# Was I right to capture the pawn in this position?

I played a blitz game against a 1900 player which I won. I'm interested in this position, with me as Black to move.

The computer recommended to take the bishop, but I didn't want to take it immediately, because I don't want to give their rook an open file. I thought if they want to take my rook, they better do it with the g pawn.

``````[FEN "r2q2k1/2p2ppp/2p2n2/p7/P2pr3/1BB2P2/1PP3PP/3Q1RK1 b - - 0 1"]
``````

I played Rh4 to keep an eye on the pawn and if they were willing to kick it by playing g3, they'll end up with a weakened position. They replied with g3 as expected. And here I could take the bishop or the h pawn then the bishop, which is what I did.

Had I followed what the computer said, I would have been up 2.5 or 3 points in computer evaluation. What I did got me half a point to 0.8, which is strange, hence I'm asking this question. The computer didn't even bother moving the rook to the h-file.

In the position, what would be the best approach?

• Basically, the problem with Rh4 is not g3, the problem is Bxa5. IMO you have a perfectly winning position, which you should simplify into a win as fast as possible, and dxc3 is more effective than Rh4 at that. The forced endgame (after cxb2) just wins. Jan 10, 2017 at 23:33

N.B. This answer is without the aid of a computer. Here I compare 1...Rh4 to 1...dxc3

## 1...Rh4

The material difference in the given position is exchange + pawn vs. minor piece. This is a difference of 3 pawns. After 1...Rh4, white can win a pawn with 2. Bxa5. On paper this is now a difference of 2 pawns, but computers tend to overvalue pawn structures, so is probably penalizing black for the doubled, isolated c-pawns.

## 1...dxc3

I think that Black is better after this sequence, because:

• Black will soon be up 2 pawns
• Black has better placed pieces
• Black has the long term strategy of making the light square bishop a spectator by placing most of the pawns on dark squares

``````[FEN "r2q2k1/2p2ppp/2p2n2/p7/P2pr3/1BB2P2/1PPQ2PP/5RK1 b - - 0 1"]

1... dxc3 2. Qxd8+ Rxd8 3. fxe4 cxb2 4. e5 Nd5 5. Rb1 Nc3 6. Rxb2 Re8
``````
• +1, good answer imo. 7. Bc4! temporarily defends the e pawn though, and then, white still has Rb7, Ra7, and some sort of counterplay. Black wouldn't want to simplify to a not-winning endgame. Jan 10, 2017 at 23:43

Well, actually it's hard to spot this line without using computer assistance, so your choice in blitz game seems reasonable (except your opponent should have played Bxa5 which gives you a material advantage of pure exchange only).

``````[FEN "r2q2k1/2p2ppp/2p2n2/p7/P2pr3/1BB2P2/1PPQ2PP/5RK1 b - - 0 1"]

1... dxc3 2. Qxd8+ Rxd8 3. fxe4 cxb2 4. e5 Nd5 5. Rb1 Nc3 6. Rxb2 Rd1+! 7. Kf2 Rd4
``````

Now Black threat with Nd1+, so they win an extra tempo to play Kf8 protecting against White's attack on 8th horizontal.

Thus Black keep their main idea to play Rb4 and c5-c4 winning material. The only way for White to save bishop is to play Ba2 and Bb1 which puts White's pieces into absolute zugzwang. So this ending is a total win for Black.

• Even without this tactical nuance, I see nothing wrong with 6...Kf8. If 7.e6, then simply 7...Rb8 followed by c5-c4 wins the bishop. May 7, 2018 at 8:41
• @Evargalo Yes, you're right. White pawns are too weak, so even if White saves the bishop, Black wins the second pawn anyway.
– Matt
May 7, 2018 at 9:47

You reasoning for not taking on c3 is flawed, as the Rook will end up on the b-file after exchanges and cxb2, suffering inactivity to get back the black b-pawn. Bad_Bishop's answer seems accurate to me.

After 6.Rxb2 then Rd1+ 7. Kf2 Rd4 (threat Nd1+) and -+. White is busted.

• downvoting because this looks to me like a comment on bad_bishop's answer. Also after 7. Bc4, white's Rb2 becomes active on the 7th rank and threatening mate on b8. Jan 10, 2017 at 23:44
• My comment was why taking on c3 is best; on not taking on c3 and playing Rh4 which is bad, and agreeing with Bad_Bishop. After 6.Rxb2 then Rd1+ 7. Kf2 Rd4 (threat Rd1+ winning the Rook), and Mr. Bishop is not going anywhere. White has no time to gather any meaningful counterplay. Prove your assessment of getting the rook on the 7th rank and mate threats, which all go away after Kf8, removing the pin and mate threats. White has nothing, and is clearly losing. Saying otherwise does not make it true, especially in chess. Jan 12, 2017 at 0:43
• Indeed, 6. Rd1+ is interesting and might be better than the 6. Re8 that Bad_Bishop suggests. You should include it in your answer if that's the line you advise. I don't mean to argue that white is winning. Jan 12, 2017 at 3:05