Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Game 2. Dzindzichashvili - Rybka 1-0
- - By Victor Zakharov (*****) [ru] Date 2008-03-04 00:43
[Event ""]
[Site ""]
[Date "2008.3.4"]
[Round ""]
[White "Dzindzichashvili"]
[Black "Rybka"]
[Result "1-0"]
[Eco ""]
[Annotator ""]
[Source ""]

1.Nc3 Nc6 2.Ne4 Nb8 3.Nc5 Nc6 4.Nxb7 Nb8 5.Nc5 Nc6 6.Na4 Nb8
7.Nc3 Nc6 8.Nb1 Nb8 {(no b7 pawn)} 9.Nf3 Nf6 10.g3 Nc6 11.Bg2
Rb8 12.d4 e6 13.O-O d5 14.b3 Be7 15.Ba3 Bxa3 16.Nxa3 Qd6 17.Nb1
O-O 18.Nc3 h6 19.Re1 Rd8 20.e3 Ba6 21.Na4 Nd7 22.Qd2 Nf6 23.Qc1
Nd7 24.Qd2 Nf6 25.Qc1 Nd7 26.Rd1 Ne7 27.Nd2 Be2 28.Re1 Bb5 29.c4
Bxa4 30.bxa4 Qb4 31.cxd5 Nxd5 32.Bxd5 exd5 33.Nb3 Rb7 34.a5 c5
35.Nxc5 Nxc5 36.Qxc5 Qxc5 37.dxc5 Rc7 38.Rec1 Rdc8 39.Rab1 Rxc5
40.Rxc5 Rxc5 41.Rb8+ Kh7 42.Rb7 Rxa5 43.Rxf7 Rxa2 44.Rd7 a5 45.Rxd5
a4 46.Ra5 g5 47.g4 a3 48.Kg2 Kg7 49.Ra6 Kf7 50.Rxh6 Rb2 51.Ra6
a2 52.Kg3 Re2 53.f3 Ke7 54.e4 Kd8 55.h4 gxh4+ 56.Kxh4 Kc8 57.Kg5
Kb7 58.Ra3 Kb6 59.f4 Kc5 60.e5 Kb4 61.Ra8 Kb3 62.Kf5 Rc2 63.g5
Rc8 64.Ra6 Rf8+ 65.Kg4 Re8 66.g6 Re6 67.Rxa2 Rxg6+ 68.Kf5 Rb6
69.Rd2 Rb5 70.Kf6 Rb6+ 71.e6 Kc4 72.f5 Kc3 73.Rd8 Ra6 74.Kf7
Ra5 75.f6 1-0
Parent - - By Debaser (***) Date 2008-03-04 00:47 Edited 2008-03-04 00:51
Is it possible to post the game with rybka evaluations? When did it start to drop,  44...a5, 49...Kf7?
Parent - - By Roland Rösler (****) [de] Date 2008-03-04 01:02
Rybka was lost after 45. ...a5. What we see from Rybka was trial and error with an 30 ply search. But this isn´t enough in rook endgames like this; may be good enough for other engines but not good enough for skilled humans like Dzindzi.
Parent - - By lkaufman (*****) Date 2008-03-04 01:41
Roman said that 43...Rxa2 was a serious mistake. Instead pitching the "d" pawn first by d4 offered much better drawing chances, he said.
Parent - By Roland Rösler (****) [de] Date 2008-03-04 01:49
Yes, Roman is right! But I think, he wasn´t so astonished by this Rybka move :-).
Parent - - By lkaufman (*****) Date 2008-03-04 01:39
The move 49...Kf7? is very suspicious, even if Rybka was already losing. Rybka does not make this move with contempt turned off. I will investigate.
Parent - - By lkaufman (*****) Date 2008-03-04 06:02
Despite my earlier comment, this turned out not to be so. Apparently the choice here just depends on the time taken, not much on the contempt setting. The basic problem is that Rybka is not materialistic enough in the endgame, the opposite of what everyone used to say about computers! But as usual, fixing this gives worse test results.
Parent - - By Ty Nance (**) [us] Date 2008-03-04 06:21
I'm really glad that these games are being played before the release of Rybka 3. Because of the handicap, Rybka gets into tough endgames rather than winning in the middlegame, and this gives Larry and Vas a chance to identify and correct endgame shortcomings.

Larry, do you feel that these games against Dzindzichashvili provide relevant positions that you'll use to improve Rybka before the release? And when can we get a timeline on that release?

ty
Parent - - By BB (****) [gb] Date 2008-03-04 11:25

> this gives Larry and Vas a chance to identify and correct endgame shortcomings


I don't know whether I'm being pedantic or simply pessimistic, but I'd attach the modifier hopefully to the latter verb.
Parent - By Permanent Brain (*****) Date 2008-03-04 12:15
The word 'chance' expresses that anyway :-)

I noticed that not only Rybka, but also many other engines 'like' Kf7 because up to a certain depth, they don't consider the simple reply Rxh6 as being good. Obviously, this has to do with big bonuses Black's a-passer gets for advancing to the 2nd (7th) rank. I guess this whole thing is a problem of balancing passer bonuses.

Rybka 232a requires depth 20 to get rid of Kf7.
Parent - By Roland Rösler (****) [de] Date 2008-03-04 15:27 Edited 2008-03-04 16:19
I´m pessemistic too. What we have seen from Rybka in game 2 is an old shortcoming; see point 1 in this thread http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=6010;
Parent - - By lkaufman (*****) Date 2008-03-04 15:28
Yes, the ...Kf7 move is one such position to correct. The release won't be quite as soon as we thought, not because of this match but just because certain changes are taking longer than expected.
Parent - - By Roland Rösler (****) [de] Date 2008-03-04 15:30
This sounds summer -ish :-(.
Parent - - By Banned for Life (Gold) Date 2008-03-04 17:15
I have July 1st in the poll but maybe I am too optimistic! :-)

Alan
Parent - By Roland Rösler (****) [de] Date 2008-03-04 17:34
Maybe Rybka 3.0 is Vas birthday present for me (July 18th) :-).
Parent - - By Permanent Brain (*****) Date 2008-03-04 15:56
Maybe you should release as soon as you are absoultely sure that 3.0 is not worse than 2.3.2a, in any aspect (while at the same time I think you will be sure that it is better, in some aspects).

P.S. I found that 2.3.2a will avoid Kf7 after depth 20 (which IMO is too late, although it doesn't need very much time on a reasonable comp).
Parent - By lkaufman (*****) Date 2008-03-04 16:04
Every change helps in some positions and hurts in others. We only keep changes that help on balance, but there will always be some situations where the newer, better program makes a worse move than the older one. Anyway the timing of release is based on completing certain work and then testing; we are long past the point of wondering whether 3.0 will be better than 2.3.2a, that is not in doubt.
Parent - By Debaser (***) Date 2008-03-04 08:53
Thanks for your comments.

And 44....a5? Some rybka moves seemed human-like to me, she was eager to queen
Parent - By Roland Rösler (****) [de] Date 2008-03-04 00:54
I´m so glad that we all have seen this rook endgame from Rybka. The same bullshit like always, when she has free edge pawn and rook ahead. Pushing the pawn cramps the rook!
Parent - - By tomski1981 (*****) [ca] Date 2008-03-04 01:54
i don't know how to ask without being stupid, and so maybe i shouldn't ask: but are the sponsors/rybka team nervous that he may have a shot at winning (was it $11k?)... or does that not affect anyone or the game...
Parent - - By lkaufman (*****) Date 2008-03-04 03:54
The 11k prize was for the match with 3 move book+White+double time handicap. For these pawn handicap matches the prize for winning is modest as a match victory at this handicap is not any big surprise.
Parent - - By Jim Walker (***) [us] Date 2008-03-04 13:13
Do you have to remove the pawns in strict order....a7,b7,c7 etc.. ?  Seems like knowing which pawn is missing next will help in preparation.
Just a thought.
Parent - By lkaufman (*****) Date 2008-03-04 15:30
We could specify any order, but it would not be fair for me to know the order and for the GM not to know it.
Up Topic Rybka Support & Discussion / Rybka Discussion / Game 2. Dzindzichashvili - Rybka 1-0

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill