Re: [computer-go] ego110_allfirst on CGOS
I'm glad that you found it. seems like winRate would end up as 0 most of the time and thus you'd make nearly random moves? s. Shape Yahoo! in your own image. Join our Network Research Panel today! http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] ego110_allfirst on CGOS
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Excellent, I am glad you found the issue. I had a strong sense it
would be something like this.
This is a fairly common type of bug in software and is why I suggested
it, however I think it's the first time I've ever directed someone to
the specific line of code without having ever seen the code! (Of
course it was a wild stab in the dark.)
- - Don
Jason House wrote:
On Fri, 2007-09-28 at 18:43 -0400, Don Dailey wrote:
Somewhere in your program I am sure there is a head-slapping error you
will find and when you do you will scream out loud! Don't give up.
Yup... A nice little omission of casting.
int nwins, nsims;
...
double winRate = nwins/nsims;
...
write(Picking move {} with win rate of, move, nwins*1.0/nsims);
What kept me from seeing this immediately (AKA bot plays
A1,A2,A3...A9,B1,B2...B9,C1...C9,...) is that there's logic that looks
at two moves with the same winning rate and picks the one with the
higher number of simulations. That can really help in non-AMAF
implementations where the number of sims are low.
It turns out that the code that I copied/pasted to make a quick AMAF bot
used floating point numbers for nwins and nsims... My so-called monte
carlo transposition reuse (MCTR) where they have to be floating point
values.
I kicked off hb-amaf-1k-v2. While it lost its first game online, it's
play was WAY better than it has been. It's rating will certainly settle
much higher.
I hope that was the issue and there are no more biggies like that! This
discussion was very helpful for me to dig up all kinds of small bugs in
other parts of my code. Once I confirm this is the big bug, I'll try
variants with the different handling to see how that affects the rating
(I'd call all past tests of variations worthless given the nature of
this bug)
While this may give an 800 ELO boost to the AMAF bot, I'm disappointed
that it won't help my other versions like that. The other changes dug
up in this long discussion will likely give them some boost though.
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFG/kqFDsOllbwnSikRAlNYAJ9R4yriIDN8UleHjlK/Ex7SqsjQVgCfaeL2
8MHiz56EXGa5GLZ4hDt/QFo=
=1nFu
-END PGP SIGNATURE-
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] ego110_allfirst on CGOS
On Sat, 2007-09-29 at 03:45 -0700, steve uurtamo wrote: I'm glad that you found it. seems like winRate would end up as 0 most of the time and thus you'd make nearly random moves? Yes, winRate would be 0.0 (unless every random simulation with that point was a victory such as in late endgame). The net effect is that it was playing the move with the most simulations, which is random. ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] ego110_allfirst on CGOS
Strangely enough, it now appears that hb-amaf-1k-v2 is significantly
stronger than genAnchor-1k, defeating it 9 out of 9 times. I still have
to wonder what the cause of the strength difference is.
On Sat, 2007-09-29 at 08:52 -0400, Don Dailey wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Excellent, I am glad you found the issue. I had a strong sense it
would be something like this.
This is a fairly common type of bug in software and is why I suggested
it, however I think it's the first time I've ever directed someone to
the specific line of code without having ever seen the code! (Of
course it was a wild stab in the dark.)
- - Don
Jason House wrote:
On Fri, 2007-09-28 at 18:43 -0400, Don Dailey wrote:
Somewhere in your program I am sure there is a head-slapping error you
will find and when you do you will scream out loud! Don't give up.
Yup... A nice little omission of casting.
int nwins, nsims;
...
double winRate = nwins/nsims;
...
write(Picking move {} with win rate of, move, nwins*1.0/nsims);
What kept me from seeing this immediately (AKA bot plays
A1,A2,A3...A9,B1,B2...B9,C1...C9,...) is that there's logic that looks
at two moves with the same winning rate and picks the one with the
higher number of simulations. That can really help in non-AMAF
implementations where the number of sims are low.
It turns out that the code that I copied/pasted to make a quick AMAF bot
used floating point numbers for nwins and nsims... My so-called monte
carlo transposition reuse (MCTR) where they have to be floating point
values.
I kicked off hb-amaf-1k-v2. While it lost its first game online, it's
play was WAY better than it has been. It's rating will certainly settle
much higher.
I hope that was the issue and there are no more biggies like that! This
discussion was very helpful for me to dig up all kinds of small bugs in
other parts of my code. Once I confirm this is the big bug, I'll try
variants with the different handling to see how that affects the rating
(I'd call all past tests of variations worthless given the nature of
this bug)
While this may give an 800 ELO boost to the AMAF bot, I'm disappointed
that it won't help my other versions like that. The other changes dug
up in this long discussion will likely give them some boost though.
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFG/kqFDsOllbwnSikRAlNYAJ9R4yriIDN8UleHjlK/Ex7SqsjQVgCfaeL2
8MHiz56EXGa5GLZ4hDt/QFo=
=1nFu
-END PGP SIGNATURE-
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] ego110_allfirst on CGOS
i don't mean necessarily that it isn't stronger, just that now is the time to not worry so much about things. if anything, try to make it even stronger! you've got another 1200 ELO to go, after all. s. - Original Message From: Jason House [EMAIL PROTECTED] To: computer-go computer-go@computer-go.org Sent: Saturday, September 29, 2007 10:44:55 AM Subject: Re: [computer-go] ego110_allfirst on CGOS Statistically speaking, I can say with 99.9% confidence that hb-amaf-1k-v2 is stronger than genAnchor-1k. No need to wait any longer to conclude that :) The math to compute the one-tailed p-value for rejecting the null hypothesis that the bots are the same strength is left as an exercise for the reader ;) On Sat, 2007-09-29 at 07:20 -0700, steve uurtamo wrote: Strangely enough, it now appears that hb-amaf-1k-v2 is significantly stronger than genAnchor-1k, defeating it 9 out of 9 times. I still have to wonder what the cause of the strength difference is. now is the time to just let it run for a few weeks and check on the stats later. s. Tonight's top picks. What will you watch tonight? Preview the hottest shows on Yahoo! TV. http://tv.yahoo.com/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ Catch up on fall's hot new shows on Yahoo! TV. Watch previews, get listings, and more! http://tv.yahoo.com/collections/3658 ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
[computer-go] Games Database
Good Evening, Sorry, I think this may have been asked but can't find it anywhere in the list archives. Anyone know of a good place to get a large sgf game database? Preferably master games or at least 1dan level or higher? -Josh ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] ego110_allfirst on CGOS
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm starting to get curious. What are you doing that is causing it to win 11 out of 11 against genAnchor_1k and yet it's only 113 ELO stronger?And it's supposedly an identical program?I don't think it's actually identical and I don't trust that you did everything the same. Of course I could have a bug too, but it doesn't explain why your program is much weaker than the results against genAnchor would indicate. How many all-as-first moves are you considering? What are the possible differences? 1. eye rule 2. number of all-as-first moves (7/8 for me + 1) 3. quality of RNG 4. correctness of random move selection strategy. 5. depth at which you stop a game. (about 1000 moves for me.) 6. stopping rule. (both program have no non-eye filling moves.) I just thought of something. I think I initialize the statistics array with 1 draw per move as a cheesy way to avoid divide by zero error. Could this be affecting the performance? Perhaps at low levels like this it has a noticeable effect?Would it make the program especially vulnerable to an identical program that doesn't do this? - - Don Jason House wrote: Statistically speaking, I can say with 99.9% confidence that hb-amaf-1k-v2 is stronger than genAnchor-1k. No need to wait any longer to conclude that :) The math to compute the one-tailed p-value for rejecting the null hypothesis that the bots are the same strength is left as an exercise for the reader ;) On Sat, 2007-09-29 at 07:20 -0700, steve uurtamo wrote: Strangely enough, it now appears that hb-amaf-1k-v2 is significantly stronger than genAnchor-1k, defeating it 9 out of 9 times. I still have to wonder what the cause of the strength difference is. now is the time to just let it run for a few weeks and check on the stats later. s. Tonight's top picks. What will you watch tonight? Preview the hottest shows on Yahoo! TV. http://tv.yahoo.com/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG/wSKDsOllbwnSikRAm+EAJ4+PrLMO7Uyf3ejn61nzZcwOgyvnACeL0tD acGx8Fhu1Co5fk8OMvDFuhE= =fFvv -END PGP SIGNATURE- ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
