[computer-go] Re: Open source real time Go server

2010-01-21 Thread Jacques Basaldúa
Petr Baudis wrote: This seems like not very productive line of argumentation unless preceded with more exact definitions of strong. My only claim is that it is a hard problem. That is unobjectionable no matter how you define strong (obviously: random strong perfect) I can't understand why

[computer-go] Re: Open source real time Go server

2010-01-20 Thread Jacques Basaldúa
Hi 1. Just a small precision first: Ingo Althöfer wrote: Terry McIntyre wrote: ... My pet peeve is the KGS score estimator, which is often wildly wrong As explained by others a strong SE for ALL positions is equivalent to a strong program. This is only true if you replace the word

[computer-go] Fuego 04 native Windows

2009-12-18 Thread Jacques Basaldúa
Michael Williams It has been some time since you made this. It was last summer (2009). I needed a strong program to test an idea on it that worked well on a weaker program. I haven't shared this with the list yet, because I am now modifying it and cannot spend as much time as I would like.

[computer-go] Biasing nodes according to pattern gammas

2009-12-16 Thread Jacques Basaldúa
Petr Baudis wrote: I wonder now, do you use separate set of gammas for simulations and node biasing? Since I've found that the performance seems very bad if I don't include some time-expensive features, since the gammas are then very off; I will probably simply generate two gamma

[computer-go] Fuego parameter question

2009-12-09 Thread Jacques Basaldúa
To run Fuego in Windows, there the build I posted in: http://www.mail-archive.com/computer-go@computer-go.org/msg12208.html There is both source and binary. It is 40% faster than the cygwin version and more efficient in all senses, just because it is a native build. cygwin has serious

[computer-go] Kinds of Zobrist hashes

2009-12-09 Thread Jacques Basaldúa
1. The 3 hashes: As Eric Boesch and Álvaro said, you can get the same codes as in the Black[], White[], Empty[] scenario using only 2 BlackXorEmpty[] and WhiteXorEmpty[] 2. What does make sense is making a difference between a POSITIONAL hash and a SITUATIONAL hash. The positional hash is the

[computer-go] Kinds of Zobrist hashes

2009-12-09 Thread Jacques Basaldúa
Álvaro Begué wrote: The legal moves might be the same now, but they might be different after I perform a particular move. You are right. Looks like my solution is not perfect, but it works good enough. I have seen trouble, before I implemented it, because the transposition table identified

Re: [computer-go] still an outstanding command

2009-10-07 Thread Jacques Basaldúa
I think it is OK to lose the one game where it crashes, but it should be able to restart for new games at least. Martin You can restart everything Ok, but you have to kill the process java.exe and then run kgsgtp again. When the program crashes, the clock is still running, I think. If it

[computer-go] [Fwd: Announcement ICGA Events 2010]

2009-10-06 Thread Jacques Basaldúa
Is it possible to participate in these tournaments remotely? I think I would like to start participating in these, but I certainly have no budget (or time) to travel outside of Europe. :-( Petr Pasky Baudis For the ICGA Computer Olympiad, there should be no

[computer-go] Fuego 04 native Windows

2009-08-09 Thread Jacques Basaldúa
I have made a native Windows Fuego build compiled with MS Visual Studio 2008. Thanks to the Fuego team for making such a nice program free software !! I will use it to measure some tree metrics to tune my own program and for a validation experiment for an evaluation function I have developed.

Re: [computer-go] Rules for remote play at the Computer Olympiad

2009-02-02 Thread Jacques Basaldúa
About the thinking process log. Enabling debugging options can result in serious performance loss. In my system only the admin thread can do such things as tree dumps and that makes all other pawn threads idle. I don't think such preventive measures are justified. In case of doubt, it should

[computer-go] [Fwd: ICGA Events 2009 in Pamplona]

2009-01-09 Thread Jacques Basaldúa
Good news! I am very happy the ICGA has chosen Pamplona. Other destinations (e.g. Beijing) are more attractive, but way out of my reach. I can travel to Pamplona easily, but cannot find the details. The website at http://www.icga.org/ is not updated and the attached .eml file contains 2 .pdf

[computer-go] MoGo v.s. Kim rematch (Jason House's paper)

2008-09-24 Thread Jacques Basaldúa
The approach of this paper is to treat all win rate estimations as independent estimators with additive white Gaussian noise. Have you tried if that works? (As Łukasz Lew wrote experimental setup would be useful) I guess there may be a flaw in your idea, but I am not a specialist. I will try

[computer-go] Counting the final score with seki

2008-08-25 Thread Jacques Basaldúa
Hi all, When you detect self atari in the playouts (something I haven't tried yet, but from recent posts in this group I am convinced that it is an important issue) a new problem arises: How can you score the board _fast_ at the end? My previous version killed all groups in seki and scanned

Re: [computer-go] 10k UCT bots

2008-05-14 Thread Jacques Basaldúa
Don Dailey wrote: [EMAIL PROTECTED] wrote: For those currently coding this up, I think the most important thing about this playout algorithm is that it is *temporary*. You will almost certainly be?replacing it with something different and better just a little bit down the road. So you

Re: [computer-go] Paper for AAAI (David Silver) PDF problem

2008-04-08 Thread Jacques Basaldúa
Thanks, Hideki I have the same problem with version 7.0.7 of Adobe Reader but version 8.1.2 works fine. That was the problem. I used the automatic upgrade feature, but that only upgrades to the latest 7.xx version not to version 8.xx. Don Dailey wrote: ... this general distaste of feeling

Re: [computer-go] Paper for AAAI (David Silver) PDF problem

2008-04-07 Thread Jacques Basaldúa
Hi David http://www.cs.ualberta.ca/~silver/research/publications/files/MoGoNectar.pdf Your paper has a PDF problem concerning embedded font BXGQFO+CMR12. I have used different versions of Adobe Reader. I even updated one of the computers to the latest version and I still cannot read any

Re: [computer-go] State of the art of pattern matching

2008-04-02 Thread Jacques Basaldúa
Jonas Kahn wrote: I guess you have checked that with your rules for getting probability distributions out of gammas, the mean of the probability of your move 1 was that that you observed (about 40 %) ? If I understand your post, there may be a misunderstanding by my fault. Here gamma is not

Re: [computer-go] State of the art of pattern matching

2008-04-01 Thread Jacques Basaldúa
Lew wrote: On Sat, Mar 29, 2008 at 3:47 PM, Jacques Basaldúa [EMAIL PROTECTED] wrote: Mark Boon wrote: I suppose there's no reason why it has to be assembler, you could just as well generate C-code. I don't think so. But I don't write that much asm as it may appear. I see what

Re: [computer-go] State of the art of pattern matching

2008-03-30 Thread Jacques Basaldúa
Mark Boon wrote: There are 16 4-distance points, so if you spill ino that by one point you get 315 or a little over 14 million patterns. Multiplied by 3 for every don't-care point within less than 4 distance. Ouch. True, but the number of patterns is learned automatically. When I first learn

Re: [computer-go] State of the art of pattern matching

2008-03-29 Thread Jacques Basaldúa
Mark Boon wrote: I suppose there's no reason why it has to be assembler, you could just as well generate C-code. I don't think so. But I don't write that much asm as it may appear. I see what the compiler writes when I debug and write critical parts only. And automatically generated code.

Re: [computer-go] State of the art of pattern matching

2008-03-28 Thread Jacques Basaldúa
Mark Boon wrote: Sorry, without a bit more explanation, the assembler code is very hard to understand. What exactly does it do? The first source code was just an example to see what kind of code is generated. The second is useful, if you understand asm you should understand it. The board

Re: [computer-go] State of the art of pattern matching

2008-03-28 Thread Jacques Basaldúa
terry mcintyre wrote: It is possible to get some remarkably high correlation between the moves played by pros and a predictor - yet still not have a good program. Why? We have a random variable, the place at which a player plays, and some variables that we can compute. The distribution of

Re: [computer-go] State of the art of pattern matching (Oops)

2008-03-27 Thread Jacques Basaldúa
Santiago wrote: ... Oops, wrong name ... (Santiago is my official name, because I was born in a dictatorship that did not allow foreign names. After that, I was too lazy to change it. Jacques, like my French grandfather, is my real name.) Jacques.

Re: endgame (Was [computer-go] Re: Should 9x9 komi be 8.0 ?])

2008-03-06 Thread Jacques Basaldúa
Petr Baudis wrote: MoGo can indeed play out some rather spectacular ko fights; unfortunately, I couldn't find any quickly, so here is at least an example of a shorter one. I see you made the following comment in that game record, which seems relevant to recent discussions here. |

Re: [computer-go] Way MC plays

2008-03-01 Thread Jacques Basaldúa
Don Dailey wrote: I personally have serious doubts about knowledge extraction from human games, but I hope you have success.I think you can get more from computer games of strong players even though the level is weaker. Here is why I say that: 1. A strong computer still plays a lot of

Re: [computer-go] Way MC plays

2008-02-29 Thread Jacques Basaldúa
Jonas Kahn wrote: I mean, if Crazy Stone played against himself from the position where Katsunari was thought to be ahead, would he win with its original color, or with Katsunari's ? That is a very big question. I hope it would win with Katsunari's stones *if* Katsunari was really ahead. I

[computer-go] Re: 19x19 Study. Nakade is difficult

2008-01-31 Thread Jacques Basaldúa
I mentioned nakade in a list including not filling own eyes. Perhaps, not filling own eyes is a simpler example: | . . . . . . . | . # # . # . . | . O # . # . . | O O O . # # . | # # O O O # . | . # # . . . . --- (Unless I made a mistake: Black to play and a1 is the only move

Re: [computer-go] 19x19 Study - prior in bayeselo, and KGS study

2008-01-30 Thread Jacques Basaldúa
Dave Hillis wrote: I've noticed this in games on KGS; a lot of people lose games with generous time limits because they, rashly, try to keep up with my dumb but very fast bot and make blunders. What Don says about humans scaling applies to humans making an effort to use the time they have,

Re: [computer-go] New scalability study : show uncertainty ?

2008-01-23 Thread Jacques Basaldúa
I don't think only uniformly random playouts will scale to perfection because what we need for playouts is not just a simple average of final scores but a maximum (in negmax sense) score. It should be the perfect evaluation function. In other words, as MC simulation is a way to get an

Re: [computer-go] New scalability study : show uncertainty ?

2008-01-22 Thread Jacques Basaldúa
Hideki Kato wrote: It's rather odd. I'm checking the log file and then I will check the source code to see if I have some artificial limits in there. Why odd? It all depends on the bias or policy of simulations. If there is a flaw in the policy, the score will converses to the score

Re: [computer-go] 19x19 cgos ranking page

2008-01-20 Thread Jacques Basaldúa
Olivier Teytaud wrote: the 19x19 CGOS ranking page is back (but might be still unstable) and Leela seemingly performs quite well. The crosstables will come back soon also. The crosstables are back, but the sgf archives ar not. I get: Forbidden You don't have permission to access

Re: [computer-go] Suicide question

2008-01-18 Thread Jacques Basaldúa
Hi Michael Another cost is undo. Superko requires undo, unless you want store a hash value with each chain of stones. I am not sure exactly what undo costs, but lets say 5% to 10%. Well, every implementation is different. In its slowest mode, my board stores information about neighbor stones

Re: [computer-go] Suicide question

2008-01-17 Thread Jacques Basaldúa
Michael Wing wrote: In my program (which implements undo), the cost of for suicide detection is around 1%, which means it would lose 1.5 ELO points. In programs that somehow maintain lists of legal moves or even probability distribution functions over the legal moves, avoiding suicide is

[computer-go] Suicide question

2008-01-16 Thread Jacques Basaldúa
Gian-Carlo Pascutto wrote: Multi-stone suicide is allowed, single stone not. I hadn't even considered suicide.(It would be a major change for me, as neither my Gui nor my board system allow such moves.) The question is Why do you do it? a. Just in case you wanted the entire program to

Re: [computer-go] Is CGOS sending TIME_LEFT?

2008-01-09 Thread Jacques Basaldúa
Hi Don I noticed that your list is upper-case. This might be the problem. I don't remember if GTP is case sensitive or not, but I'm pretty sure cgos requires lower case in these commands. Thank you. That was the problem. It was uppercase. To make it case insensitive, I convert all input to

Re: [computer-go] How to get more participation in 19x19 CGOS?

2008-01-09 Thread Jacques Basaldúa
Don Dailey wrote: My suggest to Olivier is to compromise and set time control 20 minutes per side and give 1 second gift. That's good for me. Until now, I have been running tests and gnugo clones just to support the server, but I will start serious work soon. I have some time for computer

Re: [computer-go] cgos from windows

2008-01-09 Thread Jacques Basaldúa
I think you are right about the tclkit exe. I used the one that the instruction page linked, which does throw up a GUI if you run it with no args. I wonder if it even works. It appears to be a way to cripple the server. It does work. I have played over 600 games with it. It does not

Re: [computer-go] Is CGOS sending TIME_LEFT?

2008-01-08 Thread Jacques Basaldúa
Hi Don The client does not sent time_left unless time_settings is also implemented.So your engine must also implement time_settings which is needed to inform your program of the level it will be playing at. I do implement time_settings. The server only sends a list_commands at the

[computer-go] Is CGOS sending TIME_LEFT?

2008-01-07 Thread Jacques Basaldúa
Hi.. My gtp program does not receive any time_left commands. I have checked that the command is listed without mistakes in list_commands. I even tried CGOS 9x9 and I get the same sequence. Jacques. -- This is a 19x19 game start (after

Re: [computer-go] Please have your bot resign, for your own good

2008-01-04 Thread Jacques Basaldúa
The problem is avoiding that an inferior program wins a lost position on time extending the number of moves. If you could choose, what side would you prefer to be? As a human, there is no doubt. But as a program it is even better: the strongest program. Because computers can play faster than

RE: [computer-go] misconception about game(?)

2007-12-25 Thread Jacques Basaldúa
Forrest Curo wrote (quoting David Fotland): for example, go books make a big deal about where to extend along the side, or when to play in one corner or another, but the difference between these various moves is usually only a few points. The difference between similar-appearing various moves

Re: [computer-go] rotate board

2007-12-20 Thread Jacques Basaldúa
Don Dailey wrote: You can use Zobrist hashing for maintaining all 8 keys incrementally, but you probably need a fairly good reason to do so. Incrementally updating of 1 key is almost free, but 8 might be noticeable if you are doing it inside a tree search or play-outs. Yes. Don is

Re: [computer-go] rotate board

2007-12-19 Thread Jacques Basaldúa
Ben Lambrechts wrote: Now I got the following problem: how to rotate/mirror the board for a unique representation. Both are the same board, but has anyone made an algorithm that rotates the board or an area of the board in a unique way? I don't need the move order, just the snapshot of the

Re: [computer-go] A thought about Bot-server communications

2007-12-13 Thread Jacques Basaldúa
Nick Wedd wrote: In one of the British Championship Match games, a bit over ten years ago, Zhang Shutai made an illegal ko move against Matthew Macfadyen, and immediately conceded that he had lost the game. Is the game record available? I am interested because I have only found 2 situations

Re: [computer-go] Microsoft Research Lectures: Akihiro Kishimoto

2007-12-12 Thread Jacques Basaldúa
Thank you, Harri. It sounds promising. I will have a look at that. Jacques. I coded algorithm based for that representation, it really looks another awesome thing worth of investigating. Planning to use that for at least small board search investigations as it has quite much power. That

[computer-go] Microsoft Research Lectures: Akihiro Kishimoto

2007-12-11 Thread Jacques Basaldúa
David Stern wrote: Akihiro's talk has finally been put online at: http://content.digitalwell.washington.edu/msr/external_release_talks_12_05_2005/15004/lecture.htm Good lecture. Is there a link to a binary (or source code) somewhere ? I can't find any TsumeGo Explorer website. At

[computer-go] ELO Ratings of move pattern

2007-12-06 Thread Jacques Basaldúa
Lars wrote: Anyone of you have similar or other experiences with the algorithm? I use at runtime the same Bradley-Terry formulas Remí introduces in his paper. That is a huge advance compared to naif urgency scores because it gives a measure of how hard it was to win for a move candidate. But I

[computer-go] Redirecting i/o in the Windows version of Mogo (Was in Re: Update of MoGo binary ..)

2007-12-06 Thread Jacques Basaldúa
Edward de Grijs wrote: Can someone help me with this problem, for which I cannot find a solution: I am trying to run MoGo in an automatic way, using the cygwin toolkit. I guess you are trying to do this in Windows and using the Windows binary. If this is the case, you don't need any library.

[computer-go] Re: The global search myth

2007-12-06 Thread Jacques Basaldúa
Dave Dyer wrote: In cases where the good moves are the obvious ones, you've found them anyway. Ok. Here I agree. In other cases, you prune them away. You are not really pruning, just postponing. Of course you may overlook moves of genius, who doesn't? But if your probabilities are

Re: [computer-go] New engine? From a Chess programmer perspective.

2007-12-04 Thread Jacques Basaldúa
Christoph Birk wrote: I don't think 2200 ELO on the 9x9 CGOS is equivalent to 'high dan-level' play. Neither do I. In fact the whole kyu/dan rating system applies only to 19x19. 9x9 is not deep enough to have have so many ranks. On a 9x9 board an average amateur beats a pro with handicap 3.

Re: [computer-go] Drunken sailor on payday

2007-11-22 Thread Jacques Basaldúa
Raymond Wold wrote: Do you have anything to back this up? I was under the impression that most decent assembly programmers agreed that they can't compete with the best C compilers. Absolutely NOT. Only the typical, perhaps 99.9% of the programs, made of: transfers, conditionals, integer

[computer-go] CGOS 19X19 is down

2007-11-20 Thread Jacques Basaldúa
Hello I was waiting till someone restarts, but nobody seemed to notice. CGOS was hanging yesterday morning (European time) with 3 games 4849..4851 where no black engine placed any stone. If black restarted (one of the black bots was mine) it lost on time because the 30 minutes had been used.

Re: [computer-go] Language

2007-11-14 Thread Jacques Basaldúa
Don Dailey wrote: My Go program doesn't use any libraries except the standard C libraries. I agree at 99.9% !! The only exception in my case is SFMT. http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html SFMT is 100% clean C software, easy to compile, easy to use and free. A good

Re: [computer-go] Connecting a gtp engine to CGOS 19

2007-11-04 Thread Jacques Basaldúa
Solved! It is working now. The problem was the tcl script does not pass all remaining parameters to the called application (gnugo in the example). To solve this, the command line should be delimited with a quote char. tcl cgos3.tcl testingTCL password E:\\GO\\PROGRAMS\\GnuGo\\gnugo37 --mode

Re: [computer-go] 19x19 CGOS

2007-10-29 Thread Jacques Basaldúa
Don Dailey wrote: Of course that's better, but I'm talking about a quick and dirty solution. I may never implement handicap games since it's tricky with ELO ratings. This can also be done by the programmers. E.g. If CrazyStone is too strong, Rèmi can introduce a CrazyStoneH3 which passes

RE: [computer-go] XML alternatives to SGF

2007-10-27 Thread Jacques Basaldúa
Bob Myers wrote: Many of those complaining about XML don't seem to really know too much about it. That is exactly my point. I don't know and I don't want to know! SGF is fine. It has been stable for years because there is no problem at all. Should we find a problem, there is a

Re: [computer-go] XML alternatives to SGF

2007-10-24 Thread Jacques Basaldúa
Heikki Levanto wrote: Even if a new proposed standard would have many benefits, obvious to everyone (which I have not yet seen), I would stuill urge people to consider those benefits carefully, and to weigh them against the problems that arise from having two incompatible standards. I

Re: [computer-go] Update of MoGo binary release, and windows version available!

2007-09-20 Thread Jacques Basaldúa
Thank you Sylvain! It works fine for me. A very good sparring strong and so different in style! You mean that with the .dll in WINDOWS\system32 that goes further? I thought that the .dll in the same directory as the .exe was enough. Copy/pasting from Microsoft's API reference: The function

Re: [computer-go] Re: Most common 3x3 patterns

2007-09-20 Thread Jacques Basaldúa
Cenny Wenner wrote: Care to elaborate on what you mean by scores here and how they are similar to the 9x9 equivalence? I guess Dave is using Bradley Terry scores. This idea was introduced by Rémi Coulom in Computing Elo Ratings of Move Patterns in the Game of Go The paper is available on the

Re: [computer-go] playing strength of programmers

2007-09-12 Thread Jacques Basaldúa
I agree with all those who say it is important, but there is some precision to be made: As a player you are as strong as your weakest link because you are punished for your mistakes. As a programmer you are a strong as your strongest link. You know that mistakes are just mistakes, as long

Re: [computer-go] Why Poker-GMs don't win at poker.

2007-07-27 Thread Jacques Basaldúa
Poker can be analyzed well by (even naif) Monte Carlo methods. How? Simulation! Whatever evaluation you need and don't know how to compute because it is too complex for easy formulas can be simulated. This applies to the probability of winning, but also on the betting decisions (call,

[computer-go] Go in hardware

2007-07-27 Thread Jacques Basaldúa
Hi Chrilly: You have mentioned go in hardware twice recently and I have, knowing that you have experience in hardware development, some questions: 1. What should be implemented? In your Hydra cluster I have read you implemented mobility, and somewhere you proposed something like influence. Can

[computer-go] Why Poker-GMs don't win at poker.

2007-07-26 Thread Jacques Basaldúa
Chrilly Donninger wrote: I think poker is more difficult than Go and of course chess. Poker can be analyzed well by (even naif) Monte Carlo methods. Of course, the fact that you get a better estimate than any human could dream of, won't necessarily make you win. This common misconception can

Re: [computer-go] Intelligence

2007-07-23 Thread Jacques Basaldúa
Hideki Kato wrote: Creativity here is, to generate a new method by combining methods the system already has, in order to solve a problem. That is creativity for the job of designing algorithms. Playing go, creativity is finding moves _that work_ that nobody would have thought of. I think

Re: [computer-go] Slides for Villach-EC Lecture

2007-07-22 Thread Jacques Basaldúa
The word in AI I don't feel conformable with is Artificial, not Intelligence. I use _Abstract_ Intelligence (also AI) as a replacement. Have you ever heard of artificial aerodynamics (applicable to planes but not to birds) or artificial thermodynamics (the same). I understand that AI is the

Re: [computer-go] Why are different rule sets?

2007-07-13 Thread Jacques Basaldúa
Jason House wrote: I run some really dumb bots online that play perfectly fine blitz games (10s/move) with Chinese rules and it still drives humans insane because the computer doesn't stop playing. People resign won games in endgame because they can't take it. There is some value in reducing

Re: [computer-go] UCT caveat (was in Explanation to MoGo paper wanted)

2007-07-12 Thread Jacques Basaldúa
Brian Slesinsky wrote: And this would mean that a position where black is in trouble would look stronger than in a random playout (due to black playing well only for this kind of situation) which would make it harder to tell which positions are actually good. Or in general, an improvement

Re: [computer-go] Explanation to MoGo paper wanted.

2007-07-11 Thread Jacques Basaldúa
Hi, Don I can find arguments to disagree. I think what makes humans homo sapiens is reasoning, not the ability to compute numerical simulations. As a human (I am one) I feel disappointed when the explanation I get for a best move is after x millions of simulated matches it proved to be the

Re: [computer-go] UCT caveat (was in Explanation to MoGo paper wanted)

2007-07-11 Thread Jacques Basaldúa
Brian Slesinsky wrote : When you favor defense (or attack) you may think: This is unbiased since some times it favors black and other times it favors white But the fact is when black is in danger at the root of the tree, it is in danger in most of the tree, therefore the trick gets the

Re: [computer-go] 9x9 games wanted and the next big challenge

2007-07-09 Thread Jacques Basaldúa
Except for the relation between not finding 9x9 games which is *not* real go, you can find as many 19x19 games as you want, I agree with Chrilly. Let's accept it. We are amateurs, all except those who are paid by some University to research on go. And even some of them are, because a serious go

[computer-go] Re: computer-go Digest, Vol 36, Issue 6

2007-07-06 Thread Jacques Basaldúa
terry mcintyre wrote: Lately, I've been studying joseki, and I find that it's hard to really know a joseki until you know why non-joseki moves are bad - and why moves which are locally joseki may be bad in relation to other stones on the board. No doubt. That is the most complicated part. I

[computer-go] Re: computer-go Digest, Vol 36, Issue 5

2007-07-05 Thread Jacques Basaldúa
George Dahl wrote: Is it still cheating if the program learns and discover's the patterns itself? Then isn't it just precomputing a few things? I agree. Many chess and go grand masters, e.g. Capablanca are supposed to have learned the game as children by watching others play. I do intensive

Re: [computer-go] Explanation to MoGo paper wanted.

2007-07-05 Thread Jacques Basaldúa
Hi, Magnus Magnus Persson wrote: Weak tactics is a problem of the playouts in my opinion. UCT as a general search method has thus little to do with ladders and other game specific details. If there are no tactical mistakes in the playouts the problems disappear. Also tactics has a large

Re: [computer-go] Explanation to MoGo paper wanted.

2007-07-05 Thread Jacques Basaldúa
Don Dailey wrote: I have posted before about the evils of trying to extract knowledge from human games. I don't think it is very effective compared to generating that knowledge from computer games for several reasons. I would agree if we could have quality games played by computers. In

[computer-go] How can one possibly design an optimal playout agent

2007-07-05 Thread Jacques Basaldúa
Peter Drake wrote: 1) If the computation necessary to find better moves is too expensive, performing many dumb playouts may be a better investment. 2) If the playouts are too deterministic, and the moves are merely pretty good, the program may avoid an important move and thus misjudge

Re: [computer-go] correspondence or turn-based servers

2007-06-29 Thread Jacques Basaldúa
terry mcintyre wrote: Computer Go play in general is nowhere close to dan-level play, but a computer program can read out smaller tactical problems with a very high level of accuracy. That is true. Computers can analyze closed positions as Thomas Wolf's go tools at dan level, but .. .. If

Re: [computer-go] Congratulations to GNU and to MoGoBot19!

2007-06-20 Thread Jacques Basaldúa
steve uurtamo wrote: true, and a good point. time management other than attempting to equally divide remaining time among the expected number of remaining moves (which itself isn't so easy to estimate) is complicated. But that is so much better than human time management! If the expected

[computer-go] Opening

2007-06-17 Thread Jacques Basaldúa
Heikki Levanto wrote: I am sure there is a mathematically sound way to measure how symmetric the evaluation is, but my math is a bit rusty, so I am asking if someone can come up with a good way. After that, I'm asking if various programmers would be willing to run this test, and publish the

Re: [computer-go] OSS or Free Engines

2007-06-12 Thread Jacques Basaldúa
Hola, Eduardo Eduardo Sabbatella wrote: Take a look at: http://sourceforge.net/projects/javago I get a This project has not yet created any file release packages. message at the mentioned URL. In one of your recent posts you mentioned data mining and that is what I am interested to see how

Re: [computer-go] Efficiently selecting a point to play in a random playout

2007-05-31 Thread Jacques Basaldúa
Hola, Álvaro: Álvaro Begué wrote: Could not compile libego is not a very helpful error message. What exactly did the compiler complain about? My guess is that you don't have the required boost libraries installed. Yes. It must be that. I didn't know about boost libraries. Where can I find

[computer-go] June KGS Computer Go tournament: 19x19, an hour each

2007-05-30 Thread Jacques Basaldúa
We all think about UCT all day long, ;-) but universal time is called UTC. The acronym is far from obvious, because: The International Telecommunication Union http://en.wikipedia.org/wiki/International_Telecommunication_Union wanted Coordinated Universal Time to have a single abbreviation for

Re: [computer-go] Efficiently selecting a point to play in a random playout

2007-05-30 Thread Jacques Basaldúa
Lukasz Lew wrote: Is libego too complicated? Do You have problems with compilation? Or You are not comfortable with the GNU license? Any other reason? I only wanted to compare performance ( and see what good ideas you had ;-) ) but could not compile libego. Neither with Microsoft Visual

Re: [computer-go] 9x9 vs 19x19 (was: computer-go Digest)

2007-05-22 Thread Jacques Basaldúa
Heikki Levanto wrote: I think it is better to stick to 9x9 as the beginners tournament, where it is easy to test new ideas in quick games, and 19x19 as the serious tournament where we can see how good computers are at playing the game like we humans do. I agree 100%. Other board sizes are

Re: [computer-go] analysis of UCT and other bandit algorithms for tree search

2007-04-26 Thread Jacques Basaldúa
Remi Munos wrote: .. only when the tree is very smooth, ie. when the branches that appear good (from obtained rewards so far) are actually good and the branches that appear bad are truly bad. Go is a territorial game. I.e. an accumulative game, therefore when you loose territory you could

Re: [computer-go] 19x19 Go, scalability with time vs handicap

2007-04-24 Thread Jacques Basaldúa
Christoph Birk wrote: I am sure that Daniel is wrong here ... 2 kyu difference is more like 80% likelyhood of win. That depends on strength. Between a 20 and 22 kyu, it is even lower. But in professional play Daniel should be right. Note that 2 steps means 2 stones handicap. It is clear

[computer-go] Sylvain's results

2007-04-11 Thread Jacques Basaldúa
Thanks Sylvain. Sylvain Gelly wrote: The results are that in order to keep the same winning rate, you have to increase the number of simulations by something a little larger than linear in the board area. From 9x9 to 13x13, you need something like 3 times more simulations for the same

Re: [computer-go] Simplified MC evaluator ?explained?

2007-04-10 Thread Jacques Basaldúa
Weston Markham wrote: 1. Uniform playouts, as used in practice, are not really uniform over all legal go moves. Generally, pass moves are excluded until necessary, and moves that fill eyelike points are excluded. So, I assume that when you use the word legal, you mean admissible within

Re: [computer-go] The dominance of search (Suzie v. GnuGo)

2007-04-08 Thread Jacques Basaldúa
Don Dailey wrote: I have this idea that perhaps a good evaluation function could replace the play-out portion of the UCT programs. I thought about something similar but only for initializing the counters: introduce 10 fake playouts and estimate the number of wins by a function returning

Re: [computer-go] MoGo

2007-04-06 Thread Jacques Basaldúa
Hello Sylvain Sylvain Gelly wrote: If the program blundered as you said and still wins, it means that the program already won much earlier in the game. You are totally right. I said that in my post already. But what the user thinks is: 1. He was behind, but not desperately behind. 2. The

Re: [computer-go] MoGo

2007-04-06 Thread Jacques Basaldúa
Hello Don Don Dailey wrote: Many people DO play chess after the game is over. They continue to play a game long after they could have resigned. My example wasn't very good but I meant over literally. = The king is captured (changing the rules a little). How does Japanese make any

Re: [computer-go] professional knowledge

2007-04-06 Thread Jacques Basaldúa
Darren Cook wrote: All except joseki-knowledge is board-size independent. Maybe human player's adapt to different board sizes without even noticing. But if you try to model strategy with algorithms it is totally board size dependent. The extreme case is 5x5 where black 3,3 claims the four

Re:[computer-go] MoGo

2007-04-05 Thread Jacques Basaldúa
Sylvain Gelly wrote: You should also know that we never claimed that MoGo plays 9x9 go near the level of a professional go player, . . . Just curious: Do 9x9 professionals exist? When we say professional we mean 19x19 professional. Of course, there must be a correlation. One expects an

Re: [computer-go] MoGo

2007-04-05 Thread Jacques Basaldúa
Don Dailey wrote (about big/small wins) It actually surprises me that go players care about this ... One difference with chess is that you don't play chess after the game is over. The comparison could be: the king is captured, the loser keeps playing and then the winner gives the queen for

Re: [computer-go] Time Control for the new CGOS

2007-03-28 Thread Jacques Basaldúa
Is 10 minutes a standard and if so it is standard for 19x19 or 9x9? For 19x19 I find it a little too fast. I would prefer fastest: 4 sec/move (x240 moves) = 16 min slowest: 30 sec/move (x240 moves) = 2 hours I would like to try both. Usually fast because, as you pointed, you get useful

Re: [computer-go] Help me test CGOS

2007-03-28 Thread Jacques Basaldúa
Hellwig Geisse wrote: I just finished the translation of the old TCL script cgosGtp.tcl to plain C . . . Thanks Hellwig . I will try your program tomorrow. I prefer hacking C than tcl because its more transparent. I see what the tcl sends, but I don't know what details it may hide, and these

Re: [computer-go] Help me test CGOS

2007-03-27 Thread Jacques Basaldúa
Hi Don Could you provide some minimum protocol documentation so that we do not have to use any scripting language? The tcl script seems very simple. Is it possible to just open greencheeks.homelinux.org:6867, send login and then read/write commands? This way everyone would be free to implement

Re: [computer-go] Grid Cosmos

2007-03-13 Thread Jacques Basaldúa
Hi Yoshii I have some questions about your program. 1. Is this a complete go engine? I have renamed the gnugo.exe file in the /exe directory and the program no longer works. If it is not an engine, what is it? 2. The program has a 368 Mb file named Tree_Set in the /exe directory, the pattern

Re: [computer-go] MC - Estimating a moves true probability of winning

2007-03-02 Thread Jacques Basaldúa
Hello, Just an explanation on something I may have explained badly. I see we agree in the fundamental. Correcting bias in that estimate should lead to better sampling. This is usually called continuity correction http://en.wikipedia.org/wiki/Continuity_correction. The estimator is not

Re: [computer-go] Re: Big board. Temperature

2007-03-02 Thread Jacques Basaldúa
In CGT the temperature is the difference between the value if you play and the value if you pass. The name question should be answered by a native English speaker, but I guess it is an common use of the word hot. Let's call it hotness ;-) Jacques.

  1   2   >