Komodo vs. Movsesian

[lkaufman]

This coming Wednesday and Thursday (sept 9 and 10), Komodo will play a six game handicap match against Grandmaster Sergei Movsesian. Movsesian was rated FIDE 2751 in the year 2009, at which time he was World number 10. Currently his FIDE rating is 2658, World number 89, but in rapid chess (which is perhaps more relevant) he is currently World number 30 at 2723. He was born in 1978 so is still relatively young. Komodo will be playing on my 24 core server, and may use a small handicap opening book to provide variety.
The handicap is: four games at the traditional "pawn and move" handicap, the f7 pawn, which is considered the most difficult pawn for Black to give up. The other two games are at Exchange (rook for knight) handicap; Komodo plays White without a1 rook, Movsesian plays without b8 knight and places his queen's rook on b8 instead of a8 (so Black can't castle long, since White can't). Time limit is our usual 45' plus 15" increment. Games start at 9 a.m. Eastern time, 3 pm central European time. Broadcast is planned for chess.com, other sites to be announced.
Seven years ago Rybka tried to give these handicaps to Vadim Milov, then World number 30, but failed to win any games, scoring one draw in two games at f7 handicap and three draws in four games at Exchange handicap. More recently Stockfish scored 1.5 out of two against Nakamura giving him a pawn and the White pieces, but Stockfish conceded the b7 pawn in one game and the h7 pawn in the other, much smaller handicaps than the f7 pawn.
Is Komodo sufficiently stronger than Rybka to tie or win this match? No one knows of course, but it should be a competitive match. Although the two handicaps are considered to be about equally large (both something like 1 1/2 pawn handicaps), I am more worried about the f7 games because the f7 handicap is much more susceptible to preparation than the Exchange handicap. The opening after 1.e4 at f7 handicap is rather unpleasant for Black; he doesn't get mated quickly, but he has to concede a serious deficit in space and development while remaining a pawn down.

[Laskos]

This coming Wednesday and Thursday (sept 9 and 10), Komodo will play a six game handicap match against Grandmaster Sergei Movsesian. Movsesian was rated FIDE 2751 in the year 2009, at which time he was World number 10. Currently his FIDE rating is 2658, World number 89, but in rapid chess (which is perhaps more relevant) he is currently World number 30 at 2723. He was born in 1978 so is still relatively young. Komodo will be playing on my 24 core server, and may use a small handicap opening book to provide variety.
The handicap is: four games at the traditional "pawn and move" handicap, the f7 pawn, which is considered the most difficult pawn for Black to give up. The other two games are at Exchange (rook for knight) handicap; Komodo plays White without a1 rook, Movsesian plays without b8 knight and places his queen's rook on b8 instead of a8 (so Black can't castle long, since White can't). Time limit is our usual 45' plus 15" increment. Games start at 9 a.m. Eastern time, 3 pm central European time. Broadcast is planned for chess.com, other sites to be announced.
Seven years ago Rybka tried to give these handicaps to Vadim Milov, then World number 30, but failed to win any games, scoring one draw in two games at f7 handicap and three draws in four games at Exchange handicap. More recently Stockfish scored 1.5 out of two against Nakamura giving him a pawn and the White pieces, but Stockfish conceded the b7 pawn in one game and the h7 pawn in the other, much smaller handicaps than the f7 pawn.
Is Komodo sufficiently stronger than Rybka to tie or win this match? No one knows of course, but it should be a competitive match. Although the two handicaps are considered to be about equally large (both something like 1 1/2 pawn handicaps), I am more worried about the f7 games because the f7 handicap is much more susceptible to preparation than the Exchange handicap. The opening after 1.e4 at f7 handicap is rather unpleasant for Black; he doesn't get mated quickly, but he has to concede a serious deficit in space and development while remaining a pawn down.

Seems much more challenging than against GM Petr. GM Movsesian is quite a bit stronger, and the handicap significantly larger. Would go for 2.5/6 for Komodo.

[JJJ]

3/6 for Komodo.

[lkaufman]

This coming Wednesday and Thursday (sept 9 and 10), Komodo will play a six game handicap match against Grandmaster Sergei Movsesian. Movsesian was rated FIDE 2751 in the year 2009, at which time he was World number 10. Currently his FIDE rating is 2658, World number 89, but in rapid chess (which is perhaps more relevant) he is currently World number 30 at 2723. He was born in 1978 so is still relatively young. Komodo will be playing on my 24 core server, and may use a small handicap opening book to provide variety.
The handicap is: four games at the traditional "pawn and move" handicap, the f7 pawn, which is considered the most difficult pawn for Black to give up. The other two games are at Exchange (rook for knight) handicap; Komodo plays White without a1 rook, Movsesian plays without b8 knight and places his queen's rook on b8 instead of a8 (so Black can't castle long, since White can't). Time limit is our usual 45' plus 15" increment. Games start at 9 a.m. Eastern time, 3 pm central European time. Broadcast is planned for chess.com, other sites to be announced.
Seven years ago Rybka tried to give these handicaps to Vadim Milov, then World number 30, but failed to win any games, scoring one draw in two games at f7 handicap and three draws in four games at Exchange handicap. More recently Stockfish scored 1.5 out of two against Nakamura giving him a pawn and the White pieces, but Stockfish conceded the b7 pawn in one game and the h7 pawn in the other, much smaller handicaps than the f7 pawn.
Is Komodo sufficiently stronger than Rybka to tie or win this match? No one knows of course, but it should be a competitive match. Although the two handicaps are considered to be about equally large (both something like 1 1/2 pawn handicaps), I am more worried about the f7 games because the f7 handicap is much more susceptible to preparation than the Exchange handicap. The opening after 1.e4 at f7 handicap is rather unpleasant for Black; he doesn't get mated quickly, but he has to concede a serious deficit in space and development while remaining a pawn down.

Seems much more challenging than against GM Petr. GM Movsesian is quite a bit stronger, and the handicap significantly larger. Would go for 2.5/6 for Komodo.

I quite agree with your analysis. Two pawns sounds like more than one, but it seems that a2 plus b2,d2, or e2 is actually a smaller handicap than f7. The first move is very important in these handicap games, and the f pawn is rather critical while the "a" pawn is of low value. We lost rather badly giving f2 and c2 to a 2455 GM, and it seems that this handicap is only slightly larger than f7, while the human is over 200 elo stronger. I think we have improved both the engine itself and contempt slightly since that f2/c2 match, but probably only by ten elo or so. My guess is an even split in the Exchange odds games but minus one in the f7 games, which would give the same score you predict. It is a nice dream of replicating Morphy's offer to give f7 handicap to anyone in the world, but we're probably not there yet.

[JJJ]

I predict tough game for Komodo without the f7 pawn. I don't know if this player is consistent, but I don't think Komodo can win any game at this handicap against 2700 players.

Buy he might with the Rook vs Knight exchange.

[Jesse Gersenson]

Sergei was confident about his chances at f7 odds. I'll be surprised if he loses any of those games.

There will be 15 minuet breaks between games.

The broadcast will be here:
http://u.komodochess.com/games/Live.html

(Note: this is a different url from previous matches)

Pgn's will be archived/replayable here:
http://u.komodochess.com/games/Movsesian.pgn
http://u.komodochess.com/games/Movsesian.html

[Dr.Wael Deeb]

Hi Larry,

Excuse me if it seems that I rain on your parade but why do you keep on organizing these stupid handicapped matches that have nothing to do with the great game of chess !?

Please don't percept this as a a personal attack but you of all people in this forum should know that a normal game of chess is priceless even if Komodo crushes,and he will,the GM with a perfect score....

These handicapped matches are an abomination of the classic game of chess and doesn't prove or state anything....If they do FIDE would apply them among the human players....

Now time odd matches is another story....Give the GM double or triple the time for thinking but let him play normal chess....

Just my two cents regards,
Dr.D

[JJJ]

Hi Larry,

Excuse me if it seems that I rain on your parade but why do you keep on organizing these stupid handicapped matches that have nothing to do with the great game of chess !?

Please don't percept this as a a personal attack but you of all people in this forum should know that a normal game of chess is priceless even if Komodo crushes,and he will,the GM with a perfect score....

These handicapped matches are an abomination of the classic game of chess and doesn't prove or state anything....If they do FIDE would apply them among the human players....

Now time odd matches is another story....Give the GM double or triple the time for thinking but let him play normal chess....

Just my two cents regards,
Dr.D

Thats your judgment, and nothing more. We enjoy this, you don't. Keep your negativity for yourself plz. Better, learn to live without.

[Ozymandias]

Now time odd matches is another story....Give the GM double or triple the time for thinking but let him play normal chess….

That might not be enough. I would go as far as giving the GM 90' + 30", while the engine only has 30" + 0.5". The total game time would be the same, and it'd be quite close (although the engine might still have an advantage, on strong hardware + book).

[Ferdy]

The handicap is: four games at the traditional "pawn and move"

Could you describe what is that?