Hi,
I started experimenting with implementing own Go robot and first I
created a generic infrastructure that various engines should be able to
plug into. Currently, a random player and a straightforward MonteCarlo
bot (plays as zzgobot on KGS now) engines are implemented; the sources
are at
On Mon, Nov 12, 2007 at 06:05:01PM +0100, Heikki Levanto wrote:
On Mon, Nov 12, 2007 at 04:58:49PM +0100, Petr Baudis wrote:
http://rover.dkm.cz/w/zzgo.git
I seem not to be able to find anything there. Is that link correct?
Sorry, it's http://repo.or.cz/w/zzgo.git
On Mon, Nov 12, 2007 at 08:45:14PM +0100, Heikki Levanto wrote:
On Mon, Nov 12, 2007 at 06:10:21PM +0100, Petr Baudis wrote:
Sorry, it's http://repo.or.cz/w/zzgo.git
I've had a quick look at it, and have already two comments:
1) You seem to use struct {x,y} for coordinates all the way. I
On Tue, Nov 13, 2007 at 07:10:30PM +0100, Gian-Carlo Pascutto wrote:
Petr Baudis wrote:
Hi,
is there any way to send game comments through kgsGtp on your own
(without the opponent triggering you)?
I think some possibility to send messages would be great. I could swear
I saw
On Mon, Nov 12, 2007 at 05:14:16PM -0500, Eric Boesch wrote:
On 11/12/07, Petr Baudis [EMAIL PROTECTED] wrote:
Does any frequently playing real-world bot use libEGO? It seems still
order of magnitude faster, but it looks like that is because it
simplifies some things too much. For example
On Mon, Nov 12, 2007 at 03:31:12PM -0500, Imran Hendley wrote:
Sorry I did not have time to look at your code, but here a few quick hints:
Thanks!
1) Before any optimization make sure that your code works 100%
correctly. This means testing extensively and writing tests that you
can use as
On Tue, Nov 13, 2007 at 03:32:03PM -0500, John Tromp wrote:
On Nov 13, 2007 2:48 PM, Petr Baudis [EMAIL PROTECTED] wrote:
I'm now somewhat torn. The speedup from using pseudo-liberty counts
could be huge, estimating from my profiling. On the other hand, it would
be very useful to still
On Tue, Nov 13, 2007 at 04:10:54PM -0500, John Tromp wrote:
Yes, you can generalize pseudoliberties by extending them
with another field, such that if the (summed) pseudoliberty field
is between 1 and 4, then the other (summed) field will tell you if all
these
are coming from a
On Wed, Nov 14, 2007 at 11:04:41AM -0500, Álvaro Begué wrote:
On Nov 14, 2007 10:54 AM, steve uurtamo [EMAIL PROTECTED] wrote:
I just wanted to point out that free() is not a system call. The heap is
handled by the
C library, and the OS is mostly not involved in it.
my bad.
Hi,
is anyone successfully using the kgs-chat GTP command in games?
I cannot get kgsGtp to send me the command when I make a comment inside
a game (as the bot's opponent). I receive the command when
I private-message the bot. Is there a special trick I need to do to
enable this inside games
On Mon, Nov 12, 2007 at 12:10:03PM -0800, Christoph Birk wrote:
My (not very optimized) C-code plays 12k games per seconds from the
opening position. The average game length is about 109 moves using
an early-termination rule when a big group gets captured that leaves
most stones beeing from
On Thu, Nov 15, 2007 at 12:13:33PM -0800, Christoph Birk wrote:
On Thu, 15 Nov 2007, Petr Baudis wrote:
This looks like a good technique I should implement too. What big
values are popular? I'm thinking size*size/3, but maybe that is too
conservative?
If there is a capture of more than 1
On Thu, Nov 15, 2007 at 04:41:19AM -0800, steve uurtamo wrote:
the more i think about it, the more i love whatever language
i'm using for whatever project i'm working on. some projects
would be (or are) horrifying to try to implement in some languages
[the matlab-C example springs to mind],
On Wed, Nov 21, 2007 at 09:11:25AM +, Stuart A. Yeates wrote:
In Perl you can have a line of code that looks like:
@result = (dothis $foo, $bar);
which could be either of:
@result = (dothis($foo), $bar);
@result = dothis($foo, $bar);
And the correct choice can vary each time the
On Thu, Nov 22, 2007 at 11:36:58AM +1800, Nick Apperson wrote:
Thanks for a great read there first of all.
But... and this is something I see time and again and I must admit I don't
really understand. What makes C faster than C++? I feel like C++ is all
about you don't pay for what you
On Thu, Nov 22, 2007 at 12:04:53PM +1800, Nick Apperson wrote:
right... well C++ does have this using virtual methods. I meant to add that
part.
I'm sorry, I still don't get this - what do virtual methods have to do
with branch prediction? Can you elaborate, please?
--
On Wed, Nov 21, 2007 at 11:47:05AM -0800, Dave Dyer wrote:
I disagree with almost everything Donn wrote.
Thanks to Moore's law, it is somewhere between unusual and rare for
the execution speed penalty of the language to matter, and if it
matters today (some but not all languages are fast
On Wed, Nov 21, 2007 at 09:16:48PM +0100, Raymond Wold wrote:
On Wed, 2007-11-21 at 14:11 -0500, Don Dailey wrote:
Experience in a language is a factor, but nobody refutes that properly
coded C is fastest (next to properly code assembly) and if performance
is your goal, then anything else
Hi,
On Mon, Nov 26, 2007 at 06:21:02PM +, Nick Wedd wrote:
The December 2007 KGS computer Go tournament will be next Sunday, December
2nd, in the Asian evening, European morning and American night, starting at
09:00 UTC (GMT) and ending about 13:00 UTC (GMT).
Each division will be a
On Tue, Nov 27, 2007 at 10:16:56AM -0800, terry mcintyre wrote:
From: Don Dailey [EMAIL PROTECTED]
On CGOS you will notice that a lot of the really good programs have to
throttle down. Even though they are capable of achieving 2500+ ELO
they are isolated at the top so they often choose
On Tue, Nov 27, 2007 at 08:40:11PM +0200, Harri Salakoski wrote:
command genmove w 30
reply=30 E3
cgos replys gameover 2007-11-27 B+Illegal do not understand syntax
GTP specc says : Success Responses - A successful response has one of the
syntaxes
=id
On Wed, Dec 05, 2007 at 10:16:20PM -0800, Sylvain Gelly wrote:
You should be using area scoring only and if you are playing handicap
games then either YOU or MOGO is not counting them the same. Or
perhaps Mogo has a bug in the handicap code.
MoGo uses KGS handicap counting (add 1
On Mon, Feb 18, 2008 at 09:03:17PM +0100, Alain Baeckeroot wrote:
Le lundi 18 février 2008, Michael Williams a écrit :
But as was pointed out before, these high levels of MoGo are probably still
not pro level, right?
On 9x9 Big_slow_Mogo is near pro level, maybe more.
6 monthes ago or
On Thu, Feb 28, 2008 at 01:01:08PM -0500, Don Dailey wrote:
It's naive to think some simplistic deception imposed upon the program
is going to correct the error when you don't even know if the program is
erring in the first place. How can you say, the program thinks it
is losing, but it
On Mon, Mar 03, 2008 at 11:01:52AM -0800, Christoph Birk wrote:
On Sun, 2 Mar 2008, Don Dailey wrote:
My feeling is that in lost positions, the only thing we are trying to
accomplish is to make the moves more cosmetically appealing (normal) and
at best improve the programs chances of winning
On Tue, Mar 04, 2008 at 12:01:02PM -0500, steve uurtamo wrote:
cool. do you have any examples from a 19x19 game? that's what
i was referring to when i said that i've never seen an MC player
play out a ko fight.
MoGo can indeed play out some rather spectacular ko fights;
unfortunately, I
On Wed, Mar 05, 2008 at 08:02:10PM +, Matthew Woodcraft wrote:
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
On Thu, Mar 06, 2008 at 12:55:53PM +, Jacques Basaldúa wrote:
A 4-6 kyu human is behind by 10-15 points in the midgame (at that stage the
probability of winning is correlated with territory, so the MC bot is
building fine.) He creates a 12-16 point worth nakade trick in a corner
and does
On Thu, Mar 06, 2008 at 04:33:16PM -0800, Dave Dyer wrote:
To a first order approximation, would changing the komi change the
rankings? Presumably, programs are playing the same number of games
as black and white, so any unfair advantage or disadvantage black
has would balance out.
Komi
On Fri, Mar 07, 2008 at 08:04:37AM -0500, Thomas Wolf wrote:
I assume that when you change komi dynamically, all that was learned
by MC so far under the different komi value is useless/wrong.
But what are actually your reuse rates? With the standard UCB1 formula,
I find reusing branches from
On Fri, Mar 07, 2008 at 12:43:42PM -0500, Don Dailey wrote:
I honestly think there are better ways to handle this, if you must,
other than changing the goal to a losing goal. At least give the
computer the right goal (winning) and adjust from there.
If I were trying to solve this
Hi,
I'm wondering about how to best make my Monte Carlo playouts within
UCT heavier and which pieces of domain knowledge to better use to bias
the tree and which ones to apply during the playouts, so I would like to
ask about previous experiences.
Currently, I do three basic hints that I
Hi,
does anyone know of any pre-made open framework using genetic
algorithms that one could use to tune various parameters of a bot? I
have about 6 independent parameters for my bot so far that I would like
to find best values for (from domain-specific knowledge hint rates to p
parameter of
On Sat, Mar 08, 2008 at 09:03:47AM -0500, Don Dailey wrote:
Moi de Quoi wrote:
On Sat, 2008-03-08 at 07:41 -0500, Don Dailey wrote:
But the problem you are seeing is a bug in the server. I am restarting
the server which will be a temporary cure. It seems there are
On Sat, Mar 08, 2008 at 04:56:16PM +0100, Moi de Quoi wrote:
Well, it depends on the server's policy, whether you are allowed to be
logged in more than once under the same name. (I can see no good reason
to allow it, but it's Don's server...)
I do not want to be logged in more than once, I
On Mon, Mar 10, 2008 at 02:33:03AM -0400, Michael Williams wrote:
Jonas Kahn wrote:
out, kos can go on for long. I don't know what depth is attained in the
tree (by the way, I would really like to know), but I doubt it is that
MoGo displays the depth of the principle variation in the stderr
Hi,
On Sat, Mar 08, 2008 at 10:18:34AM +0100, Petr Baudis wrote:
(By the way, pachi1-*-light are UCT bots with completely light
playouts with various UCB1 c values, if anyone wants to use that as
reference. Surprisingly, it seems that my heavy playouts do not make big
difference so far
On Mon, Mar 10, 2008 at 03:40:53PM -0700, Christoph Birk wrote:
On Mon, 10 Mar 2008, Petr Baudis wrote:
With 110k playouts per move and no domain knowledge in the playouts,
the ratings are now:
c=0.2 (pachi1-p0.2-light) ELO 1627 (285 games)
c=1.0 (pachi1-p1.0-light
On Mon, Mar 10, 2008 at 06:57:07PM -0400, Don Dailey wrote:
I think you may still have a bug. You should get well over 1700 with
110,000 playouts, even if they are light playouts.
Hmmm... That is going to be some tough debugging I suspect.
I'm pretty sure my code is fairly well debugged
Hi,
On Mon, Mar 10, 2008 at 08:07:14PM -0400, Don Dailey wrote:
What is the justification of using the parent playout count instead of
the node playout count itself?
I don't know if it makes much difference how this is done, and I don't
know how everybody else is doing it. I
On Mon, Mar 10, 2008 at 05:36:14PM -0700, Christoph Birk wrote:
On Mon, 10 Mar 2008, Christoph Birk wrote:
On Tue, 11 Mar 2008, Petr Baudis wrote:
On Mon, Mar 10, 2008 at 06:57:07PM -0400, Don Dailey wrote:
I think you may still have a bug. You should get well over 1700 with
110,000 playouts
On Mon, Mar 10, 2008 at 10:04:18PM -0400, Don Dailey wrote:
Your method is to allocate 1 node when it's been visited once or twice -
very natural I agree. My method is to allocate all the children at
once, and wait until the parent has been visited some number of times
(currently 100). If
On Tue, Mar 11, 2008 at 11:41:41AM -0700, Christoph Birk wrote:
On Tue, 11 Mar 2008, Don Dailey wrote:
I am going to keep the 25k playouts running and add a 10k play-out
version of UCT. I want to establish a standard testing size so that
Great! That way Jason can also participate.
On Thu, Mar 13, 2008 at 09:19:58AM -0400, Don Dailey wrote:
If it is agreed, I will start a 25k test.My prediction is that this
will finish around 1600 ELO on CGOS.
Looks like it's currently around 1485, so I am 115 ELO off from my
prediction at the moment.
2 more
On Thu, Mar 13, 2008 at 09:19:58AM -0400, Don Dailey wrote:
If you look at the table, drdGeneric 10k is rated 1228 and 25k is
rating 1485 which is 257 ELO for doing 2.5 x more play-outs. If
this holds, I would expect 100,000 play-outs to give well over 1700
ELO. Of course there could
On Thu, Mar 13, 2008 at 11:25:04AM -0700, Christoph Birk wrote:
On Thu, 13 Mar 2008, Petr Baudis wrote:
So I have created this page:
http://senseis.xmp.net/?CGOSBasicUCTBots
and summed up what I could find in the thread about the various bots.
Please clarify if anything there is wrong
On Thu, Mar 13, 2008 at 08:01:57PM +0100, Heikki Levanto wrote:
On Thu, Mar 13, 2008 at 06:31:19PM +0100, Petr Baudis wrote:
I got kind of lost in the thread and lost track about which bots should
I actually compare myself to. ;-)
So I have created this page:
http
Hi,
On Fri, Mar 21, 2008 at 10:14:49AM +, Nick Wedd wrote:
Saturday:
3/23/08 3:00 PM
Game I (9x9)
Game II 9x9
Game III 9x9
Played with 1.5 hours from the start of one round to the next
will this be with komi 7.5?
Petr Pasky Baudis
On Fri, Mar 21, 2008 at 05:07:01PM +0100, Olivier Teytaud wrote:
will this be with komi 7.5?
Yes. Previous records against Guo Juan, as far
as I know:
- 1/3 wins with komi 7.5
- 9/14 wins with komi 0.5 (mogo black,
i.e. komi in favor of mogo)
What computing power did have
On Fri, Mar 21, 2008 at 08:35:25PM +0100, Olivier Teytaud wrote:
What computing power did have that MoGo at its disposal?
4 cores, 2.4 GHz.
Thank you! That also puts the strength of CzechBot into some
perspective. :-)
Petr Pasky Baudis
On Wed, Mar 26, 2008 at 08:50:28PM -0700, David Fotland wrote:
You are right.
Well, I did compete for this prize about 15 times, so I hope so :)
Are there any current prized computer tournaments or does anyone know
about Ing foundation or anyone else planning to resume the challenge?
What
On Thu, Mar 27, 2008 at 12:14:06PM -0700, terry mcintyre wrote:
Suppose a group can be defended - four liberties in a
row, for example. If the opponent plays inside those
four liberties, you play to divide the area into two
eyes - unless the situation is such that the group has
a second eye
On Mon, Mar 31, 2008 at 03:12:39PM -0700, Christoph Birk wrote:
On Mar 31, 2008, at 1:05 PM, Don Dailey wrote:
Christoph Birk wrote:
On Mar 31, 2008, at 10:48 AM, Mark Boon wrote:
I don't know about this. I'm pretty sure MoGo checks if the stone can
make at least two liberties (ladder
Hello,
On Sun, Apr 06, 2008 at 08:55:26PM -0600, David Silver wrote:
Here is a draft of the paper, any feedback would be very welcome :-)
http://www.cs.ualberta.ca/~silver/research/publications/files/MoGoNectar.pdf
you are saying that in minimax, opponent moves are selected by
minimizing
On Mon, Apr 07, 2008 at 08:36:25AM -0400, Jason House wrote:
On Apr 7, 2008, at 8:22 AM, Petr Baudis [EMAIL PROTECTED] wrote:
Hello,
On Sun, Apr 06, 2008 at 08:55:26PM -0600, David Silver wrote:
Here is a draft of the paper, any feedback would be very welcome :-)
http://www.cs.ualberta.ca
On Wed, Apr 09, 2008 at 03:40:28PM -0700, Christoph Birk wrote:
On Wed, 9 Apr 2008, Matthew Woodcraft wrote:
It might be that most of those games aren't visible to the rating
system.
That might explain why a rating system may have a hard time
to follow.
Bad data in ... bad data out :-)
But
On Sun, Jan 11, 2009 at 01:24:32PM +, Nick Wedd wrote:
I suggest that instead of getting your neural players to play Go, you get
them to play a very slightly different game, in which, when both players
pass in turn, all stones remaining on the board are deemed alive. It is
not
On Sat, Jan 10, 2009 at 02:21:17PM +0100, Gian-Carlo Pascutto wrote:
Mark Boon wrote:
So it seems arbitrary to put limitations on the hardware. However, if
two programs are essentially the same, but one side manages to bring
a more powerful computer than the other, is it fair to award one
On Sat, Jan 17, 2009 at 08:29:32PM +0100, Sylvain Gelly wrote:
A small point: in PlayoutOutTree, just after if
(!played.AlreadyPlayed(move)) {, there should have a played.Play(move).
I believe it does not change the final result (as the check is also done in
the backup, and the move played in
Hi!
Just FYI, someone might find interesting that latest SVN of Fuego
still does not seem to be on par with Mogo public release 1 (not that it
would claim to be - I was just curious where do they stand against each
other).
I ran 86 19x19 games with both on the same hardware (single core of
On Mon, Feb 16, 2009 at 10:27:51AM +0900, Yamato wrote:
I ran 86 19x19 games with both on the same hardware (single core of
A64 X2 6000+, 2G RAM) with 20 minutes S.D. each, the rate is MoGo win
83.3% (+-4.1).
How did you set the time to 20 minutes S.D.? MoGo doesn't update the
clock if
Hi!
On Mon, Apr 06, 2009 at 03:52:14PM +0100, Nick Wedd wrote:
The results of yesterday's KGS bot tournament are now available at
http://www.weddslist.com/kgs/past/46/index.html
As usual, I expect there are mistakes, and I will welcome corrections.
Congratulations to the winner, CzechBot!
Hi!
Today there was a short discussion about the strongest bot currently
online on KGS and I got curious whether ManyFaces or CzechBot (bleeding
edge MoGo) is stronger, so I made it play against ManyFaces.
CzechBot is running as dual-thread pondering MoGo on slightly loaded
dual-core
Hi!
Today there was a short discussion about the strongest bot currently
online on KGS and I got curious whether ManyFaces or CzechBot (bleeding
edge MoGo) is stronger, so I made it play against ManyFaces.
CzechBot is running as dual-thread pondering MoGo on slightly loaded
dual-core
On Sat, Aug 15, 2009 at 08:33:31AM -0400, Jason House wrote:
On Aug 15, 2009, at 8:22 AM, Don Dailey dailey@gmail.com wrote:
2009/8/15 Jason House jason.james.ho...@gmail.com
On Aug 14, 2009, at 11:02 PM, David Fotland fotl...@smart-
games.com wrote:
Moves often merge two groups.
On Sat, Aug 15, 2009 at 09:13:02AM -0600, Brian Sheppard wrote:
assuming komi 7.5 and Chinese rule, playing at J3 white will win. After J3,
white has 35. It only needs to win the ko or takes two dames. If black
fills
the dame, it loses the ko. If it fills the ko, white can take two dames.
On Thu, Aug 20, 2009 at 10:47:12AM +0200, Ingo Althöfer wrote:
Does someone here know the exact starting times (in common
time zones) for the 19x19 exhibition games
* Chou(9p) vs MFoG (Fr, August 21)
* Chou(9p) vs Zen (Sa, August 22)
?
Hi!
Where were these announced? On what server will
Hi!
On Thu, Sep 10, 2009 at 12:26:06PM +0100, Christian Nentwich wrote:
I strongly suspect the low performance in the per-intersection case
is down to two reasons - please let me know what you think:
- A big proportion of moves place one stone on an empty
intersection. In this case 80 of
On Thu, Sep 10, 2009 at 01:54:49PM +0200, Magnus Persson wrote:
This is very interesting. Here is a crazy idea, maybe it the same as
Marks but I want to take it to its extreme.
Since AMAF values are so helpful, perhaps one can let go of the idea
of sequential play following the rules of go,
On Thu, Sep 10, 2009 at 08:29:31AM -0400, Jason House wrote:
I've thought of something similar in the past, but with a twist:
pre-compute a subset of moves that could be safely played in
parallel. Even if you can only play 285 moves in parallel on an
empty 19x19, it could still be a huge speed
On Thu, Sep 10, 2009 at 09:18:40AM -0400, Jason House wrote:
Somewhat... One could generate a random number (r) and combine it
with the move mask (m) as follows:
black moves = m r
white moves = m ~r
This has the drawback that the number of black and white moves may
not be equal. It can
On Sun, Sep 13, 2009 at 01:02:40AM +0200, Vincent Diepeveen wrote:
On Sep 10, 2009, at 12:55 AM, Michael Williams wrote:
Very interesting stuff. One glimmer of hope is that the memory
situations should improve over time since memory grows but Go
boards stay the same size.
Well you
On Sun, Sep 13, 2009 at 10:48:12AM +0200, Vincent Diepeveen wrote:
On Sep 13, 2009, at 10:19 AM, Petr Baudis wrote:
Just read the nVidia docs. Shifting has the same cost as addition.
Document number and url?
http://developer.download.nvidia.com/compute/cuda/2_3/toolkit/docs
On Tue, Sep 15, 2009 at 08:49:15AM -0700, David Fotland wrote:
Simple playouts with no eye fills and mogo 3x3 patterns and basic uct beat
Gnugo 40% (at version 120)
..snip..
All win rates are on 9x9 vs gnugo 3.7.20 level 10 with 5000 playouts. After
this I switched to testing 19x19, and
On Mon, Sep 21, 2009 at 08:18:37AM -0700, David Fotland wrote:
In the original Mogo paper it's the initial value for the children, rather
than try every child once.
Ah, you mean the First Play Urgency! Thanks, I will try that.
Anyway, I'm happier; after fixing many bugs and improving my
Hi!
Pachi has two RAVE/AMAF modes - in one, it counts as RAVE wins only
moves made by the player further down in the tree. In the other, it also
counts in the moves made in the playout phase.
I think most people collect AMAF statistics only from the tree phase,
at least that's what I've
On Tue, Sep 29, 2009 at 12:12:07PM +0200, Stefan Kaitschick wrote:
Every now and then somebody submits an interesting 9*9 problem,
usually rendered in X and O.
Wouldn't it be great to have a public suite, lets say a directory
with black to play and win sgf problems.
For quick testing there
Hi!
On Mon, Sep 28, 2009 at 04:09:31PM -0600, Brian Sheppard wrote:
By now, I should probably find better reference opponent than
gnugo... I wonder if to pick fuego or mogo... ;-) Strength is probably
not _as_ important as the variety of techniques used in order to avoid
selective blindness
Hi!
On Tue, Sep 29, 2009 at 01:45:40PM +0200, Lars Schäfers wrote:
I remeber a version where the call was just
./cgosview-linux-x86_32 cgos.boardspace.net 6819
Ahh, thanks, that works. I think the website should be fixed then. :-)
Petr Pasky Baudis
I guess I'm not really appreciating the difference between node value
prior and progressive bias - adding a fixed small number of wins or
diminishing heuristic value seems very similar to me in practice. Is the
difference noticeable?
On Tue, Sep 29, 2009 at 08:25:56AM -0700, David Fotland wrote:
On Thu, Oct 01, 2009 at 04:50:26PM -0400, dhillism...@netscape.net wrote:
For 9x9 games, when I added progressive widening to AntiGo (before I added
RAVE), it was low hanging fruit. I used my old program Antbot9x9 for the move
ranking and got a very nice strength increase for very little
On Fri, Oct 02, 2009 at 09:52:35PM -0600, Martin Mueller wrote:
Yes, Fuego uses just the 3x3 patterns; its strength is surprising. :-)
Someone conjenctured it is because of how well-tuned its constants
are.
I also think large part of it is that it seems to use perfect nakade
solver in
Hi!
On Tue, Oct 06, 2009 at 12:08:06AM +0900, Hideki Kato wrote:
Nick Wedd: vgf6vg99igykf...@maproom.demon.co.uk:
I wonder if anyone here has a URL for either the Tainan Computer Go
Tournament, to take place on October 30th and 31st, or the GPW Cup,
November 13th and 14th? I cannot read
On Thu, Oct 08, 2009 at 01:48:05PM -0600, Brian Sheppard wrote:
About two weeks ago I took Pebbles offline for an extensive overhaul of its
board representation. At that time Valkyria 3.3.4 had a 9x9 CGOS rating of
roughly 2475.
When I looked today, I saw Valkyria 3.3.4 rated at roughly
On Tue, Oct 06, 2009 at 01:33:10AM +0900, Hideki Kato wrote:
The tournament in Taiwan allows playing through KGS but according to
the rules http://ai.csie.ndhu.edu.tw:9898/eng/p_7.htm, it's
preferred to participate at least one person from each team. If not,
the entry fee will be doubled.
Hi!
Is there some high-level reason hypothesised about why there are
no successful programs using neural networks in Go?
I'd also like to ask if someone has a research tip for some
interesting Go sub-problem that could make for a nice beginner neural
networks project.
Thanks,
--
On Wed, Oct 14, 2009 at 05:12:58PM +0200, Heikki Levanto wrote:
On Wed, Oct 14, 2009 at 03:34:59PM +0300, Petri Pitkanen wrote:
Neural network tend to work well in those cases where evaluation function is
smooth, like backgammon. Even inbackgammon neural networks do give good
results if
On Tue, Oct 20, 2009 at 08:10:47AM +0100, Christian Nentwich wrote:
Go has been played long enough, and the proposed great wall
opening is simple enough, that is should be more than valid to argue
that if it was a good opening, it would be played more often.
A possible explanation I have read
Hi!
Does anyone know of any handicap games collection? For MCTS handicap
play research, I'm looking for at least 500 games of any handicap (1-9),
ideally of reasonable strength (KGS 2k-9d).
Thanks!
--
Petr Pasky Baudis
A lot of people have my books on
On Tue, Oct 20, 2009 at 03:06:20PM +0200, Ben Lambrechts wrote:
You can filter them from the following collections:
http://u-go.net/gamerecords/
http://u-go.net/gamerecords-4d/
Thanks! I didn't want to use the main collection, but I wasn't aware of
the 4d+ collection, which looks perfect!
On Fri, Oct 23, 2009 at 01:34:29PM -0600, Brian Sheppard wrote:
I have not done any GPU experiments, so readers should take my guesswork
FWIW. I think the code that is light is the only piece that parallelizes
efficiently. Heavy playouts look for rare but important situations and
handle them
Hi!
(Sylvain et al. 2006) describes the use of CFG-based zones in random
simulations to simulate only the local position and tune the score based
on few thousands of simulations of outside of the zone. It doesn't seem
the idea is too practical (especially with RAVE, but there seem to be
more
On Sun, Oct 25, 2009 at 06:52:56PM +0100, Folkert van Heusden wrote:
What method are you guys using for the monte carlo search? What do you
do?
I pick at random a move, then for the resulting board that comes out of
that pick another move and so on.
Then, after that I evaulate the number of
On Sun, Oct 25, 2009 at 10:16:58PM +0100, Folkert van Heusden wrote:
What method are you guys using for the monte carlo search? What do you
do?
I pick at random a move, then for the resulting board that comes out of
that pick another move and so on.
Then, after that I evaulate the
Hi!
On Mon, Oct 26, 2009 at 07:19:45PM +0100, Olivier Teytaud wrote:
For information, our Taiwanese partners(**) for a ANR grant have organized
public demonstration games between
Thanks for the information!
MoGoTW (based on MoGo 4.86.Soissons + the TW modifications developped
On Mon, Oct 26, 2009 at 04:20:24PM -0400, Don Dailey wrote:
Peter, did your comment get cut off?
Oops, indeed. Prone to tactical mistakes in high time pressure is what
I meant to say.
Anyway, I agree with you on this. Humans are not stronger on short time
settings. I believe that
On Tue, Oct 27, 2009 at 08:47:41AM +0100, Olivier Teytaud wrote:
Could you give us at least a general picture of improvements compared to
what was last published as
www.lri.fr/~teytaud/eg.pdfhttp://www.lri.fr/%7Eteytaud/eg.pdf? Is it
just
further tuning and small tweaks or are you
Hi!
Does anyone have some tips about designing effeciently-computable
coefficient functions that would have similar properties to the eg
paper?
I.e. something like:
a + b + c = 1
n(d) = 0 a ~ 0, b ~ 0, c ~ 1
n(d) smallb a
n(d) - infty a ~ 1
It sounds simple, but I'm a
On Tue, Oct 27, 2009 at 06:32:44PM +0200, Olivier Teytaud wrote:
AIUI, once upon N simulations in a node you take let's say the node with
the lowest value, pick one son of it at random within the tree and start
a simulation?
I'll try to write it clearly (for binary deterministic
On Tue, Oct 27, 2009 at 01:10:09PM +0100, Stefan Kaitschick wrote:
But, has anyone gathered stats on positions, from real games, that
require precise play by the defender/attacker/both/neither? Is defending
really easier than attacking?
Darren
Who is the defender?
One side is
1 - 100 of 309 matches
Mail list logo
