the puzzle is lichess.org/training/4iZT8
If you are puzzled by this puzzle, turn on SF16. It will explain all you need to know.
It's one of the puzzles I call "anticlimactic". You seem to be aiming for a checkmate or the queen... and end up winning a bishop (and a pawn). Makes me always feel like "Wait... that was it? Hm..."
Oh this is a tricky one. Many different motives playing together.
The main points to notice are that there are two pins going on. As the white queen on d4 is unprotected the bishop on d6 is pinned by the black queen. Additionally, the pawn on e5 is also pinned by the bishop on g7, so it doesn't actually protect the bishop d6. There also is a rook on the same diagonal on a1. Also, we notice, that the white queen is on the same black-squared diagonal with the king. A skewer or fork isn't possible right now but might become an idea later on. In conjunction with that, our light-squared bishop and whites piece placement force the white king to stay on the black-squared g1-a7 diagonal.
Having realized that the pawn e5 does nothing to protect the bishop d6 for as long as it is pinned to the queen d4, the first idea is to attack the pinned bishop with an additional piece. The only option for that without dropping the pin on the e5 pawn is 1...Re6. However, white easily defends by either Bc5 defending the queen or Qc4/Qc5 stepping out of the pin on the e-pawn.
Then I found it hard to come up with something else, but finally thought of the sacrifice 1...Rxe5. With the white king somewhat exposed 2.fxe5 might be a bad idea due to 2...Qg5+. That forces 3. Kf2 and already Dd2+ and swinging the queen back and forth guarantees us at least a draw. As a draw is already a significant improvement from being down an exchange and a pawn, there is no need to look farther into it before checking whites other alternatives on 1...Rxe5.
The other obvious response to 1...Rxe5 is capturing with the bishop 2.Bxe5. That is the sequence that playes out in the puzzle. The only move that makes sense is 2...Qxd4+ (with everything else we can be sure to have sacrificed our rook for nothing) and from there its all forcing moves: 3. Bxd4 Bxd4 4. Rf2 Bxa1. In the end black regains the rook and wins bishop and pawn in the process.
Other options for white might be 2.Qf2, letting black have the pawn and threatening to take the rook, 2.Qxe5 playing two rooks against queen or 2.Bc5.
So why does Stockfish consider 2.Bxe5 best? Lets go through the other options in depth:
2. Bc5 is not really an option. If you try 2...Rd5 with the discovered attack on the queen, you immedeatly see that white is in trouble. Move the queen away then black can take the bishop an c5 with 3...Rxc5. If white recaptures with 4.Qxc5 we get our fork on king and queen with 4...Bd4+ winning the queen. If white doesn't recapture (e.g. after 3.Qf2) then they are still in trouble because Bd4 (and Qd4 and Qd5) threaten to skewer the queen exploit the weak king position. An example: 1...Rxe5 2.Bc5 Rd4 3.Qf2 Rxc5 4.Qe1 and know 4...Qd4+ 5.Rf2 Qxa1 wins the rook similar to how it went in the puzzle - 4...Qd5 threatening mate might be even stringer. Other fourth moves for white end much the same.
With 2. Qf2 white will run into similar problems. Black can let the rook hang and take the bishop 2...Qxe6. If white doesn't want to be down in material on top they have to take the rook 3. fxe5 Bxe5. Again black threatens the skewer Bd4 and also Qd5 with the threat of Qh1 mate is nasty. White can try many ideas to defend but all of them lose material.
2. Qxe5 Bxe5 3. Bxe5 keeps the material balanced for the moment, but we know now that 3...Qd5 is very dangerous for white. White has to prevent mate so 4.Kf2 Qf3+ 5.Ke1 Qe3+ 6.Kd1 and now black wins the exchange with 6...Bf3+ forcing 7. Rxf3 Qxf3+. Even stronger is 6...Bg2 winning the whole rook.
As the last option that only leaves 2.fxe5. Here Stockfish can actually be helpful for once, telling us it a forced mate. After 2...Qg5+ 3.Kf2 Qd2+ 4.Kg3 black doesn't need to swing back and can instead give 4...Qg2+ 5.Kh4 Qh2+ 6.Kg5 h6#.
So that is why Stockfish chooses 2.Bxe5 for white. It wants to fight for a draw with a rook against the bishop pair, because everything else turns out to be even worse.
@Felleagle Stockfish does not tell you all you need to know. It only tells you what the best move is (most probably) but it can't tell you *why* it is good. You have to figure that out by yourself.
The main points to notice are that there are two pins going on. As the white queen on d4 is unprotected the bishop on d6 is pinned by the black queen. Additionally, the pawn on e5 is also pinned by the bishop on g7, so it doesn't actually protect the bishop d6. There also is a rook on the same diagonal on a1. Also, we notice, that the white queen is on the same black-squared diagonal with the king. A skewer or fork isn't possible right now but might become an idea later on. In conjunction with that, our light-squared bishop and whites piece placement force the white king to stay on the black-squared g1-a7 diagonal.
Having realized that the pawn e5 does nothing to protect the bishop d6 for as long as it is pinned to the queen d4, the first idea is to attack the pinned bishop with an additional piece. The only option for that without dropping the pin on the e5 pawn is 1...Re6. However, white easily defends by either Bc5 defending the queen or Qc4/Qc5 stepping out of the pin on the e-pawn.
Then I found it hard to come up with something else, but finally thought of the sacrifice 1...Rxe5. With the white king somewhat exposed 2.fxe5 might be a bad idea due to 2...Qg5+. That forces 3. Kf2 and already Dd2+ and swinging the queen back and forth guarantees us at least a draw. As a draw is already a significant improvement from being down an exchange and a pawn, there is no need to look farther into it before checking whites other alternatives on 1...Rxe5.
The other obvious response to 1...Rxe5 is capturing with the bishop 2.Bxe5. That is the sequence that playes out in the puzzle. The only move that makes sense is 2...Qxd4+ (with everything else we can be sure to have sacrificed our rook for nothing) and from there its all forcing moves: 3. Bxd4 Bxd4 4. Rf2 Bxa1. In the end black regains the rook and wins bishop and pawn in the process.
Other options for white might be 2.Qf2, letting black have the pawn and threatening to take the rook, 2.Qxe5 playing two rooks against queen or 2.Bc5.
So why does Stockfish consider 2.Bxe5 best? Lets go through the other options in depth:
2. Bc5 is not really an option. If you try 2...Rd5 with the discovered attack on the queen, you immedeatly see that white is in trouble. Move the queen away then black can take the bishop an c5 with 3...Rxc5. If white recaptures with 4.Qxc5 we get our fork on king and queen with 4...Bd4+ winning the queen. If white doesn't recapture (e.g. after 3.Qf2) then they are still in trouble because Bd4 (and Qd4 and Qd5) threaten to skewer the queen exploit the weak king position. An example: 1...Rxe5 2.Bc5 Rd4 3.Qf2 Rxc5 4.Qe1 and know 4...Qd4+ 5.Rf2 Qxa1 wins the rook similar to how it went in the puzzle - 4...Qd5 threatening mate might be even stringer. Other fourth moves for white end much the same.
With 2. Qf2 white will run into similar problems. Black can let the rook hang and take the bishop 2...Qxe6. If white doesn't want to be down in material on top they have to take the rook 3. fxe5 Bxe5. Again black threatens the skewer Bd4 and also Qd5 with the threat of Qh1 mate is nasty. White can try many ideas to defend but all of them lose material.
2. Qxe5 Bxe5 3. Bxe5 keeps the material balanced for the moment, but we know now that 3...Qd5 is very dangerous for white. White has to prevent mate so 4.Kf2 Qf3+ 5.Ke1 Qe3+ 6.Kd1 and now black wins the exchange with 6...Bf3+ forcing 7. Rxf3 Qxf3+. Even stronger is 6...Bg2 winning the whole rook.
As the last option that only leaves 2.fxe5. Here Stockfish can actually be helpful for once, telling us it a forced mate. After 2...Qg5+ 3.Kf2 Qd2+ 4.Kg3 black doesn't need to swing back and can instead give 4...Qg2+ 5.Kh4 Qh2+ 6.Kg5 h6#.
So that is why Stockfish chooses 2.Bxe5 for white. It wants to fight for a draw with a rook against the bishop pair, because everything else turns out to be even worse.
@Felleagle Stockfish does not tell you all you need to know. It only tells you what the best move is (most probably) but it can't tell you *why* it is good. You have to figure that out by yourself.
It boils down to king safety. The king has nowhere to hide. I used the Fen code and added a pawn in front of the king to show that nothing would have happened the way it did, if it was not for that missing g2 pawn. With less pawns on the chessboard, pieces get exposed.
lichess.org/analysis/3qr1k1/1p3pbp/2bB2p1/p3Pp2/P2Q1P2/1P1P4/2P3PP/R4RK1_b_-_-_0_25?color=black
lichess.org/analysis/3qr1k1/1p3pbp/2bB2p1/p3Pp2/P2Q1P2/1P1P4/2P3PP/R4RK1_b_-_-_0_25?color=black
You need to explain what it is about it that you don't understand.
@mkubecek said in #3:
> It's one of the puzzles I call "anticlimactic". You seem to be aiming for a checkmate or the queen... and end up winning a bishop (and a pawn). Makes me always feel like "Wait... that was it? Hm..."
If you call this one "anticlimactic", imagine the same but losing material in the process...
> It's one of the puzzles I call "anticlimactic". You seem to be aiming for a checkmate or the queen... and end up winning a bishop (and a pawn). Makes me always feel like "Wait... that was it? Hm..."
If you call this one "anticlimactic", imagine the same but losing material in the process...
@Lukasel said in #4:
> Oh this is a tricky one. Many different motives playing together.
Thanks for such a great and detailed explanation!
> Oh this is a tricky one. Many different motives playing together.
Thanks for such a great and detailed explanation!
@irridente said in #1:
> the puzzle is lichess.org/training/4iZT8
As soon as you realise that fxe5 results in #, the solution should be clear, since we can (relatively) easily calculate that Bxe5 and Qxe5 and not recapturing are all bad. Hard puzzle though.
> the puzzle is lichess.org/training/4iZT8
As soon as you realise that fxe5 results in #, the solution should be clear, since we can (relatively) easily calculate that Bxe5 and Qxe5 and not recapturing are all bad. Hard puzzle though.
This topic has been archived and can no longer be replied to.