Tuesday 16 August 2016

Beating a grandmaster

A lot of interesting things have happened in chess since August 4: the end of the British Championships, the announcement of the world championship venue, the Sinquefield Cup....

....and I'm afraid I don't give a rat's arse about any of these things, because also, between that date and this, I beat a grandmaster for the very first time. In the first round of the Prague Summer Open, about forty-six years after I learned the moves, about forty years after I first played competitive chess, I knocked over a grandmaster for the first time. I'd been close once or twice but never even managed to draw. But this time I got across the line.

Prague Summer Open 5 August 2016, Round One

White: Sergei Domogaev
Black: Justin Horton
This is a psychological error common to chessplayers of all ranks from beginner to grandmaster: to lose one's objectivity is almost invariably to lose the game as well.

David Bronstein
1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 e6 5. e3 Nbd7 6. Bd3 dxc4 7. Bxc4 b5 8. Bd3 Bd6

You won't even find this variation in a lot of books (8...Bb7, 8...a6 and sometimes 8...b4 are favoured) but this game is quite a good advertisement for its charms and it's the recommendation of Larry Kaufman's The Chess Advantage In Black And White, probably the best repertoire book I know and one I looked at on the flight from Barcelona to Prague the day before the game.

9. O-O O-O 10. Bd2 Bb7 11. Qc2 Rc8

I changed my mind many times about what openings I wanted to play in Prague. But I'd opted for this one enough times to have had this position (and almost the same one, but with different eleventh moves for White and hence not always 11...Rc8) on my smartphone, with Stockfish taking the White bits, dozens of times in the last few months. I don't think I even got a single draw in those games, but I learned enough to know that if White's going to move a knight here, it probably ought to be be the one on f3.

12. Ne4? Nxe4 13. Bxe4 f5!

This is why.

14. Bd3 c5!

Black gets his break in without even having to play ...a6 first, the queen is vulnerable and so, for that matter, is h2. White is really struggling.

15. Qd1 c4 16. Bc2 Nf6 17. Ne5?! Bxe5 18. dxe5 Qd5

19. f3?!

After this Black's just a pawn up, albeit White's two bishops are lurking with the potential to do more than lurk. I hadn't noticed that White actually has 19. Qf3 after which best is 19...Qxd2 20. Qxb7 Qd5 21. Qxd5 Nxd5 and although Black is obviously rather better, at least I'm not a pawn up.

19...Qxe5 20. Qc1 Nd5 21. Kh1 b4 22. e4 c3 23. bxc3


I thought at the time that this was automatic and that White would then take on d5, but in fact Black doesn't necessarily want that to happen. 24. exd5 cxd2 25. Qxd2 Qc3 26. Qxc3 Rxc3 27. Bb3 Bxd5 28. Bxd5 exd5 29. Rfd1

and although Black has all the chances, aren't all double rook endgames drawn? So maybe 23...fxe4 might have been preferable. As it turned out it didn't matter.

24. Bg5 fxe4 25. Bxe4 c2 26. Re1 Qd4

Now I'm sure that 27. Qd2 is objectively best, followed by exchanging queens and assuming that Black hasn't got the firepower left to force the c-pawn home or win a second one. White wanted to keep the queens on - but this way he's losing, or ought to be.

27. Qa3? h6 28. Bh4


Missed it!

29. Bg3 Rxe4! and after 30. fxe4 Nb4! wins whatever way White wants to play it. It's a hard one to see, in my view, but its secret is - in part - that it solves the problem of the white bishop, which currently inhibits the knight from adding its further weight to the attack. The move Black played was also designed to address that problem.

29. Rac1?!

The computer prefers 29. Bg3, though Black is still winning (29...Nf4! 30. Bxa8 Nd3 31. Bxb7 Qxa1 32. Rxa1 c1=Q+ 33. Rxc1 Rxc1+ 34. Qxc1 Nxc1). It stops liking 29. Qd6 after it sees that 29....Qxa1 is effective, though I can imagine 30. Qxe6+ Kh8 31. Qg6 Nf6 being pretty hairy to navigate in time trouble.

Another chance...


...another chance missed, and having played 28...Ba8 I really should have seen 29...Nf4. If White takes the c2 pawn Black will play ...Qd2 and White maybe has to try 30. Bg3 after which Black wins with 30...Bxe4 31. Rxe4 and now Black needs to see that he wins by putting a queen on d3 or by taking the rook and playing ...Ne2.

Black shouldn't get a third chance to add to the two he's missed.

30. Qd6

Now Black has to defend, which should have been straightforward enough - provided he was able to see 30...Rc6 31. Qd7 Rf7, which he was not.

30...Rfe8 31. Bg6

Whoops! Hadn't seen that one either. Now 31...Rc6 is absolutely compulsory, but instead Black went

31...Qb4 32. Qxb4 Nxb4

33. a3?

I have no idea why he played this, as opposed to the natural 33. Bxe8 Rxe8 34. Re2 followed by picking up the c-pawn and almost certain victory. Maybe he wanted to hang on to the admittedly strong bishop on g6 or maybe he saw some ghosts involving 33...Nxa2, ghosts which did not exist. Either way, having seen my likely win against a grandmaster turn suddenly into an almost certain defeat, I could see at least the outlines of a possible, plausible draw.

33...Na2 34. Rxc2 Rxc2 35. Bxc2 Bd5

The sort of move you play thinking "that ought to hold it for a while".

36. Bf2 Rc8 37. Bg6 a5 38. h4 a4 39. Bd4

Presumably played with an eye to rounding up the knight, but in fact that piece has a safe haven on b3 thanks not least to Black's last couple of moves.

39...Nc1 40. Kh2 Nb3

Time control! A very welcome extra half-hour.

Very welcome, but they had a clock setting that only added on the extra time once one of the two players had exhausted the previous allocation. This made me nervous, so I played move 41 quickly and was very pleased when he took time over his 42nd and spared me the burden of seeing if I'd remembered rightly what the time control was.

41. Be5 Nd2 42. Bd6 Nb3 43. Kg3 Rc6 44. Bb4 Rc4 45. Bd3

The turning point.


This is a mistake, but it's a winning mistake so I'm not inclined to care too much, or indeed at all.

The point is that - maybe - White can't actually make progress here, if Black plays it really well. Or, perhaps, unless White plays it really well. I don't have the skill or the application or the time to try and find out for sure, but I can see that this may be the case, and that Black's best plan is to just sit tight.

But at the time, all I could imagine happening was White gradually moving his pieces, and perhaps especially his king, into stronger positions until I couldn't sit tight any longer. I didn't fancy that, so I went looking something that would give me a little bit of initiative. This was it.

46. axb4

Funnily enough though, when I'd calculated the exchange sacrifice, for most of the time I'd managed to envisage positions where for some reason this pawn had ceased to exist and mine was the only pawn in a threatening position. Ah, I thought, that's a bit awkward...


47. b5?

What a stroke of luck! But why did he not play any of these moves? What did he have in mind instead?

47...a2 48. b6


I'd worked it out - 48...a1Q?? loses immediately to 49. Rxa1 Nxa1 50. Be4!. I also had to see 48. Be4 Bc4! - but having seen both lines, I also saw that I really shouldn't lose this game any more, not even against a grandmaster.

49. Kf4 a1=Q 50. Rxa1 Nxa1

51. Ke5?!?

If 45...Rxb4 was the winning mistake, this is the losing...

...losing whatever. I confess I cannot think of an appropriate antonym for "mistake". It's not a losing move, as such, that's not the problem at all. The problem is that it just gives White an opportunity to lose, and it's an opportunity which he eventually takes.

I was amazed when he didn't move the bishop, but after 51. Be4 Ba6 52. b7 Bxb7 53. Bxb7 Nc2 54. Ke5 Kf7 55. Bc8 Ne1 the draw basically signs itself and 54. Be4 is a trap (54...Ne1??) which presumably he didn't feel worth playing for. So presumably he was keeping the draw in hand as long as possible and hoping something came up. But he overdid it.

It's only thinking about this now that I grasp why, maybe, he didn't play 19. Qf3.

51...Nb3 52. Kd6 Kf7 53. Bc4 Na5 54. Bxe6+ Kf6 55. Bd7

OK it's still drawn, but how is he expecting me to go about losing this?

55...g6 56. Be6 h5 57. Bd7 g5

58. Be6

It was surely time now to find a way of getting a draw: my computer suggests lines like 58. hxg5+ Kxg5 59. Kc5 Kf4 60. Kb5 Nb3 61. Kc4 Nc1 62. Kb5 Ne2 63. Kc5 Ke5 64. Bc6 and there are all sorts of other variants. But 58. Be6 is one subtlety too many, at least once you realise that just because it's on the c8-h3 diagonal doesn't mean the bishop can stop the h-pawns.

58...gxh4 59. Bh3

Trying to plug the gap, but the bishop isn't big enough to fill it. Still, 59. Bd5 Ba6 60. Be4 Kg5 61. Kc5 Kf4 62. Bd5 Kg3 is hopeless and so is everything else.

59... Kg5 60. Kc7

making it easy...

60...Kf4 61. Kd6 Kg3 62. Ke5


...though I confess that it was only now that the thought "I'm winning here!" came back into my head, from whence it had fled more than thirty moves previously.

63. Kf6 Nxb6 64. Kg5 Bxf3

Even now, it surprised me when Domogaev stretched out his hand to offer his resignation. It might have been interesting had he been able to exchange his bishop for the two h-pawns and make me mate with bishop and knight: but in fact, he can't even achieve that.

The wheel turns. Domogaev won his next six games and played on top board in round eight. I collapsed at the end of the tournament and lost my last three games, one from about +8 and the other two in long games against sub-1900 players. (I finished the tournament with thirteen scoresheets from nine games.) But it doesn't matter. Maybe no game of chess ever need matter to me again, because after all this time, during most of which I never even imagined it might even happen, I have finally beaten a grandmaster.

Only once, to be sure. Quite likely I'll never do it again. But the difference between zero and one is greater than the difference between one and one hundred.

-- -- -- -- -- --

[This blog is off to the UK for business, seeing friends and family and the e2e4 August Bank Holiday Congress. Martin will be making some posts in the meantime. I should resume some time in early September.]


Anonymous said...

Domogaev himself is a bit of a late bloomer to make it to GM.
Born 1964, and FIDE lists 3 norms from GM International Tournaments in 2010 and 2011 (Donskoj Kireevsk Bogoroditsk), the last one leading to a 2514 rating.

Jonathan Rogers said...

Justin seems to be saying "it doesn't matter how I did it - I beat a GM!"

And he's quite right too. Good for you, sir!

Paul said...

You did seem to beat a solid > 2400 IM here: https://www.chessgames.com/perl/chessgame?gid=1560158 so perhaps the fact that you went on to beat a GM is a bit less surprising than you make it sound. I'm a bit surprised that you don't mention the IM scalp (Kozhuharov). A few words about that game would be welcome if you have the time?


Paul Epstein

ejh said...

Thanks for noticing! Kozhuharov is probably the best player by rating I've ever beaten but in fact, I don't think he officially had the title at the time of the game.

Paul said...

Giant-killer games are always interesting. Great that you've posted about these experiences. How do you think your performances against elite players (say 2400 FIDE or higher) compares with other players of your strength? If you feel that you can pose more problems to the elite than other players of your strength, it would be instructive to explain why.

Paul Epstein

ejh said...

I suspect I pose fewer