# Why didn't the Black rook capture the White bishop?

I am a total newbie in chess. I was playing with an ~intermidiate computer~ as White and made a pretty stupid mistake by playing Bf4. I waved the bishop goodbye, yet Black's rook did not capture it. Was there any good reason for the Black not to? The white pawns are moving downwards here. As far as I know, the pawn cannot move back to capture, so I reckon the bishop to have been 100% unprotected.

``````[FEN "8/4R3/8/4P1B1/6k1/PBP5/1P3r2/1KN5 w - - 0 1"]

1. Bf4 Rf3
``````
• In which direction are the pawns moving? If the position is from White's perspective (i.e. the pawns move "up") then the pawn on d4 guards the bishop on c5. Oct 25, 2020 at 21:43
• @Scounged sorry for not saying it directly: the pawn is moving downwards, I meant to embed this in the last sentence. Oct 25, 2020 at 22:15
• @ZhiltsoffIgor please note that a1 is the bottom left of the board from white's position. In this case your bishop did not move to c5, it moved to f4 Oct 26, 2020 at 13:01
• @DarrenH thank you. I am sorry, I am not quite good with this board-viewer yet, is it possible to add coordinates to the picture (like a-h on the top and 1-8 on the left side)? Oct 26, 2020 at 16:06
• Obviously the computer panicked under time pressure and dropped the rook a little too short. Oct 27, 2020 at 1:57

When a "total newbie" achieves such an overwhelming position against a computer, it most probably means that the computer was forced to make a "sub-optimal" move from time to time - to give the player a chance to win the game. The problem is that the computer has no clue what is a "reasonable" mistake from a human point of view.

"Rxf4" might have been the best choice (not avoiding defeat eventually). But the computer decided not to make the best move, since the programmers told him to do so. In this particular case, this resulted in choosing the second(?) best move, which from our human perspective is absolutely incomprehensible.

• Also White's position is so desperate anyway that some computers may not see much difference (depending on how the evaluation function was programmed) Oct 25, 2020 at 23:02
• Naively I would expect the computer to play better when there are fewer pieces, because it's much faster to look many turns ahead in that case.
– Mark
Oct 26, 2020 at 8:21
• @Mark Actually, that is not the case. With more space on the board there are far more possible options available per move. This massively increases the number of moves to check. Depending on how well the program is written, this can be a serious problem if the computer is time-limited. Oct 26, 2020 at 8:36
• @theblitz Unintuitive, that's interesting, thanks.
– Mark
Oct 26, 2020 at 8:50
• @theblitz: It's possible for positions with many pieces but few pawns to be cramped, but I would think in general the dominating factor for evaluation difficulty would be the number of pawns (having fewer pawns would make things harder). A piece may block at most six of the squares another piece could have used, but in most cases would have six or more moves of its own. A pawn, by contrast, would often preclude multiple potential moves by other pieces while often having only one (very seldom more than three, and never more than four) moves of its own. Oct 26, 2020 at 15:58

Because the bishop is protected by the pawn -

``````[fen "5NK/2r3P/5PBP/8/1B1P/1k/3R/8 w - - 0 0"]

1. Bc5 Rxc5 2. dxc5
``````

Not that it makes any difference. Black's position is hopeless.

• Thank you for your answer. Yet, I am a bit lost. You suggest playing dxc5. Are you assuming the White pawns to move upwards (which is false, hence the question; sorry for not saying it in the post directly right away), or is it within the rules to capture with a pawn backwards? Oct 25, 2020 at 22:29
• @ZhiltsoffIgor In your diagram you have a pawn on d4. When you move the bishop to c5 it is protected by the pawn on d4. The pawn on d4 can move to d5 or capture a black piece on either c5 or e5 (if there are any black pieces on those squares). That is the forward direction for the white pawns. You say you are a chess newbie. That would explain your confusion. Oct 25, 2020 at 22:59
• This answer is false, considering the last sentence (and comment) of OP. OP drew the board backwards, white started up top, so this move is not allowed. Oct 26, 2020 at 7:26