Re: [computer-go] Black/White winning rates with random playout?
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 difficult to write a scoring algorithm for this game. I believe this game is called Go with Tromp-Taylor rules. ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] UEC cup
Nick Wedd wrote: So what _is_ reality nowadays? Your previous email did not make this clear. Are Japanese pro grades now closer together than a third of a stone, or farther apart? The reality is that the correlation between ranks and playing strengths is very low, and that knowing that player A is x-dan professional and player B is y-dan professional does not actually tell you very much about how likely player A is to win against player B. Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] UEC cup
dave.de...@planet.nl wrote: What you are saying is that many professionals are overrated or underrated (sometimes by as much as two stones). No, what I'm saying is that professional ranking systems are not meant to be rating systems and should not be treated as if they were rating systems. (As Hideki pointed out, professional ranks never go down, only up.) The set of professional players whose strength is within a third of stone of Lee Sedol or Lee Changho is surely smaller than the set of professional players who are ranked 9dan. If you wanted a rating system which tops out at 9dan, you'd have to demote people. Professional organisations don't want that. So you should consider professional ranks more of a lifetime achievement indicator. They are not a rating system. Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] UEC cup
Hi Mark, I'm not claiming to be an authority on the matter, but I beg to differ. Name me an EGF 7-dan that's not professional level. And then explain how come they are listed among players that are anywhere from 1p to 5p in different Asian countries. I used to be an EGF 6-dan and have beaten top 9p players with 3 stones on occasion. For a while I had a Japanese 2p teacher but stopped taking lessons when I started to beat him on black pretty consistently. That was when I was still 5-dan. So I don't think it's so far off to say 7-dan amateur is pro level. this is actually a rather complicated topic because you can have different definitions for professional strength. For instance, I could make an argument that S. Shikshina 3p does not have professional strength, AFAIK she did not become a professional in the regular way and has never won a professional tournament game. So, if you define professional strength as someone who could become a 1p in Korea, China or Japan today, I think most European 7dans would fail. (Dragos Bajenaru, while only calling himself 6dan, has a rating higher than some 7dans and failed to become a professional in Japan in the past.) If you define professional strength as the lowest strength of anyone who currently holds a professional rank, then most European 7dans qualify, yes. I've heard 2nd-hand reports of Noguchi Motoki losing a 4- or 5-handicap game against an active professional player. The collective record of European 7dans against professionals at the WMSG was 0 wins, 6 losses. Regrads, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] UEC cup
Mark Boon wrote: All the examples given to support the argument either way are at best anecdotal. But looking at the EGF ranking list, the 7-dan players are interspersed with players of professional ranks, with very few 6 dans among them. That is based on a considerable amount of data. Maybe you are correct to have doubts about S. Shikshina, but how about Guo Juan, Catalin Taranu, Alexander Dinnerschtein and others? These things tend to average out over larger numbers. Let's say that active Pros should have 2800+, though players with 2750+ might still be professional strength. That would give 2 or 3 EGF 7dans of professional strength, which doesn't contradict anything I said earlier. About the pros in the rating list, since you asked: Dinerstein is at the bottom end of professional strength, note that his promotion from 1p to 3p (like Shikshinas) had nothing to do with his playing achievements. Catalin was over 2800 during his time as an active pro (peaking at 2821 in 2004). He has obviously gotten weaker since he stopped playing pro tournaments, just like Guo, who has been out of the pro scene for so long that I think it's fair to say she doesn't have pro strength anymore. Finally, Diana has only been rated for 3 tournaments since she became a pro - there was a gap of 2 years where she was studying Baduk in Korea - and one of them was the WMSG which obviously has questionable effects on the rating because of so many participants who weren't in the rating database beforehand (for instance she lost against a girl from Taiwan who will have entered the database at 2600 but for all I know will be a professional soon). I think it's fair to say that her current rating probably does not reflect her current playing strength. I don't quite see the large numbers over which this is averaging out. ;) Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] UEC cup
dave.de...@planet.nl wrote: Also, a 4p is not a 7p. The difference should be about one stone. 4p is equivalent to 8d EGF. I wish people would stop spreading such incorrect information. The correlation between professional ranks and playing strength is quite bad, and EGF 7dans are not, generally speaking, professional strength. Also, please note that some professional associations have different rules for male and female players. If you find a Japanese 7p who can give a Korean 1p 2 stones and win, I will eat my hat... Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Re: WMSG - Scoring
Ingo Althöfer wrote: Michael Goetze wrote: I doubt that this rule has a significant effect on playing strength, either of computers or humans. After all, the average effect is about half a point per game, which you probably won't notice below the level of amateur 6d or 7d. You are right, and I did not state clearly what I had in mind: play on small boards (like 9x9 or even 7x7). Especially, there Seki situations are (much) more frequent than on larger boards. I think this effect is often overstated. None of the lines believed to be optimal play on 7x7 ends in a seki. http://senseis.xmp.net/?7x7BestPlay http://homepages.cwi.nl/~tromp/java/go/7x7.sgf Concerning the next Computer Olympiad and having in mind the discussion on the last one (how fair is 7.5 komi for 9x9 computer games?) the WMSG scoring should be worth to be discussed for 9x9. I wasn't subscribed to the list at the time, but I've read that discussion, and must say I consider many of the things said there to be silly. If you have a problem because you're using komi of 7.5, and the correct komi is probably 7, then changing the komi to 6.5 won't solve that problem at all! Similarily, there will be a komi which is correct for 9x9 in a game theoretical sense for WMSG rules, and it might be 7, or 6, or 8, but it certainly won't be 6.5 or 7.5. So, please, don't try to alleviate the symptoms of this problem. Instead, solve it at the root: change the komi to 7, and come up with a tournament system which can deal with jigos. Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Re: WMSG - Scoring
Ingo Althöfer wrote: Ok, that is a technical answer. But ... ... what does the rule change mean for strengths of programs - especially in play against (strong) humans? Would this rule help the computers or the humans? I doubt that this rule has a significant effect on playing strength, either of computers or humans. After all, the average effect is about half a point per game, which you probably won't notice below the level of amateur 6d or 7d. And since there are currently no programs playing at that level, it's impossible to say how they would be affected... Positions where the WMSG rules (as a whole, not just this simple element) differ more significantly from normal rules, such as seki with one-sided dame, are very rare. So I think this is pretty much a non-topic... Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Re: WMSG - Scoring
David Fotland wrote: AGA rules also have the effect of changing the komi depending on which side makes the last pass. No, they don't. AGA rules are area-scoring rules and the komi is fixed. (They also provide a method to determine the area-scoring result via territory counting, but that's irrelevant.) WMSG rules are different. Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] Re: WMSG - Scoring
David Fotland wrote: Semantics. If white passes first she has to give one more prisoner to black than if black passes first. This changes the score by one point relative to Japanese rules, which has the same effect as changing the komi by a point. Of course I'm aware that the komi is not actually changed. So you would also say that Chinese and New Zealand rules change the komi depending on who passes first? I think that's pretty silly... (Note that playing with AGA rules does not require the use of pass stones. It is perfectly valid to play with AGA rules and not use any pass stones.) WMSG rules, however, change the komi relative to other area-scoring rules. That's a fundamental difference. Regards, Michael ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/