Nakamura vs. Komodo

[Jesse Gersenson]

Game 4
[pgn][Event "Live Chess"]
[Site "Chess.com"]
[Date "2016.01.07"]
[White "Hikaru"]
[Black "KomodoChess"]
[Result "0-1"]
[WhiteElo "2787"]
[BlackElo "3368"]
[TimeControl "45|15"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/3PP3/5N2/PPP2PPP/RNBQKB1R w KQkq - 0 1"]
[Termination "KomodoChess won by resignation"]

1.c4 d6 2.Nc3 g6 3.Qb3 c6 4.Be2 Qb6 5.Qc2 Bg7 6.O-O Nf6 7.h3 O-O 8.Be3 Qc7 9.Rac1 Nbd7 10.Rfd1 b6 11.Qd2 Bb7 12.b3 Rad8 13.Bh6 e5 14.Bxg7 Kxg7 15.d5 a5 16.Qe3 Ra8 17.Ne1 Nc5 18.Nd3 Qe7 19.Nb2 Nfd7 20.Nca4 f5 21.Nxc5 bxc5 22.f3 Nf6 23.g4 Rf7 24.Rf1 Raf8 25.Kh2 fxg4 26.fxg4 h5 27.gxh5 Nxh5 28.Rxf7+ Qxf7 29.Rf1 Nf4 30.Bg4 cxd5 31.exd5 Qc7 32.Qg3 Rf6 33.Rf2 Bc8 34.Nd1 Bxg4 35.Qxg4 Qf7 36.Qh4 Rf5 37.Ne3 Rh5 38.Qg4 Qe7 39.Rf3 Rg5 40.Qc8 Ne2 41.Qe6 Qxe6 42.dxe6 Nf4 43.Rg3 Rh5 44.Nd5 Nxe6 45.Nc3 Nd4 46.Ne4 Nf5 47.Rd3 Rh4 48.Nxd6 Rf4 49.Nxf5+ gxf5 50.Rd5 e4 51.Rxc5 Kf6 52.Rc8 Rf2+ 53.Kg3 Rf3+ 54.Kg2 Ke5 55.h4 Kf4 56.h5 e3 57.h6 Rg3+ 58.Kh2 Rg6 {Nakamura resigns} 0-1
[/pgn]

[Graham Banks]

Well played game by Komodo and a great defensive fight from the GM.
Thanks to all concerned for the match.

[JJJ]

Nakamura played well. So, now we know that draw is best thing to expect at these handicap, will you change them Larry ?

What is the currently elo of komodo based only with his game against human ? Can you determine it ?

[Dan Cooper]

I think depth handicaps would be interesting. Is Komodo at a max depth of 4 good enough to take out a super GM?

[lkaufman]

Nakamura played well. So, now we know that draw is best thing to expect at these handicap, will you change them Larry ?

What is the currently elo of komodo based only with his game against human ? Can you determine it ?

Well, the handicap obviously depends on the opponent. I don't know when or if we'll have another chance to play a 2800 level opponent. I'll try to think of some new ideas for matches with "ordinary" GMs.

I plan to calculate Komodo's performance rating at each handicap so far. The question is how much to add to those numbers for the handicap. Based on both my own tests and those of Kai, we have some idea, but it's not very exact. Anyway, I'll post something on this once I have the time.

[Steve Maughan]

Hi Larry,

Thanks for doing this event - it was so much fun to watch.

- Steve

[duncan]

larry,

that was great. imo an important contribution to computer chess.

thanks

[lkaufman]

I think depth handicaps would be interesting. Is Komodo at a max depth of 4 good enough to take out a super GM?

Komodo at four ply would probably be something like 1800 elo or so, I would guess. Any serious player can search more than that. Somewhere around 10 to 12 ply might be in the (ordinary) GM range (always assuming the human has standard tournament time), but it is probably a poor handicap for several reasons. For one thing, the meaning is not consistent; 10 ply on a 24 core machine is much stronger than 10 ply on a single core. Also, the quality would be very high in the middlegame but poor in the endgame, because humans can rarely search much more than 10 ply accurately in a complicated middlegame, but can often search much more than that in endgames. I think huge time (or hardware) handicaps (without ponder) are better than fixed depth.

[Uri Blass]

Nakamura played well. So, now we know that draw is best thing to expect at these handicap, will you change them Larry ?

What is the currently elo of komodo based only with his game against human ? Can you determine it ?

I do not know that draw is best to expect at these handicap.

Maybe there are humans who are better in winning against komodo from these positions.

[Laskos]

Nakamura played well. So, now we know that draw is best thing to expect at these handicap, will you change them Larry ?

What is the currently elo of komodo based only with his game against human ? Can you determine it ?

Well, the handicap obviously depends on the opponent. I don't know when or if we'll have another chance to play a 2800 level opponent. I'll try to think of some new ideas for matches with "ordinary" GMs.

I plan to calculate Komodo's performance rating at each handicap so far. The question is how much to add to those numbers for the handicap. Based on both my own tests and those of Kai, we have some idea, but it's not very exact. Anyway, I'll post something on this once I have the time.

Thanks Larry and Mark for this match. By now you are able pick the right handicaps in order to have balanced matches, as the last matches had shown. People are getting pretty good too at predicting the result, 2.5/4 for Komodo was the central value in the poll answers.

Naka played very well, congratulations.

The overall performance of Komodo versus Naka is probably somewhere around 3200 ELO. Also, this performance is much more stable than computer ratings. An improvement of 200 ELO points of Komodo on CCRL will be reflected only in some minor increase of the balanced handicap against top human. This "compression" of improvement comes from the dilation of the handicap with the strength, engine improves not much faster than the same handicap increases too in its ELO value with the strength improvement of the engine. Probably Naka can draw at f7 + c7 odds even against the perfect engine, because there are only 1000-1200 more ELO points to Komodo, and the value of the handicap against top human increases too to huge value at this level of engine strength.

Other ways of giving odds seem less spectacular. Fixed depth is meaningless, a 1992 Mephisto is probably not far away from Komodo at fixed depth (in fact I don't know which is stronger). Time odds: not intuitive for most chess players, not clear in its magnitude even to many CC specialists. That Russian GM commentary on twitch was very interesting, at one time he asked something like: "do computers really need time to think? I don't see any difference in the level of playing whether it thinks for 1 second or 1 minute". Time odds would surely not impress him. At 1 second/move Komodo is roughly 2900 or so in human ratings, at 1 minute/move maybe 3100, both beyond the usual human play. Maybe "taking back" is something new to consider, as humans, even top ones, are particularly vulnerable to some "stupid" blunders, which ruin almost all good intentions.

Thanks again for this entertainment!

PS What was the value of the contempt in these games?