I play the Caro-Kann defense for black and every time after the game in computer analysis the move c6 is shown as an inaccuracy, why does this happen in stockfish 15? I didn't notice this before in previous versions of the engine.
Why do you care? Ignore comp evals in the early opening.
Yesterday I played a formerly „flawless“ game...
Yeah
And it recommends playing the French instead
1c6 was an inaccuracy. 1e6 was best (!)
And it recommends playing the French instead
1c6 was an inaccuracy. 1e6 was best (!)
I must agree with Mr. Pushwood.
www.youtube.com/watch?v=nCwi_Pbb9Pk
comps are confused in closed positions.
It's a calculator - no experience, no sense of danger or any other intuition, horizon-effect, no understanding.
Jonathan Schrantz beat Stockfish even in semi-closed positions, normally by sac.ing a piece and stucking a pawn on f6 / f3 (when opponent played 0-0) because comps don't understand long-term advantages either.
I pitted Stockfish (14) against itself in the King's Indian Defense, at first White is almost 'a pawn' up, then it's equals, with both sides sacrificing their dark-sqrd bishop on knight - which is a strategic error.
If you give a position to a computer: one side has a queen, the other side has 6 bishops of the same color, comp thinks the bishops win... until the king and queen go on the other sqrs of the bishops, get closed to the king, and checkmate him.
We need to know how seriously and when, to take computer eval.
According to GM Ben Finegold, 3 world-champions were the best - Karpov, Kasparov, Fischer - dominating the tournaments, The first liked the Caro Kann, the other two liked the King's Indian. so what's now? a computer will tell us otherwise? with no explanation even.
Ruy Lopez - 400 years old, now suddenly it's not good... I Do respect Kramnik as a theoritician, very much, but idk... 400 years of experience don't just go to waste.
www.youtube.com/watch?v=nCwi_Pbb9Pk
comps are confused in closed positions.
It's a calculator - no experience, no sense of danger or any other intuition, horizon-effect, no understanding.
Jonathan Schrantz beat Stockfish even in semi-closed positions, normally by sac.ing a piece and stucking a pawn on f6 / f3 (when opponent played 0-0) because comps don't understand long-term advantages either.
I pitted Stockfish (14) against itself in the King's Indian Defense, at first White is almost 'a pawn' up, then it's equals, with both sides sacrificing their dark-sqrd bishop on knight - which is a strategic error.
If you give a position to a computer: one side has a queen, the other side has 6 bishops of the same color, comp thinks the bishops win... until the king and queen go on the other sqrs of the bishops, get closed to the king, and checkmate him.
We need to know how seriously and when, to take computer eval.
According to GM Ben Finegold, 3 world-champions were the best - Karpov, Kasparov, Fischer - dominating the tournaments, The first liked the Caro Kann, the other two liked the King's Indian. so what's now? a computer will tell us otherwise? with no explanation even.
Ruy Lopez - 400 years old, now suddenly it's not good... I Do respect Kramnik as a theoritician, very much, but idk... 400 years of experience don't just go to waste.
@mammadrezajamshidi said in #4:
> Yeah
> And it recommends playing the French instead
> 1c6 was an inaccuracy. 1e6 was best (!)
Yeah, crazy stuff! I would never ever play the French!
> Yeah
> And it recommends playing the French instead
> 1c6 was an inaccuracy. 1e6 was best (!)
Yeah, crazy stuff! I would never ever play the French!
That would be because it's Stockfish.
If GMs play it, it's fine.
If GMs play it, it's fine.
Maybe SF knows better`?
This topic has been archived and can no longer be replied to.