It seems to have evaluated that the threat of having a passed pawn outweighs winning the exchange. Other than that I can't see why it would decide to do that.
So it seems like it's either a bishop plus creating a passed pawn, versus trading a rook for a bishop and a pawn. 23. Bxc5 also cuts off castling in both directions (if ... 0-0-0, then white can capture the a-pawn), with 23. Bxg7 black loses the ability to castle kingside. And after capturing the pawn on c5, I'd guess really good players would be able to make that pawn advantage on the left flank count.
Just speculating, I'm not qualified enough to even pretend I can answer the question.
Just speculating, I'm not qualified enough to even pretend I can answer the question.
Looking at the computer's evaluations, there's not much to choose between taking the Rook and taking the Bishop. The reason is basically 2 factors:
1) your bishop pair becomes very strong in the endgame, almost compensating for the exchange by itself
2) 23. Bxg7 is going to make Black's pawns look very silly on both sides of the board - by eliminating the dark-squared bishop, at least Black keeps the kingside intact.
1) your bishop pair becomes very strong in the endgame, almost compensating for the exchange by itself
2) 23. Bxg7 is going to make Black's pawns look very silly on both sides of the board - by eliminating the dark-squared bishop, at least Black keeps the kingside intact.
It's a bug man. Bad evaluation I think.
Not only you take the rook, but also the pawn on a2 after that, so I don't know what was the computer thinking but it wasn't correct. Maybe the engine did not have enough time to evaluate the position. I was trying to justify c5 but I couldn't honestly.
Not only you take the rook, but also the pawn on a2 after that, so I don't know what was the computer thinking but it wasn't correct. Maybe the engine did not have enough time to evaluate the position. I was trying to justify c5 but I couldn't honestly.
#5 Not sure I agree with you. A line could go something like 22...Bxb1 23.Bxg7 Rg8 24. Bc3 Bxa2 25. Bd3 and what does Black play? White is just going to win the exchange back with his Bishop pair, and most likely finish up in an ending with B+R vs B+R with good chances to pick up pawns on the Q-side.
This is superb positional judgment on the part of the computer. White's bishop pair is very important on the open board and it is his source of counterplay. After Bxb1 and Bxa2, Black's light squared bishop is ridiculously placed on a2 and Black's kingside is exposed to White's deadly bishop pair. Perhaps the position is objectively equal, but I would rather have white.
A sample line could go 22..Bxb1 23. Bxg7 Rg8 24. Be3 Bxa2 25. Bd3 where 25..Bxb3 only opens a new file for White's rook. Black's rook on g8, on the other hand, is very close to being trapped by White's two bishops.
On the other hand, 21..c5 is very difficult to understand. The motive is the same (protecting the bishop on d4), however castling is a much better way to do it since it activates the rook on h8 and gives White no chances to trap it. After centralizing his rooks, Black can infiltrate into White's seventh rank and have a comfortable advantage.
A sample line could go 22..Bxb1 23. Bxg7 Rg8 24. Be3 Bxa2 25. Bd3 where 25..Bxb3 only opens a new file for White's rook. Black's rook on g8, on the other hand, is very close to being trapped by White's two bishops.
On the other hand, 21..c5 is very difficult to understand. The motive is the same (protecting the bishop on d4), however castling is a much better way to do it since it activates the rook on h8 and gives White no chances to trap it. After centralizing his rooks, Black can infiltrate into White's seventh rank and have a comfortable advantage.
This topic has been archived and can no longer be replied to.