Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / GM Ehlvest - Rybka. Game 4
- - By Victor Zakharov (*****) [ru] Date 2007-07-07 16:20
[Event ""]
[Site ""]
[Date "2007.7.7"]
[Round ""]
[White "*Ehlvest"]
[Black "*Rybka"]
[Result "0-1"]
[Eco ""]
[Annotator ""]
[Source ""]

1.e4 d5 2.exd5 Qxd5 3.Nc3 Qd6 4.d4 Nf6 5.Bg5 h6 6.Bh4 Qb6 7.Rb1
Na6 8.Bc4 Bf5 9.Nf3 Nb4 10.Bb3 O-O-O 11.O-O e6 12.Qe2 g5 13.Bg3
Nc6 14.Nb5 a6 15.Bxc7 Qxb5 16.c4 Qb4 17.Bxd8 Nxd8 18.Rbc1 Bg7
19.Rfd1 Nd7 20.c5 Nc6 21.d5 exd5 22.Rxd5 Be6 23.Rd6 Nxc5 24.Ne5
Bxe5 25.Rxc6+ bxc6 26.Qxe5 Re8 27.g3 Bd7 0-1
Parent - - By Carl Bicknell (*****) [gb] Date 2007-07-07 17:07
what's the score now? 4-0?
Parent - By Fulcrum2000 (****) [nl] Date 2007-07-07 17:08
3.5-0.5 or better 0.5-3.5
Parent - - By turbojuice1122 (Gold) [us] Date 2007-07-07 17:28
Wow, this is one that Ehlvest could have prepared beforehand if he so chose to go that route, as this is a well-known line.
Parent - - By Hetman (*****) Date 2007-07-07 17:34
Hi,
but the openning character suit more programm then human.  Psychology ;-).
We can congratulate Rybka with the win of the match.!!

My forecast was 4:2 but it seems that GM is tired, so the score could be even larger. :-(.
2 heavy games per day are too much for the human.
3rd and 4th games was easier for Rybka.

Regards
Hetman
Parent - - By turbojuice1122 (Gold) [us] Date 2007-07-07 17:46
I'm sure that if Ehlvest requests it, Larry wouldn't have much of a problem in "bending the rules" and allowing the match to go a little longer so that Ehlvest can get more rest.
Parent - - By lkaufman (*****) Date 2007-07-07 18:10
He is leaving for the Continental Championships the next day, no chance to change the schedule. I don't think he is too tired, he often plays far more grueling schedules than this one.
Parent - By Hetman (*****) Date 2007-07-07 21:24
I do not know the psychic of GM Ehlvest but all games were full of play so he may be a bit exhausted.
Regards
Hetman
Parent - By turbojuice1122 (Gold) [us] Date 2007-07-07 17:47
I just realized--wasn't Game 2 the only game against Hydra in which Michael Adams was able to get a draw?
Parent - - By Uri Blass (*****) [il] Date 2007-07-07 17:59
I do not understand the opening choice of ehlvest.

It could be better for him to stick to one first move and not to change it in every game.
He chose the following moves in his games:
1.c4
1.d4
1.Nf3
1.e4

It is interesting that only with d4 ehlvest got a draw.
It reminds me the following: 
I remember that about 30 years ago when  Ofer Bruk(today IM with fide rating of 2468) was a child and played his first tournament against unrated players he was afraid of his opponents that most of them were older than him so he decided to play 1.d4 as first move and he explained to me his choice by the fact that based on statistics there are more draws with 1.d4 relative to 1.e4

Note that his choice did not prevent him to win that tournament.

Uri
Parent - By George Tsavdaris (****) Date 2007-07-07 18:07
Choosing 1.e4 against a computer is always risky.
What i don't understand like you, is why he changes the opening move all the time?
He should stick with 1.d4 in my opinion....
Parent - - By lkaufman (*****) Date 2007-07-07 18:11
Probably he didn't want to give me a chance to do more work on the book against one specific opening move, or else he simply thought that variety would be more interesting for the fans, and just as good for him.
Parent - By Hetman (*****) Date 2007-07-07 18:26
That is possible that he is playing for the fans. I hope that experience got with Rybka helps him in the nearest tournament.
I remember when I played internal tournament vs my computer and lost to it, then in play against humans I got good result.
Rgds
Hetman
Parent - By Hetman (*****) Date 2007-07-07 18:28
d4 was favourable move of Capablanca. He wrote that it is the best 1st move.
Rgds
Hetman
Parent - - By lkaufman (*****) Date 2007-07-07 18:09 Edited 2007-07-07 18:26
     Of course he could not guess we would play this line. He did play 5 Bg5 here in a game before, but 5...h6 and 6...Qb6 were new. I think White must sac a pawn with 7Qd2 to try for advantage, but it is understandable that Ehlvest was reluctant to do so against Rybka.
     I should also mention here that this 3...Qd6 line was suggested to me for this match by Jeroen Noomen. He made several suggestions against 1e4, but this was the one I liked best and I included it in the book.
Parent - - By Dom Leste (**) Date 2007-07-07 20:10 Edited 2007-07-07 20:22
I guess if rybka gets a score of 5.0 or higher, then time to downgrade it to single or 2 threads and 128-256mb hash limit ;)

Then if things even up just increase processoring power/book depth i.e. up to 5, 8, 10 moves :)

Limit technology, dont change the game of chess into taking back moves,  or giving humans computer help! Time differences i can understand.

rybka team should call Intel and ask them for Pentium 4/Intel Centrino then you see GM's rolling in ;)

Dom
Parent - - By lkaufman (*****) Date 2007-07-07 20:31
    Reducing hash size would only cost a few Elo points. Reducing to single thread would cost about a hundred or a bit more, which would make it closer but certainly not even. In principle we aren't too interested in doing things like this that make Rybka play worse, we're trying to show how well she can play. A straight time handicap is a little more interesting I think, but it would probably have to be something like FIDE time limit vs. blitz to be truly equal (without the other handicaps), and this has the same objection of just weakening Rybka.
     So probably future matches will have handicaps that deal with the conditions of the game itself. This mostly means material (pawn or perhaps Exchange) handicaps, though it could also include multimove, castling, or draw handicaps.
Parent - - By Dom Leste (**) Date 2007-07-07 20:39
I still think time compromise would be more appealing to the majority of chess public :)
Parent - By SR (****) [gb] Date 2007-07-07 20:58
Time handicaps are in my view complete nonsense. Since the search in chess to a large extend can be parallelized, one can also always give Rybka 8 times less, if one compensate by speeding up the search, by for example letting the future version of Rybka use 8 times as many processors. Well, I suppose in practice its a bit more complicated since searching in parallel on 8 machines can never be quite as efficient as giving one machine 8 times as much time. Anyway instead of taking of time handicaps, it is really hardware handicaps that matters. But I agree we want to see Rybka play for full strength and glory.

Also in real life the time handicap (without changing hardware) might in fact make it more difficult for the human, who have less time between his/her own moves.  
Up Topic Rybka Support & Discussion / Rybka Discussion / GM Ehlvest - Rybka. Game 4

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill