: Hi, would be nice, if one could arrange a typical date, where one can hope to find more programs on cgos. E.g. once a week or once every two weeks on monday? Depending on the opponents one could let the programs run for some hours or some days. For the last weeks it was quite empty with only the standard programs running. You can only get rated around 1800ELO with some accuracy... Detlef ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] April KGS bot tournament: 13x13 boards, slow
I changed the room name and it works. ManyFaces1 still might crash or hang. If so, please kick it, and it will restart immediately. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Sunday, April 01, 2012 12:14 AM To: Hiroshi Yamashita Cc: computer-go@dvandva.org Subject: Re: [Computer-go] April KGS bot tournament: 13x13 boards, slow On 01/04/2012 06:41, Hiroshi Yamashita wrote: Hi, My bot failed to login in KGS. kgsGtp says, Client cannot find room to join. Logging out. It seems Computer Go room name has changed to Confuser Go by April fool. Do I have to rewrite room name for today's KGS tournament? This may be a problem. I shall start thinking about it - I have just woken up. Nick Regards, Hiroshi Yamashita - Original Message - From: Nick Wedd n...@maproom.co.uk To: computer-go@dvandva.org Sent: Saturday, March 31, 2012 5:31 PM Subject: Re: [Computer-go] April KGS bot tournament: 13x13 boards, slow Reminder: it's tomorrow. Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Computer go for iPhone
The free Igowin is artificially weak. Igowin Pro supports 19x19 on iPhone. It's not as strong as Many Faces on the PC due to memory limitations, but it is still quite strong. Igowin HD for the iPad is stronger. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Jason House Sent: Friday, March 23, 2012 6:06 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Computer go for iPhone Isn't igowin 9x9 only? When I beta tested it, the monte carlo engine was artificially weak in order to maintain fast response times / low resource usage. It was borderline too weak at 9x9 and I suspect it would be too weak at 19x19. I understand that the Many Faces engine can be considerably stronger given enough resources, but how strong is it on the iPhone? Sent from my iPhone On Mar 23, 2012, at 1:42 AM, David Fotland fotl...@smart-games.com wrote: The iPhone igowin apps have the Many Faces of Go engine. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ray Tayek Sent: Thursday, March 22, 2012 6:46 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Computer go for iPhone At 06:14 PM 3/22/2012, you wrote: I'm trying to buy a go AI for my iPhone. I'm kgs 3k and am trying to figure out what my options are. i suspect that http://www.smartgo.com/ would be one of the better ones having been around for a while. it does not seem to play 19x18 though. thanks --- co-chair http://ocjug.org/ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Computer go for iPhone
Smart Go Kifu is from Anders Kierulf, not me. My company is Smart Games :) Smart Go Kifu does not use the Many Faces engine. He has his own engine, which is quite a bit weaker. Anders invented the SGF format, and his software has always had exceptionally good SFG editors. -David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Jouni Valkonen Sent: Friday, March 23, 2012 12:43 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Computer go for iPhone To be fair in commercial issues. For Android phone, Lauri Paatero published new AI and good SGF-editor called Hactar. The AI is server based so it needs an Internet connection to play. For iPhone and iPad, top AI is R�mi's CrazyStone (called Champion Go). I have hard time to beat it. Champion Go is also available for Android. However David's Smart Go Kifu is probably the best SGF-editor and viewer for iPad and iPhone. It also comes with rather good ManyFaces engine, but I have not tested the iPhone version. However iPad version is something that I can highly recommend and as a SGF-editor and viewer I doubt that there is even remotely as good SGF software available for iOS. �Jouni Sent from my iPad On 23 Mar 2012, at 03:14, Jason House jason.james.ho...@gmail.com wrote: I'm trying to buy a go AI for my iPhone. I'm kgs 3k and am trying to figure out what my options are. Which programs support which board sizes? How strong are they on those boards? A built in SGF viewer that can accept URL to load would be a nice bonus. Sent from my iPhone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Computer go for iPhone
The iPhone igowin apps have the Many Faces of Go engine. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ray Tayek Sent: Thursday, March 22, 2012 6:46 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Computer go for iPhone At 06:14 PM 3/22/2012, you wrote: I'm trying to buy a go AI for my iPhone. I'm kgs 3k and am trying to figure out what my options are. i suspect that http://www.smartgo.com/ would be one of the better ones having been around for a while. it does not seem to play 19x18 though. thanks --- co-chair http://ocjug.org/ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] oakfoam/Zen9 game
Thanks Nick. I've been using 3.5.0, so I also need to update. I see the latest is 3.5.4. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Tuesday, February 14, 2012 1:51 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] oakfoam/Zen9 game On 13/02/2012 10:36, Yamato wrote: A similar thing happened in past tournaments, and we sent the log file to wms that time. I guess he doesn't fix this bug yet. I have now had a reply from wms. He wrote Both of these players were on version 3.5.0 of the GTP client. This is a fixed bug. If Zen9 had had version 3.5.3 or later, then it would not have lost. So, everyone, please download and use the latest version of kgsGtp.jar. It is at http://www.gokgs.com/download.jsp Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Handicap-29 Website is up now
The 29 stone game wasn't really a demonstration game. My memory is that I had just finished igowin 9x9, and Martin played through all its levels up to professional, without losing a game. I was surprised, and while we were talking the subject of a 29 stone game on 19x19 came up. I dont remember if I challenged him or if he said he could beat 19x19 on that handicap. So we made a very small bet (I dont remember, maybe $5), on the game and he played. I was very confident Many Faces would win and was really surprised when it didn't. I dont remember if we announced the game before it was played. I think we didnt, but there were a few people watching. I dont remember how the game record got out to the internet. Martin, how do you remember it? It was a long time ago... regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Friday, February 10, 2012 1:42 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Handicap-29 Website is up now Hello, after meditating carefully about all the constructive feedback on the original announcement of a handicap-29 prize, now the website for the prize is there: http://www.althofer.de/handicap-29-prize.html Please, let me know when you find errors or inconsistencies in the text. Thanks in advance, Ingo. Original-Nachricht ... Now I set up a prize for the programmer of a bot that beats the old ManyFaces at 29 handicap stones: 1,000 Euro for the first bot that achieves this at least three times in a five games match. The offer ends on December 31, 2020. Details will be clarified in a forthcoming website. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Handicap-29 Website is up now
mfgo1998 is set up now to put the 29 handicap stones in exactly the same place every time, to reproduce the conditions of Martin's game. The random option will prevent it from making the same moves. Do you want me to change it so it generates a different setup each time? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Friday, February 10, 2012 9:18 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Handicap-29 Website is up now Hi David, The 29 stone game wasn't really a demonstration game. My memory is that I had just finished igowin 9x9, and Martin played through all its levels up to professional, without losing a game... I don�t remember if I challenged him or if he said he could beat 19x19 on that handicap. So we made a very small bet (I don�t remember, maybe $5), on the game and he played. Canadian Dollars or US? Martin once told/wrote me that the bet gave him three tries and he managed it already in the first one. Regards, Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Handicap-29 Website is up now
mfgo1998 is crashing once in a while. I can fix the crash, but then it won’t be the same program. Otherwise you probably want to change the rules so that crashes don’t count. Please let me know if you want me to fix the crash. I wouldn't make any other changes, but the fix might make it a little stronger. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Friday, February 10, 2012 9:18 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Handicap-29 Website is up now Hi David, The 29 stone game wasn't really a demonstration game. My memory is that I had just finished igowin 9x9, and Martin played through all its levels up to professional, without losing a game... I don�t remember if I challenged him or if he said he could beat 19x19 on that handicap. So we made a very small bet (I don�t remember, maybe $5), on the game and he played. Canadian Dollars or US? Martin once told/wrote me that the bet gave him three tries and he managed it already in the first one. Regards, Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
please do not set handicaps just by passing, start with at least 6 handicap, then pass. If you set handicap to 6 or more the game will be free. Otherwise kgs treats it as a rated game, without handicap and it will mess up the raking system. mfgo1998 expects a 9 stone handicap then many passes to get to 29 stones. Otherwise its own internal handicap concept will be incorrect. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Hiroshi Yamashita Sent: Sunday, January 29, 2012 2:01 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Thanks! Aya tried 6 games. Aya uses 2 cores, 10sec/move. I think 14 handicaps is best for Aya. handicaps 5 win 11 win 9 win 16 loss 14 loss 14 win http://www.gokgs.com/gameArchives.jsp?user=ayamc Some small problem. I don't know how to select B, and set 9 handicap on KgsGtp. So I just passed 14 times in 14 hadicaps game, but KGS set komi to 6.5, and does not minus handicaps stones in final result. My config file is like this. --- engine=c:/go/kgs/AyaMC/aya.exe firstargument=--mode gtp verbose=f server.host=goserver.gokgs.com server.port=2379 name=AyaMC password=** room=Computer Go #mode=custom mode=wait talk=Good evening. I am a bot and do not talk. gameNotes=monte-carlo based Aya opponent=mfgo1998 rules=chinese rules.boardSize=19 rules.time=10:00+5x0:30 reconnect=t --- Hiroshi Yamashita - Original Message - From: David Fotland fotl...@smart-games.com To: computer-go@dvandva.org Sent: Sunday, January 29, 2012 10:33 AM Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones mfgo1998 is up on kgs if anyone wants to try it. To get a big handicap, set up a 9 stone game and pass until you get up to 29 stones in the same pattern as Martin's game. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
I asked for it to be rated. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Sunday, January 29, 2012 9:24 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Or he requested it but a KGS admin has not set the value yet. On Sun, Jan 29, 2012 at 12:09 PM, ds d...@physik.de wrote: Should be no trouble with the rating system, it seems that you did not register mfgo1998 as rated bot. So all games should be free. Am Sonntag, den 29.01.2012, 08:59 -0800 schrieb David Fotland: please do not set handicaps just by passing, start with at least 6 handicap, then pass. If you set handicap to 6 or more the game will be free. Otherwise kgs treats it as a rated game, without handicap and it will mess up the raking system. mfgo1998 expects a 9 stone handicap then many passes to get to 29 stones. Otherwise its own internal handicap concept will be incorrect. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Hiroshi Yamashita Sent: Sunday, January 29, 2012 2:01 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Thanks! Aya tried 6 games. Aya uses 2 cores, 10sec/move. I think 14 handicaps is best for Aya. handicaps 5 win 11 win 9 win 16 loss 14 loss 14 win http://www.gokgs.com/gameArchives.jsp?user=ayamc Some small problem. I don't know how to select B, and set 9 handicap on KgsGtp. So I just passed 14 times in 14 hadicaps game, but KGS set komi to 6.5, and does not minus handicaps stones in final result. My config file is like this. --- engine=c:/go/kgs/AyaMC/aya.exe firstargument=--mode gtp verbose=f server.host=goserver.gokgs.com server.port=2379 name=AyaMC password=** room=Computer Go #mode=custom mode=wait talk=Good evening. I am a bot and do not talk. gameNotes=monte-carlo based Aya opponent=mfgo1998 rules=chinese rules.boardSize=19 rules.time=10:00+5x0:30 reconnect=t --- Hiroshi Yamashita - Original Message - From: David Fotland fotl...@smart-games.com To: computer-go@dvandva.org Sent: Sunday, January 29, 2012 10:33 AM Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones mfgo1998 is up on kgs if anyone wants to try it. To get a big handicap, set up a 9 stone game and pass until you get up to 29 stones in the same pattern as Martin's game. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
Here is the 1998 Many Faces of Go against the released version 12, 40 minute game, two cores, 29 stones. White did not make a living group. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go mfgo98-mfgo.sgf Description: Binary data ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
I tried 29 stones once and it crushed me J David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Saturday, January 28, 2012 5:43 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Nice. This should be interesting... On Sat, Jan 28, 2012 at 8:33 PM, David Fotland fotl...@smart-games.com wrote: mfgo1998 is up on kgs if anyone wants to try it. To get a big handicap, set up a 9 stone game and pass until you get up to 29 stones in the same pattern as Martin's game. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
ok, its set up to only play ajahuang. give it a try. From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Huang Sent: Saturday, January 28, 2012 5:47 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones You should let me try. :) Aja From: David Fotland mailto:fotl...@smart-games.com Sent: Saturday, January 28, 2012 6:45 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones I tried 29 stones once and it crushed me J David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Saturday, January 28, 2012 5:43 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Nice. This should be interesting... On Sat, Jan 28, 2012 at 8:33 PM, David Fotland fotl...@smart-games.com wrote: mfgo1998 is up on kgs if anyone wants to try it. To get a big handicap, set up a 9 stone game and pass until you get up to 29 stones in the same pattern as Martin's game. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go _ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
ok, mfgo98 is back open to everyone. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Huang Sent: Saturday, January 28, 2012 8:03 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Yes, thanks. Actually in the last try, I lost only 30 points and was having a chance as well. ManyFaces is not unbeatable with 29 stones. I also tried 2 games against Minirock2 with 6 stones, 1 win 1 loss. The first game was a test. Minirock2 is obviously a MCTS program and it handles ladders very well. In the second, I killed a big group. Minirock2 also has the popular semeai problem. I think its playing strength is maybe around 1d. My guess would be Minirock2 is the Japanese program blast. Aja From: Stefan Kaitschick mailto:stefan.kaitsch...@hamburg.de Sent: Saturday, January 28, 2012 8:41 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Awesome attempt! 285 at R8 and you would have crushed it in the first try. Stefan On Sun, Jan 29, 2012 at 2:47 AM, Aja Huang ajahu...@gmail.com wrote: You should let me try. :) Aja From: David Fotland mailto:fotl...@smart-games.com Sent: Saturday, January 28, 2012 6:45 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones I tried 29 stones once and it crushed me J David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Saturday, January 28, 2012 5:43 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Nice. This should be interesting... On Sat, Jan 28, 2012 at 8:33 PM, David Fotland fotl...@smart-games.com wrote: mfgo1998 is up on kgs if anyone wants to try it. To get a big handicap, set up a 9 stone game and pass until you get up to 29 stones in the same pattern as Martin's game. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of David Fotland Sent: Saturday, January 28, 2012 11:34 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones The game with Martin was played on 8/5/1998 at the US Go Congress. My closest source is dated 7/3/1998. I have this built into a gtp front end now, and it's playing different moves from Martin's game, so the engine is not identical. Do you still want to continue with this bet? David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 1:22 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Which old Manyface, I have a DOS version from 1990 :) In 1997 ManyFaces 10.0 came to the market (and is sol still nowadays). David Fotland wrote that in the game against Martin Mueller he likely used a source from July 1998. He will try to reanimate that. ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go _ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go _ ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Pachi on cygwin
maybe you need a fflush(stdout)? David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Saturday, January 28, 2012 9:11 PM To: computer-go@dvandva.org Subject: [Computer-go] Pachi on cygwin I just build the latest Pachi on the patest Cygwin. I got some warnings but no errors. It does not crash when I run it. It does show it's thinking process when I give it a genmove command. But when it is done thinking, it does not print the result and prompt for the next command as expected. Instead it just hangs. Any ideas? $ ./pachi.exe Random seed: 1327809195 play b e5 IN: play b e5 got move 1,5,5 Fresh board with random seed 1327809195 Warning: Cannot promote move node! Several play commands in row? Move: 1 Komi: 0.0 Handicap: 0 Captures B: 0 W: 0 A B C D E F G H JA B C D E F G H J +---++---+ 9 | . . . . . . . . . | 9 | : : : : : : : : : | 8 | . . . . . . . . . | 8 | : : : : : : : : : | 7 | . . . . . . . . . | 7 | : : : : : : : : : | 6 | . . . . . . . . . | 6 | : : : : : : : : : | 5 | . . . . X). . . . | 5 | : : : : : : : : : | 4 | . . . . . . . . . | 4 | : : : : : : : : : | 3 | . . . . . . . . . | 3 | : : : : : : : : : | 2 | . . . . . . . . . | 2 | : : : : : : : : : | 1 | . . . . . . . . . | 1 | : : : : : : : : : | +---++---+ = genmove w IN: genmove w Fresh board with random seed 1327809195 [1] best 0.395954 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.396) F6(0.397) F7(0.423) F8(0.410) [2] best 0.405934 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.406) F6(0.388) G7(0.394) F7(0.423) [3] best 0.412510 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.413) F6(0.385) G7(0.387) G8(0.393) [4] best 0.417998 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.418) F6(0.385) G7(0.387) G8(0.392) [5] best 0.420043 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.420) F7(0.431) F6(0.385) G7(0.386) [6] best 0.420043 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.420) F7(0.428) F6(0.385) G7(0.386) [7] best 0.420043 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.420) F7(0.425) F6(0.385) G7(0.386) [8] best 0.420043 komi 0.0 | seq E6 F6 F7 D6 | can E6(0.420) F7(0.430) F6(0.385) G7(0.386) ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
I dont want to compete in this contest. Even if I were interested, the conflict of interest is too clear. The archived source I found is the version I took to the US Go Congress in 1998. It was stable enough that I released it in IgoWin (9x9 only). I remember that Martin played Igowin then and went through all of its levels without losing a game. Later that year a slightly modified version won the World Computer Go championship, so its strength was state of the art for the time. The algorithm was a highly pruned one ply full board search (typically looking at 10 to 30 moves), followed by an alpha-beta quiescence search. The evaluation function estimated the score and returned the move that led to the best score. The top level pruning was done with a pattern database and many rules related to group attack and group safety. The quiescence search tried to stabilize all groups to get an accurate score estimate. The evaluation function did local tactical search for all strings with 3 or fewer liberties, and some four liberty strings. It evaluated eyes of all shapes and sizes, with some local reading, and determined group strength using features such as eyes, ability to run, nearby weak enemy groups, etc. There was so much local search in the evaluation function that it only evaluated about 5 to 10 positions a second on the hardware of the time. I should be able to get it on KGS this weekend so it can get a rating. I expect it will be around KGS 12 kyu. regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Friday, January 27, 2012 8:46 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Petr Baudis pa...@ucw.cz: Ingo Althofer wrote: David Fotland is willing to provide the old bot - also for sparring sessions on KGS. Will David also try to compete? :-) Serious point. I did not have that in mind. In principle I would allow him to participate - as long as he does not exploit the exact structure of the pseudo-random generator within MF. Knowing David for several years now as an honorable man, I would believe him when he gives his word on this. Of course, for experiments it would be neat to have a binary of this version, but maybe that would make the challenge too easy (or prone to overtuning). As long as MF would play randomly enough, it would remain a difficult task. Anyway, sounds wonderful, and Martin Mueller's play in that game is quite something. I'm looking forward to giving this challenge to Pachi. I think, sophisticated dynamic komi will not be enough to achieve the level. You will need some sort of opponent modelling. But future will show. And looking over the fence: When a win at handicap 29 is achieved how much further could strong bots go? 33? or 37? or 49? Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
The random number generator is not very interesting for beating the program. All it does is add a small random bias to the evaluation function. The initial seed (from the time) means that every random game is a little different. I just used the system rand() function. There arent many move generation patterns that apply when the board starts so full, so its probably easier to exploit some bad local responses that are forced because only one pattern matches. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Friday, January 27, 2012 8:58 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones I hate to address it, but access to the binary would pretty much give away the structure of the prng. s. On Jan 27, 2012 8:46 AM, Ingo Althöfer 3-hirn-ver...@gmx.de wrote: Petr Baudis pa...@ucw.cz: Ingo Althofer wrote: David Fotland is willing to provide the old bot - also for sparring sessions on KGS. Will David also try to compete? :-) Serious point. I did not have that in mind. In principle I would allow him to participate - as long as he does not exploit the exact structure of the pseudo-random generator within MF. Knowing David for several years now as an honorable man, I would believe him when he gives his word on this. Of course, for experiments it would be neat to have a binary of this version, but maybe that would make the challenge too easy (or prone to overtuning). As long as MF would play randomly enough, it would remain a difficult task. Anyway, sounds wonderful, and Martin Mueller's play in that game is quite something. I'm looking forward to giving this challenge to Pachi. I think, sophisticated dynamic komi will not be enough to achieve the level. You will need some sort of opponent modelling. But future will show. And looking over the fence: When a win at handicap 29 is achieved how much further could strong bots go? 33? or 37? or 49? Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
The 18 kyu level of many faces should be weaker than the challenge bot, and it uses the same old pattern based algorithm. So you can practice already on kgd, or at home if you have a copy of many faces version 12. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of ds Sent: Friday, January 27, 2012 8:41 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones I unpacked my nice packed version from 1991 again, but it seems that it does not implement even the go modem protocol. At least it is not in the printed (!) user manual:) If somebody knows more, I would let my version play on KGS. This should not make it to easy to tune for the 1999 version, but to get an idea how to play a weak opponent. Detlef Am Freitag, den 27.01.2012, 16:50 +0100 schrieb Stefan Kaitschick: Of course, for experiments it would be neat to have a binary of this version, but maybe that would make the challenge too easy (or prone to overtuning). My guess is, that opponent modeling will have to be part of the package. For a generic weak opponent, maybe one sided simulation balancing would work. But it may very well be necessary to specifically model the opponents weaknesses. I hope that this version of MF is non-deterministic. :-) Stefan ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Beating old ManyFaces at 29 handicap stones
Yes. Randomization is an option in the new game menu. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Thursday, January 26, 2012 2:49 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Beating old ManyFaces at 29 handicap stones Maybe it's an obvious question but: Does the old ManyFaces have some randomization when picking moves? It is an important question. It has, at least my ManyFaces 10.0. Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Game 2 goes to Zen: 1-1
try the go teaching ladder: http://gtl.xmp.net/ hundreds of commented games at all levels, commentary by players a few stones stronger. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Friday, January 20, 2012 4:25 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Game 2 goes to Zen: 1-1 Oh, I followed the breadcrumbs back to this listing: http://gogameguru.com/get-better-at-go/commented-go-games/ On Fri, Jan 20, 2012 at 7:23 PM, Michael Williams michaelwilliam...@gmail.com wrote: Thanks for the link. I think I learned a few things about how to play Go. Does anyone know of a good collection well-commented 19x19 games? On Fri, Jan 20, 2012 at 12:56 PM, Stefan Kaitschick stefan.kaitsch...@hamburg.de wrote: This game is very nicely commented by David Ormerod at gogameguru: http://gogameguru.com/man-machine-match-final-results-game-commentary/ He did an amazing job of explaining the complicated tactical encounters. It's a lot better than what I could have come up with. Stefan ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Game 2 goes to Zen: 1-1
Ingo is correct. The territory chart's fluctuations are because it uses the old territory evaluation without any lookahead. So for example if a move threatens a big group, and the reply saves it, there will be fluctuation, since the group's territory went from solid to unsettled, back to solid. If I turned on the quiescence search it would be much smoother, but also much slower to draw. If you have the program you can click on a move (which goes to that position), and ask for a score display, and see what is going on. The early fluctuations in the win rate chart show that MCTS is not very good at fuseki. -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Sunday, January 15, 2012 6:33 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Game 2 goes to Zen: 1-1 Hi Darren, I think, I an also answer your questions on the charts. Winning percents for White, as seen by Many Faces of Go http://www.althofer.de/zen-tromp-02-percents.jpg Territory estimates for White, as seen by ManyFaces http://www.althofer.de/zen-tromp-02-territory.jpg ... David, what is the territory chart saying? Is it based on playouts (i.e. based on terminal positions), or is it counting the current board position (i.e. as if both players were to pass)? No playouts at all involved. Territory is just counting the current board position in the traditional pre Monte-Carlo way. It is one of the nice old MF-10 features within ManyFaces, and thanks to David for leaving them in the code. It is quite typical that the fluctuations in Territory are much larger than in the relevant part of the Monte-Carlo percent chart. (The territory charts says 15pts to black, with 52-55% to white at around move 45; then says 40pts to white, with about 52% to white at move 80.) The two charts are generated completely independently of each other. (This independence gives additional information especially to experienced users.) In the Monte-Carlo percent diagram each data point stems from 1,000 random games starting at the corresponding position. For reasons of computational speed only every third position is evaluated: after moves 3, 6, 9, 12, ... The positions after moves of Black are related to black dots, the others after moves of White. It is a bit strange that ManyFaces generates rather large fluctuations for the very first data points. Perhaps it would make sense to start the data points only at move 20 or so (David, do you hear me?). For me it would be also nice to have an option computing the % not only for every third but for every position in the game. Regards, Ingo. -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] John is Tromp in Game 1
Around move 121, it looks like Zen has the common MCTS problem that it plays the ko threats before starting the ko. Ko threats are essential for wining, so RAVE gives them very high biases. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Saturday, January 14, 2012 12:16 AM To: computer-go@dvandva.org Subject: [Computer-go] John is Tromp in Game 1 Hello, John Tromp got the white stones and his been winning game 1 of the slow-time match aagainst Zen on KGS. The sgf, including chat, you can download here: www.althofer.de/tromp-zen-01.sgf I let ManyFaces run across the game and got the following two figures. Monte-Carlo percents: http://www.althofer.de/tromp-zen-01-percents.jpg Expected Territory: http://www.althofer.de/tromp-zen-01-territory.jpg Around move 118-120 experts were still undecided about the possible outcome. According to ManyFaces, the game may have turned to John's favour around move 130. At move 146 Hideki said that Zen had still 37 % self-confidence. Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zur�ck-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] John is Tromp in Game 1
I had the same problem with Many Faces until recently, and I think Aja fixed it in Erica also. The symptom is pretty clear: an important ko appears and the program starts playing ko threats rather than starting the ko. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Saturday, January 14, 2012 1:27 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] John is Tromp in Game 1 i'd be surprised if the ko threats had more utility for mcts than as potential chances for winning due to mistake. i.e. if you have 50 ko threats, they all look like (especially after digging into the tree) ways to win due to mistake. s. On Sat, Jan 14, 2012 at 1:24 PM, David Fotland fotl...@smart-games.com wrote: Around move 121, it looks like Zen has the common MCTS problem that it plays the ko threats before starting the ko. �Ko threats are essential for wining, so RAVE gives them very high biases. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Alth�fer Sent: Saturday, January 14, 2012 12:16 AM To: computer-go@dvandva.org Subject: [Computer-go] John is Tromp in Game 1 Hello, John Tromp got the white stones and his been winning game 1 of the slow-time match aagainst Zen on KGS. The sgf, including chat, you can download here: www.althofer.de/tromp-zen-01.sgf I let ManyFaces run across the game and got the following two figures. Monte-Carlo percents: http://www.althofer.de/tromp-zen-01-percents.jpg Expected Territory: http://www.althofer.de/tromp-zen-01-territory.jpg Around move 118-120 experts were still undecided about the possible outcome. According to ManyFaces, the game may have turned to John's favour around move 130. At move 146 Hideki said that Zen had still 37 % self- confidence. Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zur?ck-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Lines of code
For Many Faces, all engine code is c, and line counts are from wc (includes blank lines and comments for .c and .h files) today: The uct/playout code is 10K lines. The old go engine is 55K lines. Some of the is code is not used by the strongest level, but the weaker levels still use the old alpha-beta searcher. There are about 2500 playout pattern gamma constants, about 62K joseki patterns, and about 2K old program patterns (with 51K move-value pairs). version 10, in 1997, had 42K lines of code in 2001 the old engine had 53K lines of code version 11, in 2002, had 52K lines of code The original version 12, 10/2008, had 5.2K lines in the uct/playout code. This was the first release that used MCTS. I don't have any on-line source older than version 10, and the older backups are on floppies, so I can't read them any more J I don't think I even have backups any more for any code before 1990. Everything from version 10 forward is in Git, so I could in theory see how many lines are unchanged since 1997. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Sunday, January 08, 2012 12:14 PM To: computer-go@dvandva.org Subject: [Computer-go] Lines of code Has there been any studies into the number of lines of code in the top chess/go programs over time? Another measure would be bytes of executable or bytes of executable+data. Obviously the latter grows in chess with endgame databases, so maybe that's less interesting. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] replacing dynamic komi with a scoring function
Yes. That's why MCTS prefers the center. Sometimes the playouts let an invasion work, so corner territory is not evaluated as being as secure as it should be. It seems easier for MCTS to kill groups in the center in the playouts. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Sunday, January 08, 2012 8:03 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] replacing dynamic komi with a scoring function underestimation? s. On Sun, Jan 8, 2012 at 7:30 PM, Yamato yamato...@yahoo.co.jp wrote: (2012/01/09 9:56), Jouni Valkonen wrote: There is indeed a problem with Dynamic komi with Zen. Zen often loses the handicap games if black tries to minimize the move count. Often if it is possible to bring game to small yose in around move 180 or so and if not too much behind, then Zen most likely will lose. I have played few games where Zen noted only when filling last dames that it is losing the game by few or half points and then resign. Although, one game was that I lost by ½ points, because I accidentally defended unnecessarily instead of taking the last dame. One game was that i was about ten points behind around move 180, but then Zen played a slack small yose, and lost by 2½ points. Also good and very easy strategy against zen in handicap games is to take all the sides and give center territory to the Zen. Zen almost always will take the center territory as too small and gives sides as too big. I think people often confuse the evaluation problem and the dynamic komi problem. A more urgent problem is the underestimation of the edge and corner territory. -- Yamato ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] replacing dynamic komi with a scoring function
middle game playouts are far out of the joseki library, and in any case the joseki library moves are not used in the playouts, only the uct tree. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Sunday, January 08, 2012 8:23 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] replacing dynamic komi with a scoring function even with the joseki libraries, MCTS prefers center? that seems to favor the idea of center actually being good. s. On Sun, Jan 8, 2012 at 8:13 PM, David Fotland fotl...@smart-games.com wrote: Yes. That's why MCTS prefers the center. Sometimes the playouts let an invasion work, so corner territory is not evaluated as being as secure as it should be. It seems easier for MCTS to kill groups in the center in the playouts. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Sunday, January 08, 2012 8:03 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] replacing dynamic komi with a scoring function underestimation? s. On Sun, Jan 8, 2012 at 7:30 PM, Yamato yamato...@yahoo.co.jp wrote: (2012/01/09 9:56), Jouni Valkonen wrote: There is indeed a problem with Dynamic komi with Zen. Zen often loses the handicap games if black tries to minimize the move count. Often if it is possible to bring game to small yose in around move 180 or so and if not too much behind, then Zen most likely will lose. I have played few games where Zen noted only when filling last dames that it is losing the game by few or half points and then resign. Although, one game was that I lost by ½ points, because I accidentally defended unnecessarily instead of taking the last dame. One game was that i was about ten points behind around move 180, but then Zen played a slack small yose, and lost by 2½ points. Also good and very easy strategy against zen in handicap games is to take all the sides and give center territory to the Zen. Zen almost always will take the center territory as too small and gives sides as too big. I think people often confuse the evaluation problem and the dynamic komi problem. A more urgent problem is the underestimation of the edge and corner territory. -- Yamato ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] win rate bias and CLOP
Yes, I understood this. The particular case I was trying to optimize used two tunables, and it looked like there were two optimal combinations, but CLOP chose a point between them that was less optimal. I was checking the interaction between the beta constant and the MFGO bias. -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Rémi Coulom Sent: Tuesday, January 03, 2012 7:19 AM To: computer-go@dvandva.org Subject: [Computer-go] win rate bias and CLOP It is important to understand that CLOP claims very little in terms of win rate. That is to say the win rate estimates it reports are all biased. Win rate over all samples underestimates the real win rate. Win rate near the maximum (central, and weighted) tend to be over-estimated. CLOP finds the location in parameter space that has the highest win rate. It may be the highest because it is the best, but also because it is the most lucky. That's why it is necessarily biased toward optimistic values. If the win rate over all samples is an improvement, then you can be sure you have an improvement. Otherwise you cannot be sure unless you actually play a lot of games with the suggested parameters. Rémi On 3 janv. 2012, at 14:09, Ingo Althöfer wrote: Hi David, David Fotland on CLOP-optimization: I tried it, but got no benefit so far. It claimed to find better settings for most parameters, but when I used them the program wasnt any stronger. Interestant. Had it similar strength or did it even become weaker? How often did the move proposals by your older ManyFaces and the CLOP-MF differ? Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen
It was weaker, but I have to do more experiments. I didn’t compare move proposals. I just ran 2000-game tournaments vs gnugo before and after. Win rage went from 84.9% to 78.9%. I need to try it again with fewer variables and more games. I still think it is a great tool. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Tuesday, January 03, 2012 5:09 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Hi David, David Fotland on CLOP-optimization: I tried it, but got no benefit so far. It claimed to find better settings for most parameters, but when I used them the program wasn�t any stronger. Interestant. Had it similar strength or did it even become weaker? How often did the move proposals by your older ManyFaces and the CLOP-MF differ? Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zur�ck-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen
Very impressive. Does anyone know the relative hardware? I think the 5 dan Zen was running on 26 cores. Remi, can you tell us how much of the benefit was due to CLOP? It's an excellent tool BTW. Thank you for sharing it. -David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Monday, January 02, 2012 11:04 AM To: computer-go@dvandva.org Subject: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Hello everybody, let us start the new year also in the mailing list. It seems that Zen is finding a strong contender for its position as leading bot on KGS. Starting on December 29, Crazy Stone (by Remi Coulom) has started a long playing session in KGS computer room, and has now established a stable 5-dan rating. See the rating diagram for CrazyStone at http://www.gokgs.com/graphPage.jsp?user=crazystone and the list of games at http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2011month=12 and http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2012month=1 Ingo (wishes to see bots with 6-dan ranks on KGS before the European Go Congress in July 2012) -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen
A 24 core SMP PC should be a little more efficient/stronger than the 26 core cluster that Zen uses. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of terry mcintyre Sent: Monday, January 02, 2012 11:43 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen This is the info on KGS user crazystone: running on a 24-core PC. I've watched a few games. It impresses me, and more to the point, impresses the high-dan players. Terry McIntyre terrymcint...@yahoo.com Unix/Linux Systems Administration Taking time to do it right saves having to do it twice. _ From: David Fotland fotl...@smart-games.com To: computer-go@dvandva.org Sent: Monday, January 2, 2012 2:12 PM Subject: Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Very impressive. Does anyone know the relative hardware? I think the 5 dan Zen was running on 26 cores. Remi, can you tell us how much of the benefit was due to CLOP? It's an excellent tool BTW. Thank you for sharing it. -David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Monday, January 02, 2012 11:04 AM To: computer-go@dvandva.org Subject: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Hello everybody, let us start the new year also in the mailing list. It seems that Zen is finding a strong contender for its position as leading bot on KGS. Starting on December 29, Crazy Stone (by Remi Coulom) has started a long playing session in KGS computer room, and has now established a stable 5-dan rating. See the rating diagram for CrazyStone at http://www.gokgs.com/graphPage.jsp?user=crazystone and the list of games at http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2011month=12 and http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2012month=1 Ingo (wishes to see bots with 6-dan ranks on KGS before the European Go Congress in July 2012) -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen
I tried it, but got no benefit so far. It claimed to find better settings for most parameters, but when I used them the program wasnt any stronger. From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Michael Williams Sent: Monday, January 02, 2012 11:22 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Did you use it on Many Faces? How much benefit did you see there? On Mon, Jan 2, 2012 at 2:12 PM, David Fotland fotl...@smart-games.com wrote: Very impressive. Does anyone know the relative hardware? I think the 5 dan Zen was running on 26 cores. Remi, can you tell us how much of the benefit was due to CLOP? It's an excellent tool BTW. Thank you for sharing it. -David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Ingo Althöfer Sent: Monday, January 02, 2012 11:04 AM To: computer-go@dvandva.org Subject: [Computer-go] CrazyStone in the 5-dan footsteps of Zen Hello everybody, let us start the new year also in the mailing list. It seems that Zen is finding a strong contender for its position as leading bot on KGS. Starting on December 29, Crazy Stone (by Remi Coulom) has started a long playing session in KGS computer room, and has now established a stable 5-dan rating. See the rating diagram for CrazyStone at http://www.gokgs.com/graphPage.jsp?user=crazystone and the list of games at http://www.gokgs.com/gameArchives.jsp?user=crazystone http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2011month=12 year=2011month=12 and http://www.gokgs.com/gameArchives.jsp?user=crazystone http://www.gokgs.com/gameArchives.jsp?user=crazystoneyear=2012month=1 year=2012month=1 Ingo (wishes to see bots with 6-dan ranks on KGS before the European Go Congress in July 2012) -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen
For the last year or so, all the bots have been using this time control (nine 15-second periods), so the bot ratings can be compared. All the periods are the same 15 seconds. So any time over 15 seconds uses one of the periods (over 30 seconds would use two periods, etc). 15 seconds is pretty reasonable for a quick game, and 9 periods allows a couple of long thinks. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Monday, January 02, 2012 2:28 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] CrazyStone in the 5-dan footsteps of Zen On Mon, Jan 2, 2012 at 4:02 PM, Ingo Althöfer 3-hirn-ver...@gmx.de wrote: Hello Don, What time control was used for these games or did it vary? In the last few days it was 15 seconds per move, with 9 byoyomi phases. I don't think that specifies the time control does it?So if a player exceeds 15 seconds he starts to use one of his byoyomi periods?How long for the byoyomi phases? I don't think 15 seconds per move average is bad for humans at all, but if you have to make each move in 15 seconds it's horrible and I can see that this is going to make the computer really look good. For making strength claims there should be some sort of standard. You could take almost any program and get 5 dan or more by setting the time to 1 second per move. Don Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Results of first day of the UEC Cup
Thanks. David Fotland 4863 Capistrano Ave San Jose CA 95129, USA I'll try to do better next year :) -David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Sent: Sunday, December 04, 2011 9:27 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Results of first day of the UEC Cup Dear David, Thank you for pointing out the error. We have fixed it. By the way, we like to send you the certificate of the fifth prize together with a supplementary good. Please let me know your mailing address. Best regards, Masakazu Muramatsu ; muram...@cs.uec.ac.jp 2011/12/4 David Fotland fotl...@smart-games.com: There is a small mistake I think. MFGO beat blast in the last round. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Sent: Saturday, December 03, 2011 6:55 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Results of first day of the UEC Cup The 1st day results are now on our web page; http://jsb.cs.uec.ac.jp/~igo/eng/result1.html SGF files will be uploaded within a week (or two). Masakazu Muramatsu ; muram...@cs.uec.ac.jp 2011/12/4 David Fotland fotl...@smart-games.com: Many Faces is using a single Core i7-2600 (4 cores, 8 threads, 3.4 GHz). David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Hiroshi Yamashita Sent: Saturday, December 03, 2011 6:57 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Results of first day of the UEC Cup Hi Remi, Your Japanese skill is perfect. Preliminary league has 24 paticipants include one Gnu Go guest. Top 16 programs plays tommorow's eliminate tounament. 1. Zen 7-0 2. Erica 6-1 3. Aya 5-2 4. Pachi 5-2 5. Many Faces 5-2 6. Nomitan 5-2 7. blast 4-3 8. Fuego 4-3 9. GNU Go 4-3 10. Katsunari 4-3 In the beginning of tournament, there were some network disconnections in remote paticipants. Pachi, MFG and Fuego lost one game for this. But this trouble seems to be cleared. I hope no network error tommorow. Aya uses 4 machines cluster. Each has 12 cores, so 48 cores all. I will use 10 machines 120 cores tommorow. Hiroshi Yamashita - Original Message - From: Rémi Coulom remi.cou...@free.fr To: computer-go@dvandva.org Sent: Saturday, December 03, 2011 8:20 PM Subject: [Computer-go] Results of first day of the UEC Cup Hi, A video with the results of the first day is online there: http://www.ustream.tv/channel/uec-go-2011 If I understood correctly, the top players may be: 1. Zen 2. Erica 3. Aya 4. Pachi 5. Many Faces 6. Nomitan But my japanese is not so good :-) Rémi ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Results of first day of the UEC Cup
There is a small mistake I think. MFGO beat blast in the last round. Regards, David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Sent: Saturday, December 03, 2011 6:55 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] Results of first day of the UEC Cup The 1st day results are now on our web page; http://jsb.cs.uec.ac.jp/~igo/eng/result1.html SGF files will be uploaded within a week (or two). Masakazu Muramatsu ; muram...@cs.uec.ac.jp 2011/12/4 David Fotland fotl...@smart-games.com: Many Faces is using a single Core i7-2600 (4 cores, 8 threads, 3.4 GHz). David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Hiroshi Yamashita Sent: Saturday, December 03, 2011 6:57 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] Results of first day of the UEC Cup Hi Remi, Your Japanese skill is perfect. Preliminary league has 24 paticipants include one Gnu Go guest. Top 16 programs plays tommorow's eliminate tounament. 1. Zen 7-0 2. Erica 6-1 3. Aya 5-2 4. Pachi 5-2 5. Many Faces 5-2 6. Nomitan 5-2 7. blast 4-3 8. Fuego 4-3 9. GNU Go 4-3 10. Katsunari 4-3 In the beginning of tournament, there were some network disconnections in remote paticipants. Pachi, MFG and Fuego lost one game for this. But this trouble seems to be cleared. I hope no network error tommorow. Aya uses 4 machines cluster. Each has 12 cores, so 48 cores all. I will use 10 machines 120 cores tommorow. Hiroshi Yamashita - Original Message - From: Rémi Coulom remi.cou...@free.fr To: computer-go@dvandva.org Sent: Saturday, December 03, 2011 8:20 PM Subject: [Computer-go] Results of first day of the UEC Cup Hi, A video with the results of the first day is online there: http://www.ustream.tv/channel/uec-go-2011 If I understood correctly, the top players may be: 1. Zen 2. Erica 3. Aya 4. Pachi 5. Many Faces 6. Nomitan But my japanese is not so good :-) Rémi ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Four questions, for competitors in KGS bot tournaments
Three hours seems reasonable. I agree that 2 hours is a little fast. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Nick Wedd Sent: Sunday, November 20, 2011 11:17 AM To: computer-go@dvandva.org Subject: [Computer-go] Four questions, for competitors in KGS bot tournaments __Question 1.__ The next KGS SLOW bot tournament will start next Sunday, November 27th, at 22:00 UTC. I have said that the time limits will be two hours each per game. I now think this is rather fast for a SLOW tournament, and am considering changing it to something slower (it will use 19x19 boards). The SLOW tournaments already run this year had time limits of 4 hours, 3 hours, and 8 hours. When I ran the one with 8 hours, the players generally took less than half their allocated time. So I think I may use 3 hours each (instead of the 2 planned). What do you think? Please send me your views soon (by email or to this list) so that I can get the tournament set up. __Question 2.__ Not really a question, this - I have realised that the regular December KGS bot tournament (Sunday December 4th, starting at 16:00 UTC, 9x9 boards) will clash with the UEC Cup (December 3rd and 4th, in Tokyo). This is unfortunate, but my personal schedule does not leave room to move it, all my other weekends this year are booked. It is also puzzling to me, as I have already moved it once in the hope of avoiding a clash. __Question 3.__ Was the a computer Go tournament at the GPW Cup http://sig-gi.c.u-tokyo.ac.jp/gpw/2011/index-e.html ? If so, can anyone tell me the results? __Question 4.__ Was the a computer Go tournament at the TAAI 2011 http://taai2011.cse.yzu.edu.tw/ ? If so, can anyone tell me the results? Nick -- Nick Wedd n...@maproom.co.uk ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] KGS highest rank Bot
The American Go Association has allowed bots in tournaments since at least the mid 80's. This is from their 1988 Tournament Rules. Later they gave people the option to declare at the start of the tournament if they refused to be paired with computers. In practice very few people refused. C. Computer entry. Computers may enter tournaments under certain conditions: 1. Only the inventor of the hardware/program or his/her designated agent may enter the computer (hereafter, either inventor or agent are called the operator.); 2. The computer must correctly handle any move legal for it or its opponent to make and must not make any illegal moves; 3. Both computer and operator must be AGA members; 4. The operator must play computer moves on a regular board and punch the clock for the computer; 5. The operator may enter or adjust playing parameters before a round begins, but not during a round; 6. The computer's clock must be left ticking if the operator must fix hardware or software problems. 7. The operator may offer to resign on the computer's behalf. -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of steve uurtamo Sent: Tuesday, August 09, 2011 6:21 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] KGS highest rank Bot kgs recently had a tournament where bots were allowed to play -- it was on nonstandard-sized boards, and zen did fantastically well, taking second place in the 21x21 tournament, in both american/european and asian/european divisions. there are also a stable of people throwing themselves at zen in the computer go room on kgs, solidifying its rank at 5d (as it slowly creeps toward 6d). (to be clear, this is the version playing at roughly (15s?/move), which in my experience is at-speed or slower than most non-tournament play happens in practice without a clock, so totally fair for humans to play at). so even if it can't play in human tournaments, everyone knows that it is at least as strong as the strongest 5d's on KGS. i think that it'd be great if bots could play in the 19x19 tournaments on kgs. that is a far cry from playing as an actual player over the board on a regular basis at regular tournaments. does anyone have an example of *any* game that existed before computers where computers have been accepted/allowed to play as a regular practice (instead of as a highly debated issue?). s. On Tue, Aug 9, 2011 at 6:06 AM, Jouni Valkonen jounivalko...@gmail.com wrote: Ingo wrote: The ranks you mention are from KGS. Is there something like a KGS World Championship, let it be with or without prize money? Winning such an online championship might be easier for a bot then winning over the board. Is it allowed for gobots to participate to online Kgs tournaments? It would very nice if they could. I think that there should be 2-4 places open for gobots, because computer go is such an important aspect of go. Chessbots could participate into some offline tournaments until they were too strong to play with humans. This is the best way to observe the development of gobots. -Jouni ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Master Thesis: Information Sharing in MCTS
Great thesis. Many Faces also uses rule-based playouts, so Pachi is not the only rule-based strong program. You mention that in the playouts you check ataris and extensions to avoid growing a losing ladder. Do you do a full ladder search, or just some local heuristics? Many Faces does not have any ladder search in the playouts. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Petr Baudis Sent: Wednesday, August 03, 2011 3:13 PM To: computer...@computer-go.org Subject: [Computer-go] Master Thesis: Information Sharing in MCTS Hi! If anyone is interested, you can read my master thesis at: http://pasky.or.cz/go/prace.pdf It could give a good introduction to current Monte Carlo techniques in Computer Go in general, and discusses some approaches for improvement (nothing too dramatic). It also gives a mid-level technical description of Pachi (with some important stuff left out, but we are preparing a paper). Kind regards, -- Petr Pasky Baudis UNIX is user friendly, it's just picky about who its friends are. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] testing improvements
Did each fuego play the same number of games vs gnugo, and did each play half its games on each color? -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Vlad Dumitrescu Sent: Thursday, August 04, 2011 9:57 AM To: computer-go@dvandva.org Subject: [Computer-go] testing improvements Hi all! I finally got an idea that is worth investigating. Luckily, it is something that can be tested by modifying existing programs and so I started to set up an environment to test it. In order to have a reference, I started this morning a small tournament with two identical versions of fuego (@1k) and gnugo (@level 10) and they played 863 rounds so far. The scores towards gnugo are almost identical, but the two fuegos score 449-415, which is 52% and the 95% confidence is ~3%, i.e. ~10 ELO. Now this is within limits, and it varies a bit, but it is always on the side of one of the instances, never less than 51.5%. Is this something normal? Sorry for the n00b question :-) best regards, Vlad ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] testing improvements
Remember that the confidence interval is two sided, so 3% means plus or minus 3%. So 52% win rate is within +- 3% of 50%. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Vlad Dumitrescu Sent: Thursday, August 04, 2011 1:14 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] testing improvements Hi, On Thu, Aug 4, 2011 at 19:29, David Fotland fotl...@smart-games.com wrote: Did each fuego play the same number of games vs gnugo, and did each play half its games on each color? Yes, I set up an all-play-all competition with gomill. On Thu, Aug 4, 2011 at 19:55, Erik van der Werf erikvanderw...@gmail.com wrote: On Thu, Aug 4, 2011 at 6:57 PM, Vlad Dumitrescu vladd...@gmail.com wrote: The scores towards gnugo are almost identical, but the two fuegos score 449-415, which is 52% and the 95% confidence is ~3%, i.e. ~10 ELO. That 3% is not a 95% confidence interval, more like 1 standard deviation... (so nothing with high confidence yet) I took the easy way out and used a formula mentioned by David Fotland on this list for a while ago There is a simple formula to estimate the confidence interval of a result. I use it to see if a new version is likely better than a reference version (but I use 95% confidence intervals, so over hundred of experiments it gives me the wrong answer too often). 1.96 * sqrt(wr * (1 - wr) / trials) Where wr is the win rate of one version vs the reference, and trials is the number of test games. On Thu, Aug 4, 2011 at 20:21, Kahn Jonas jonas.k...@math.u-psud.fr wrote: All the more since you're testing the same idea on two bots simultaneaously. So if you want to be wrong at most five percent of the time, and consider you are better as soon as one of the bots gets better, you have to make individual tests at the 2.5% level. At the moment I ran the bots without any modification, to see if everything works fine. So I think that the results between the identical bots should have been closer to 50% or at least to swing sometimes to the other side of 50%. Right now it's 625-566, which is 52,5% and 2.83% confidence according to the formula above. The results are fuego-1.1 v fuego-new (1199/2000 games) unknown results: 1 0.08% board size: 9 komi: 6.5 wins black whiteavg cpu fuego-1.1569 47.46% 386 64.33% 183 30.55% 2.69 fuego-new629 52.46% 415 69.28% 214 35.67% 2.67 801 66.81% 397 33.11% I realize that statistic results don't always match what one would expect, but this should be a straightforward case... Thanks a lot for all the answers! regards, /Vlad ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] testing improvements
I often see that one side gets lucky early and over a few hundred games the win rate moves back toward what I expected. Small numbers of games (like a few hundred) can be very misleading. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Vlad Dumitrescu Sent: Thursday, August 04, 2011 2:02 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] testing improvements On Thu, Aug 4, 2011 at 22:41, David Fotland fotl...@smart-games.com wrote: Remember that the confidence interval is two sided, so 3% means plus or minus 3%. So 52% win rate is within +- 3% of 50%. Yes, of course. What I reacted to was that under the whole test, one bot always had around 52% wins (well, after some 100 games, at least). I would have thought it would move around the real value. Thanks, Vlad -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Vlad Dumitrescu Sent: Thursday, August 04, 2011 1:14 PM To: computer-go@dvandva.org Subject: Re: [Computer-go] testing improvements Hi, On Thu, Aug 4, 2011 at 19:29, David Fotland fotl...@smart-games.com wrote: Did each fuego play the same number of games vs gnugo, and did each play half its games on each color? Yes, I set up an all-play-all competition with gomill. On Thu, Aug 4, 2011 at 19:55, Erik van der Werf erikvanderw...@gmail.com wrote: On Thu, Aug 4, 2011 at 6:57 PM, Vlad Dumitrescu vladd...@gmail.com wrote: The scores towards gnugo are almost identical, but the two fuegos score 449-415, which is 52% and the 95% confidence is ~3%, i.e. ~10 ELO. That 3% is not a 95% confidence interval, more like 1 standard deviation... (so nothing with high confidence yet) I took the easy way out and used a formula mentioned by David Fotland on this list for a while ago There is a simple formula to estimate the confidence interval of a result. I use it to see if a new version is likely better than a reference version (but I use 95% confidence intervals, so over hundred of experiments it gives me the wrong answer too often). 1.96 * sqrt(wr * (1 - wr) / trials) Where wr is the win rate of one version vs the reference, and trials is the number of test games. On Thu, Aug 4, 2011 at 20:21, Kahn Jonas jonas.k...@math.u-psud.fr wrote: All the more since you're testing the same idea on two bots simultaneaously. So if you want to be wrong at most five percent of the time, and consider you are better as soon as one of the bots gets better, you have to make individual tests at the 2.5% level. At the moment I ran the bots without any modification, to see if everything works fine. So I think that the results between the identical bots should have been closer to 50% or at least to swing sometimes to the other side of 50%. Right now it's 625-566, which is 52,5% and 2.83% confidence according to the formula above. The results are fuego-1.1 v fuego-new (1199/2000 games) unknown results: 1 0.08% board size: 9 komi: 6.5 wins black white avg cpu fuego-1.1 569 47.46% 386 64.33% 183 30.55% 2.69 fuego-new 629 52.46% 415 69.28% 214 35.67% 2.67 801 66.81% 397 33.11% I realize that statistic results don't always match what one would expect, but this should be a straightforward case... Thanks a lot for all the answers! regards, /Vlad ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] EGC computer-go tournament 9x9
I agree that 7.0 would be a better komi for the computer Olympiad, but it should be decided soon, since the komi affects the opening book, and an integer komi requires some changes in the playout algorithm to properly account for draws. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Jean-loup Gailly Sent: Monday, August 01, 2011 7:08 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] EGC computer-go tournament 9x9 At the end Zen was a convincing winner, with 9 wins out of ten games. Pachi on rank 2 with 5 points, The actual result is Zen 8 wins (won all its games, didn't play in two games because of parity), and Pachi still rank 2 but with 4 points only. In the last game against Pachi, Zen did not implement the KGS cleanup phase correctly, it passed while leaving dead stones on the board. So KGS indicates that Pachi has won. However the tournement rules http://www.egc2011.eu/index.php/en/computer-go/75-computer-go-reglement indicate In case of trouble in the KGS counting of the score, the human referee is allowed to change the result - the situation on the board as the priority. So it is fair to say that Zen has won the game. ManyFaces also failed to implement the cleanup phase correctly in another game, but this did not affect the result. Jean-loup 2011/8/1 Ingo Althöfer 3-hirn-ver...@gmx.de Today a 9x9 computer-go tournament was played within the European Go Congress: 5 bots participating, for each pairing two rounds (with different colours). At the end Zen was a convincing winner, with 9 wins out of ten games. Pachi on rank 2 with 5 points, ahead of ManyFaces, MyGoFriend, and MoGo. http://www.gokgs.com/tournEntrants.jsp?sort=s http://www.gokgs.com/tournEntrants.jsp?sort=sid=608 id=608 Komi was 7.5. 16 out of the 20 games ended with wins for White. From the four wins with Black three were achieved by Zen, the other one by Pachi. I think it is really necessary to play in the Tilburg Olympiad with a smaller komi value. Probably 7.0 would be a fair choice. Ingo. -- NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie! Jetzt informieren: http://www.gmx.net/de/go/freephone ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] Aja's PhD thesis
Congratulations on completing your Ph.D.! Well done. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Aja Sent: Wednesday, July 27, 2011 5:27 AM To: computer-go@dvandva.org Subject: [Computer-go] Aja's PhD thesis Dear all, If you are interested, my PhD thesis, entitled New Heuristics for Monte Carlo Tree Search Applied to the Game of Go, can be found in the following link. http://www.grappa.univ-lille3.fr/~coulom/Aja_PhD_Thesis.pdf Due to some personal reasons, I am sorry to announce that the sharing of Erica's binary is indefinitely postponed. Best regards, Aja ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] 19x19 opening books
Against strong players you can do better than that with a full board opening book, since there is not so much variety in human openings. I'd expect 6 to 10 moves typically, and sometimes as many as 25. But even this is not enough to really make a difference. A joseki book also helps, since so many joseki sequences have been worked out, and some mistakes can be quite costly. David From: computer-go-boun...@dvandva.org [mailto:computer-go-boun...@dvandva.org] On Behalf Of Don Dailey Sent: Monday, July 11, 2011 6:12 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] 19x19 opening books Opening books for 19x19 very limited usefulness because the board is so big. You are out of book after just a couple of moves. It could be used to place the first stone and have an instant response to the first stone the opponent places but it's not likely to take you farther than that. In 9x9 it could be critical as you could cover most reasonable plays to a modest depth.In 7x7 you could almost have a book that plays the toughest part of the game for you. Don On Mon, Jul 11, 2011 at 8:13 AM, Hendrik Baier hendrik.ba...@googlemail.com wrote: Good morning, I'm sorry I have to repeat my question: Can anyone point me to a paper on opening books in 19x19 Go? Or have you seen only cosmetic improvements by using them, but no strength improvements? It would be great if some of you program authors could give a one-line answer on that. best regards, Hendrik Baier ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
Re: [Computer-go] 19x19 opening books
On 19x19, Many Faces has books. A full board opening book made from strong player games is a hash of all positions (rotation/refection invariant). It keeps statistics of player strength and win rate, and is only used to bias the search, not to choose a move quickly. It also has a joseki book with all published lines in english, entered from books (about 60K positions). On 9x9 I have a full board book made from strong cgos games, but I don't ship it with the program. David -Original Message- From: computer-go-boun...@dvandva.org [mailto:computer-go- boun...@dvandva.org] On Behalf Of Aja Sent: Monday, July 11, 2011 5:35 AM To: computer-go@dvandva.org Subject: Re: [Computer-go] 19x19 opening books Hi Hendrik, In the Olympiad last year in Kanazawa, I noticed that Fuego was using a small 19x19 opening book. It successfully forced Fuego to play at the corner in the beginning. In Erica, I never tried using a book in 19x19 board. As far as I know, Many Faces is not using a book. pachi is apparently without a book either. Aja - Original Message - From: Hendrik Baier hendrik.ba...@googlemail.com To: computer-go@dvandva.org Sent: Monday, July 11, 2011 8:13 PM Subject: [Computer-go] 19x19 opening books Good morning, I'm sorry I have to repeat my question: Can anyone point me to a paper on opening books in 19x19 Go? Or have you seen only cosmetic improvements by using them, but no strength improvements? It would be great if some of you program authors could give a one-line answer on that. best regards, Hendrik Baier ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
