Dragon versus Nakamura

[Ozymandias]

I thought Dragon would trash him, but not by so large a margin. Was he tired?

[JJJ]

These results seems incredible ! How about a 3 pawns handicap now with Dragon ?

[Nordlandia]

Nakamura should have been given the white pieces. Just a couple of inaccuries and Dragon may equalize. Human does not have much margin of error to go on against top engines, even with odds.

[Uri Blass]

Nakamura psyched himself out.

I don't think he realized he was playing the MCTS version.

I mentioned we used MCTS mode in the chess.com chat along the side of the game board. He speculated there was a more aggressive setting. The only change was MCTS mode. In both regular and MCTS mode we used a low Time Usage to make Dragon move faster. We even had the same Contempt. We have speculated in the past MCTS mode (although much weaker against engines) might be better against some opponents, since it should prefer to leave them with positions with just a few good moves over positions with many good moves. Regular alpha-beta does not care since if one move is a refutation, the rest are never searched at that point in the tree. We cannot prove this though since it would take a lot of games to do so.

It shouldn't actually be very complicated to build a search function that in each position selects objectively sub-optimal move that, however, leads to positions where all but one or very few moves for opponent are bad. That kind of engine would easily win against best human even with a rook handicap.

I totally disagree about it.

rook is a rook and I think nakamura did not do his best against dragon with 2 pawns handicap.
I do not claim that he lost on purpose but you do not do your best when you talk during the game and explain what you think to other people.

[Cornfed]

Final score : Dragon wins 6.5-1.5 (===+++++)

An excellent result for Dragon.

It shouldn't actually be very complicated to build a search function that in each position selects objectively sub-optimal move that, however, leads to positions where all but one or very few moves for opponent are bad. That kind of engine would easily win against best human even with a rook handicap.

That would be a very interesting and useful function to have for chess analysis (not just so an engine can beat up a human in handicap matches).

I've certainly proposed it. Team K/D seems to indicate K/D in MCTS mode already 'sort of does this'...but an engine that would basically say 'option 2' gives the opponent 75% (or a number) of 'reasonable' looking moves that lead to an even worse continuation, would be ideal. Developers are only interested in elo for best lines though it seems.

Also, it would require a special GUI I think...and 'engine developers' specialize in tweaking parameters and such (or organizing odds exhibitions) not coding GUI's, so don't expect it to ever happen.

[Nordlandia]

Someone proposed Rook for Two Minors in the Twitch chat. This is a considered a huge advantage.

Is this something to consider as future odds configuration ?

[Chessqueen]

Nakamura should have been given the white pieces. Just a couple of inaccuries and Dragon may equalize. Human does not have much margin of error to go on against top engines, even with odds.

I agree with you but only 4 games with White and 4 with Black pieces and if Tablebases only up to 5 men since Non of the super GM can play 6 or 7 men endgames perfectly.

[Paloma]

....
I do not claim that he lost on purpose but you do not do your best when you talk during the game and explain what you think to other people.

I agree completely.

[Cornfed]

Final score : Dragon wins 6.5-1.5 (===+++++)

An excellent result for Dragon.

It shouldn't actually be very complicated to build a search function that in each position selects objectively sub-optimal move that, however, leads to positions where all but one or very few moves for opponent are bad. That kind of engine would easily win against best human even with a rook handicap.

That would be a very interesting and useful function to have for chess analysis (not just so an engine can beat up a human in handicap matches).

I've certainly proposed it. Team K/D seems to indicate K/D in MCTS mode already 'sort of does this'...but an engine that would basically say 'option 2' gives the opponent 75% (or a number) of 'reasonable' looking moves that lead to an even worse continuation, would be ideal. Developers are only interested in elo for best lines though it seems.

Also, it would require a special GUI I think...and 'engine developers' specialize in tweaking parameters and such (or organizing odds exhibitions) not coding GUI's, so don't expect it to ever happen.

Not really the proper thread for this specific idea. But to follow up, I would add that Komodo/Dragon using their MCTS...in conjunction with Chessbase would be ideal since about everyone has Chessbase and us it to work on their game, prepare opening idea, etc. A boy can dream...

[lkaufman]

I thought Dragon would trash him, but not by so large a margin. Was he tired?

I wouldn't say "tired". But he played too quickly, was too distracted reading messages, and seemed to be upset that we had changed something (mcts), even though there was no rule against changing settings and we had made the very same change midway in our two previous matches. But when I reviewed the games, most of the critical decisions (but not all) by Dragon would have also been chosen by normal Dragon, and in one game Dragon MCTS missed a brilliant win that regular Dragon saw, only to win anyway after later errors by Nakamura.