Komodo vrs Human being. Bullet time control. 2016-Only One

[yurikvelo]

Your approach is different from ARB (actually ARB has no approach, it's pure fake, mistification, he compose games without engine participation).

1) You give engine starting position with 16-pawn fortress already built on the board (starting FEN is somewhere around 20th move).

2) Play thousand of games in this position, use learning of lost games

3) Finally draw 1 game, post it's PGN online

Profit!

[Father]

yurikvelo (Yuri Pylypenko):

As I told, I cannot win one game in a tactic combat, against super engines.

An engines every day are better, and better playing chess, and faster and faster. In 2003, I could search score of 50% or more. Nowadays, my goals are 10%, more or less. I won games in 2003 in 60 minutes plus 15 per move, running the super engine to 60.000 Kilo nodes per second. Who? To close the paw structure.

Or in one match, where only will be playing one game, the first one and unique, if I win the game, then the score will be 100%, and man versus machine will be winning for human being. But this is fake too. Why)? Because engines could be winning, of course in a longer match.

Then the target is other: To prove, that human being could be winning or drawing chess computers for ever. But only some games. And to prove that a Master or an amateur can win or draw engines, the best ones computers, why not, much better that Deep Blue was. And this goal is nice.

With best respect,


Pablo