Re: [Computer-go] Optimistic Majority Parallelization

> I tried parallelization of two machines. > > 1. Run same program independently on 2 machines. > 2. A plays E5 and its winrate is 0.51 > 3. B plays F3 and its winrate is 0.53 > 4. Program plays F3 (select higher winrate move). > > Thus, very simple. > Use same hardware, and random seed must be d

[Computer-go] Optimistic Majority Parallelization

I tried parallelization of two machines. 1. Run same program independently on 2 machines. 2. A plays E5 and its winrate is 0.51 3. B plays F3 and its winrate is 0.53 4. Program plays F3 (select higher winrate move). Thus, very simple. Use same hardware, and random seed must be different. Resul

Re: [Computer-go] UEC2010 tournament pairing

Aja wrote: >> >Fuego won the tournament. >> >> Congratulations to Fuego. >> Was it the same version as in SVN or do you have any secret improvements? >> What was the command line you used? > >Hi Yamato, > >I asked Professor Mueller and Rich the same question yesterday. They replied >that this vers

Re: [Computer-go] UEC2010 tournament pairing

>Fuego won the tournament. Congratulations to Fuego. Was it the same version as in SVN or do you have any secret improvements? What was the command line you used? Hi Yamato, I asked Professor Mueller and Rich the same question yesterday. They replied that this version of Fuego is almost ident

Re: [Computer-go] UEC2010 tournament pairing

>Fuego won the tournament. Congratulations to Fuego. Was it the same version as in SVN or do you have any secret improvements? What was the command line you used? -- Yamato ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/

Re: [Computer-go] The 4th UEC Cup, today and tommorow

Ingo Althöfer: <20101128152718.215...@gmx.net>: >Hi Hideki, > >thnaks for your information. > >> Attached is the (unofficial) game record of Zen vs. Kaori Aoba 4p. >> Zen ran on a pc cluster (6 x 4 GHz Intel Xeon and 5 x 4 x 3 GHz Intel >> Core2). > >I have two questions on this sgf: >*

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

At the end of the game mcts is pretty good at life and death, but rarely gets the exact score. > -Original Message- > From: computer-go-boun...@dvandva.org [mailto:computer-go- > boun...@dvandva.org] On Behalf Of Dave Dyer > Sent: Sunday, November 28, 2010 11:02 AM > To: computer-go@dvandv

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

> >Back to Dyer's research into scoring games "where humans typically finish >them" - did that include wins by resignation? No, only games where an exact score was published. ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

At 07:16 AM 11/28/2010, Michael Williams wrote: >On Sun, Nov 28, 2010 at 4:15 AM, Dave Dyer wrote: >> At 10:39 PM 11/27/2010, David Fotland wrote: >>>Accurate scoring, even at the end of a game, is very difficult. You have to >>>read accurately, and evaluate semeai and seki. >> >> Yup. I spent y

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

It's not much more difficult to score Japanese (vs Chinese) with MCTS. You still have to implement seki in the playouts to score the end of a game. I agree with you that if I were starting from scratch I'd write an MCTS solver. Still hundreds of hours of work. > -Original Message- > Fro

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

From: Magnus Javerberg One way around this would be to not do the scoring on the phone, but on a server. This way you can use a DLL, and if the image processing is done in the phone the data traffic should be very low. Maybe not an optimal solution, but IMH

Re: [Computer-go] The 4th UEC Cup, today and tommorow

Hi Hideki, thnaks for your information. > Attached is the (unofficial) game record of Zen vs. Kaori Aoba 4p. > Zen ran on a pc cluster (6 x 4 GHz Intel Xeon and 5 x 4 x 3 GHz Intel > Core2). I have two questions on this sgf: * Which were the rules of play? Chinese or Japanese or ...? * In "g

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

On Sun, Nov 28, 2010 at 10:16 AM, Michael Williams wrote: > ... For instance, if I wanted to score a > Chinese endgame starting with no code, I would probably write an MCTS > move generator that played the game to the bitter end and then score > it using simple area counting.  But you can't do the

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

On Sun, Nov 28, 2010 at 4:15 AM, Dave Dyer wrote: > At 10:39 PM 11/27/2010, David Fotland wrote: >>Accurate scoring, even at the end of a game, is very difficult.  You have to >>read accurately, and evaluate semeai and seki. > > Yup.  I spent years developing the capability to score endgames > at

Re: [Computer-go] The 4th UEC Cup, today and tommorow

The results of the 2nd day of the 4th UEC Cup is here . About the Exhibition games, Zen won against Kaori Aoba 4p with 6 stones handicap and Fuego lost against Meiko Tei 9p with 6 stones handicap. Attached is the (unofficial) game record of Zen vs

[Computer-go] intel i5 760 vs amd Phenom II X6 1055T

These two has simular prices. Any one know which cpu is better for Go MC simulation? ___ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

One way around this would be to not do the scoring on the phone, but on a server. This way you can use a DLL, and if the image processing is done in the phone the data traffic should be very low. Maybe not an optimal solution, but IMHO clearly acceptable and much better then spending several hun

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

>> I'm starting to suspect that I'll be forced to implement this myself, >> obnoxious and time consuming as that is. i'm starting to suspect that you regard this as a solved problem, somehow. it's very unsolved, merely approximated. the most obnoxious part about it isn't "trivial details", it's re

Re: [Computer-go] I need an off-the-shelf final position live/dead evaluator

At 10:39 PM 11/27/2010, David Fotland wrote: >Accurate scoring, even at the end of a game, is very difficult. You have to >read accurately, and evaluate semeai and seki. Yup. I spent years developing the capability to score endgames at the point where humans typically leave them. Getting within