The trouble with good positions

I think it was Donner who said he hated good (or was it winning ?) positions. I’ve often found my heart rate increasing when I begin to think I am winning. I suppose, if one wants to win, one had better learn to put up with good positions, but I have recently had some reason to dislike them.

Here are three positions I recently had.

You need to activate javascript to enhance chess game and diagram visualization.
4rbk1/pb3p1p/1p3Pp1/r7/2pPP3/2P3Q1/5RBP/1qB1R1K1 w - - 4 28
You need to activate javascript to enhance chess game and diagram visualization.
4rr1k/pp4pp/3bb3/2pp3q/1n2n3/1P2NNP1/PB2PPBP/1QR2RK1 b - - 18 19
You need to activate javascript to enhance chess game and diagram visualization.
3r2k1/1pq3p1/2ppbr1p/8/1PPQ1P2/4P1P1/6BP/R1R3K1 w - - 5 25

I am White to move in the first and last, and Black to move in the second. I think a reasonable prediction of my score, against similarly graded players and without severe time trouble, might be about 2.5. As I only managed one point, I need a story. Here goes.

The first game was played for Dulwich in the London League. One way to mess up a good position is to spend too long trying to win it. I have taken up recording both players’ clock times after each move and this time they told a sorry tale.

You need to activate javascript to enhance chess game and diagram visualization.
1. c4 {75 minutes for 30 plus quickplay, against an opponent I had beaten twice before.} 1... Nf6 2. Nc3 d5 3. cxd5 Nxd5 4. g3 {Holding back the d pawn denies Black his usual Gruenfeld target.} 4... g6 5. Bg2 Nxc3 6. bxc3 Bg7 7. Rb1 Nc6 8. Nf3 (8. h4 {does quite well.}) 8... O-O 9. O-O e5 10. d3 (10. Ne1 {has been played a couple of times.}) 10... Rb8 11. Qc2 (11. Ng5 {perhaps followed by Ne4-c5 is worth considering.}) 11... b6 12. Ba3 Re8 13. Nd2 Bb7 14. Rbe1 Qd7 15. f4 { <span class="PgnWidget-anchor-diagram">[]</span> I have tried similar f4 pushes in a few games in the English, so far without satisfactory results. The computer doesn&#8217;t like it. There is a positional logic to securing the central majority, but it tends to create weaknesses in White&#8217;s position as well.} 15... exf4 16. gxf4 Ne7 17. e4 c5 18. f5 Nc6 ({The computer is happy to eat the pawn with} 18... gxf5) 19. Nc4 Ne5 20. Nxe5 Rxe5 (20... Bxe5 {looks better. The game begins to turn my way.}) 21. Bb2 c4 {I had ideas of c4 and f6. This move cramps my bishop but at too much cost in the centre, I think.} 22. d4 Ra5 23. f6 Bf8 24. Bc1 Qa4 {Bold, but White&#8217;s central and kingside control should outweigh the queenside demonstration.} 25. Qf2 Qxa2 26. Qg3 Re8 27. Rf2 Qb1 { <span class="PgnWidget-anchor-diagram">[]</span> At this point we had the same amount of time, but I now started thinking too long.} 28. e5 ({Played quite quickly, but} 28. Qc7 {is probably better here. I had thirteen minutes left before move 30 so should perhaps have thought more.}) 28... Bxg2 29. Qxg2 Qb5 30. e6 {Ten minutes on this, mainly thinking about the follow-up. I couldn&#8217;t see a clear win but hoped something would appear. But the risk is greater than I realised.} 30... Rxe6 31. Rxe6 {Another long think looking for a win, after which I had nine minutes for the rest of the game.} 31... fxe6 32. f7+ Kg7 33. Qh3 Qh5 {I thought this was forced.} ({The computer prefers} 33... h5 34. Qxe6 Qd5) 34. Qxe6 {Now White has enough for a draw, but not more.} 34... Rf5 35. Bf4 Qg4+ 36. Bg3 (36. Kh1 {seems to be better. Now I had four minutes left and some difficulties over the board. My opponent has twenty minutes left and although both of us play inaccurately from here on my mistakes are bigger.}) 36... Qd1+ (36... Qg5 37. Qxc4 Qe3 {may be better.}) 37. Kg2 Rxf2+ 38. Bxf2 (38. Kxf2 {should draw.}) 38... Qc2 (38... h5 {and Black is better.}) 39. Kg1 (39. Qe5+ {gives a perpetual, but this takes calculation and it&#8217;s not easy to jettison ones best asset.}) 39... Qf5 40. d5 {Finally losing.} (40. Qxc4 Qxf7 {and Black should have good chances of converting his extra pawn.}) 40... Qxe6 41. dxe6 Kf6 {and I resigned.} *
() -
Commented by

The second game, played as Black for Dulwich in the London League, shows something about calculation. What one calculates is always a choice, but some things are mandatory, though I’ve never seen an author clearly say which they are. The moral of this tale is: keep an eye on the oponent’s candidates.

You need to activate javascript to enhance chess game and diagram visualization.
1. d4 Nf6 2. Nf3 c5 3. d5 e6 4. dxe6 fxe6 5. g3 (5. c4 b5 {goes into the Blumenfeld}) 5... d5 6. Bg2 Bd6 7. O-O O-O 8. c4 Nc6 9. cxd5 exd5 {White&#8217;s play is rather passive, though Black has to be careful about d5.} 10. b3 Bf5 11. Bb2 Kh8 {the check on d5 is annoying in some lines} 12. Nbd2 Qd7 13. Rc1 Rae8 14. Re1 Ne4 15. Nf1 Qf7 16. Ne3 Be6 17. Rf1 Qh5 18. Qd3 Nb4 19. Qb1 { <span class="PgnWidget-anchor-diagram">[]</span> Black has built an excellent position and now has to look for concrete ways to win. I still had almost half an hour to move 30, then twenty minutes for the rest of the game, so there should be time.} 19... Bh3 {Four minutes here, looking for a win.} 20. a3 Bxg2 (20... d4 {wins according to the computer. If} 21. Nc4 ({another line is} 21. Bxh3 Qxh3 22. axb4 dxe3 {with a huge attack}) {Black has} 21... Bxg2 22. Kxg2 Nxg3 23. hxg3 Rxe2 {and if} 24. Ne1 {then} 24... Bxg3 25. Kxg3 Nd5 {with a forced mate coming}) 21. Kxg2 Rxf3 {Here, despite six minutes&#8217; thought, is the fatal mistake. I remember verbalising the sequence of actions, as the variations were confusing me. But the one crucial thing when sacrificing to force a win is to search carefully for the opponent&#8217;s candidates on each move. You can be more relaxed about your own candidates if you have know you have at least one good line.} ({I also considered} 21... Nxf2 {which turns out to be very good} 22. Rxf2 Rxe3 {and} 23. axb4 ({better} 23. Bxg7+ Kxg7 24. Qb2+ Kg8 25. axb4) {can be met by} 23... Rxe2 {with a decisive X-ray that I hadn&#8217;t seen}) ({Alternatively} 21... Nc6 {keeps the pressure on with d4 to follow.}) 22. exf3 Nd2 (22... Ng5 23. Qf5 {is also good for White.}) 23. Qf5 {I simply missed this. f5 was previously obstructed by the knight and multiply protected but a careful search should have spotted the move. White gains a tempo and wins at least a piece.} ({After sacrificing the exchange I had seen} 23. Bxg7+ Kg8 24. Qa1 (24. Qf5 {wins as in the game}) 24... Qxf3+ 25. Kg1 {which is very good for Black.}) 23... Qxf5 24. Nxf5 {I thrash around a bit as the team needed the point but the game is lost.} 24... Be5 25. Bxe5 Nd3 26. Bxg7+ Kg8 27. Bc3 Nxf1 28. Rxf1 Re2 29. f4 d4 30. Ba1 Ra2 31. a4 b6 32. Kf3 Nxf2 33. Bxd4 (33. Rxf2 Rxa1) 33... Nd3 34. Ke4 Nb4 35. Bc3 Re2+ 36. Kf3 Rxh2 37. Bxb4 cxb4 38. Rd1 Rc2 39. Rd3 Kf7 40. Nd4 Rb2 41. Nc6 a5 42. Ne5+ Ke6 43. Nc4 Rc2 44. Nxb6 Rc3 45. Ke4 h5 46. Nc4 {and I resigned.} *
() -
Commented by

I don’t usually play on successive days, but the day after this debacle I had to play for Pimlico. It was a relief to discover that some positions, despite their excellent appearance, can actually be won.

You need to activate javascript to enhance chess game and diagram visualization.
1. c4 {75 minutes with a 15 second increment} e5 2. Nc3 Nc6 3. g3 f5 4. Bg2 Nf6 5. d3 ({I had played} 5. e3 {a few weeks before. I am still not sure of White&#8217;s best approach here.}) 5... Bb4 {This reversed Grand Prix does quite well for Black and my opponent said he has good results with it.} 6. Bd2 O-O 7. a3 Bxc3 8. Bxc3 d6 (8... d5 {is rare but worth considering.} 9. cxd5 Nxd5 {when} 10. Qb3 {seems not to work after} 10... Be6 11. Qxb7 Qd6) 9. e3 {Rare.} (9. Nf3 {and}) (9. b4 {are more popular.}) 9... Be6 10. Nf3 h6 {With the thought of starting operations in the centre. In similar positions Black often plays Qe8-h5 with a kingside attack.} 11. b4 e4 {Afterwards my opponent wondered if this was too early.} 12. Nd4 Nxd4 13. Bxd4 Qd7 14. O-O Rae8 15. Rc1 a5 { <span class="PgnWidget-anchor-diagram">[]</span> Committing to a pawn sacrifice, in the hope of using the half-open files on the kingside.} 16. Bxf6 Rxf6 17. dxe4 fxe4 18. Bxe4 axb4 19. axb4 c6 (19... Bg4 20. Qd5+) ({Against} 19... Bh3 {I had planned} 20. Qd5+ Kh8 21. Bg2 {with the thought of bringing the Queen to g2.} 21... c6 {seems to stop this but the computer points out} 22. Bxh3 Qxh3 23. Qg2) (19... Bxc4 {loses a piece to} 20. Rxc4 c6 21. Bg2) 20. Bg2 Qf7 21. Qd4 Rf8 22. f4 {Making Black&#8217;s last few moves retrospectively mysterious. I think my opponent missed this idea when sacrificing the pawn.} 22... Ra8 23. Ra1 {I was pleased with this, further frustrating my opponent.} 23... Rd8 (23... Rxa1 24. Rxa1 Bxc4 {runs into} 25. Ra8+ Kh7 26. Be4+ g6 27. Bxc6 {which the computer likes even more than I did.}) 24. Rfc1 Qc7 25. Rd1 Rff8 26. Qc3 {Trying to restrain all of Black&#8217;s pawn breaks} ({the computer suggests} 26. b5) 26... Qf7 27. Rac1 Rfe8 28. e4 b5 29. cxb5 {I thought this won a pawn, but in any case it seems right to start opening lines.} (29. c5 {is also possible}) 29... cxb5 30. Qc6 Bg4 {I missed this and was lucky it didn&#8217;t matter.} 31. Re1 ({Of course not} 31. Rd2 Rc8) 31... Qa7+ 32. Kh1 Qa4 33. Qd5+ Be6 34. Qd4 Bc4 {I hadn&#8217;t really meant to allow the bishop to this active-looking square, but in fact it may be no better than it was on the kingside.} 35. Qd2 {Although White seems to be retreating, the threat of Ra1 and Re3 pushes Black&#8217;s queen back.} (35. Bf1 {was tempting but the ending after} 35... Qxb4 36. Bxc4+ bxc4 37. Qxc4+ Qxc4 38. Rxc4 d5 39. Rd4 dxe4 40. Rdxe4 Rxe4 41. Rxe4 {is drawn}) 35... Qa7 36. Red1 Qc7 37. Qd4 Kh8 38. Ra1 Ra8 39. h3 ({avoiding} 39. Rxa8 Rxa8 40. Qxd6 Qxd6 41. Rxd6 Ra1+) (39. e5 {is possible}) 39... Rxa1 ({Black could simply wait as the ending after} 39... Red8 40. Rxa8 Rxa8 41. Qxd6 Qxd6 42. Rxd6 Ra1+ 43. Kh2 Ra2 44. Rd1 Rb2 {could still be tricky though White should win.}) 40. Rxa1 Re7 41. Ra6 Rd7 42. h4 { <span class="PgnWidget-anchor-diagram">[]</span> With only three minutes plus increments, Black starts to cave in.} 42... d5 (42... Bf7) 43. Bh3 Qb7 44. Rxh6+ Kg8 45. Be6+ Kf8 46. Rh8+ (46. Qc5+ {is more clinical.}) 46... Ke7 47. Qxg7+ {and Black resigned. Capturing the bishop allows immediate mate and otherwise White can exchange off on d7 and win the ending, though the computer says Kd6 Qf6 is mate in seven.} *
() -
Commented by