Chess engine against level 10 chess.com computer

Sort:
Daunted

Hi, I'm a low ranked amateur and fairly new to computer chess.  I was practicing against chess.com computer's lowest levels and I got to think to try something: I have Rybka 2.3.2a on 64bit pc, and I think its rating is like 2900-3000 or somewhere around that.  I played two games using that engine against the level 10 chess.com engine, both times as black, and both times got a draw.
Now I read in another thread that level 10 is like 2600 rating, so if that's true, Rybka should have beat it easily both times. 

I think there are two possibilities: Either Rybka errs on the side of draw, instead of trying to win, when it's playing as black and sees the opponent is a highly rated engine too (even if 300 below its level), or me limiting the engine's time (10 minutes/40 moves) has negatively affected its calculation power (even though the level 10 computer makes each move in less than 10 seconds, which ends up being less time than I've given Rybka).

Any ideas?

llama

Well, both are using your computer's resources. If the chess.com engine thinks during the opponent's time, and if its process is given preference, Rybka may only be getting a few % of your system's resources.

Trying for wins or draws is a human idea. Engines just calculate, apply their evaluation functions, spit out a number, and play the move that led to the best number. It doesn't think in terms of winning or losing.

EscherehcsE

I tried that idea for a few games. Looking at my Windows task manager, it doesn't appear that the Chess.com engine ponders. I can only guess that the Chess.com engine is significantly stronger than 2600, at least on my desktop PC. Maybe it's only 2600 on mobile apps?

llama

Looks like it doesn't ponder, you're right.

Maybe opening book matters? I tried it with stockfish once without opening book and once with. The first game (without book) was a draw. Black played an unambitious setup (although of course engines don't know that) and it was never far from dead equal the whole game.

 

Second game white played a few strategically poor moves in the opening and ended up with no counterplay to black's natural queenside ideas. The position after 13...e6 in particular I think is instructive. White has no pawn break in the center or kingside, and e6 highlights that white's hole on d4 has no counterpart for black. This sort of position that offers only long term defense is something engines don't mind playing, and they give an eval of basically equal. In practical play though I'd rate black's chances as very promising.