Wednesday, July 25, 2012

Penarth 2012 : Streatham and Brixton against the grandmasters

Just as in our last outing to an international tournament, the Streatham and Brixton contingent at the South Wales International played three games against grandmasters*. Although the pattern was different - this time we all played one grandmaster each - the overall result, you'll not be astonished to hear, was precisely the same, to wit three points to the grandmasters and none at all to the Streatham and Brixton team.

However, just as with last year's series, we came away from the games feeling that we ought to have done better. Especially so with the two games that took place in the very first round: Angus French's game against Keith Arkell and mine against Boris Chatalbashev, from Sofia.

Not one of our victims at Penarth

I had the White pieces. Larger diagrams indicate positions from the game, smaller ones positions from analysis.

1. d4 Nf6
2. Nf3 c5
3. d5 b5

I've been on the White side of this three times, in which capacity I have achieved three losses. To be fair, two of these games have been against grandmasters.

4. Bg5 Ne4
5. Bf4 Bb7
6. a4
If this looks at all familiar to blog readers, that's possibly because I played the same line against German GM Henrik Teske a couple of years ago while Stephen Gordon was doing exactly the same thing at the adjacent board. However, Teske - and indeed Gordon's opponent - played 6...b4.

7. c3 b4
8. c4 e6
9. Qd3

This is the idea that Gordon played at Benasque (and I did not). But though in this slightly different position it's still not a bad move, White can play more naturally with 9.Nbd2. There's no need to worry about the d5 pawn as after 9...exd5 10.cxd5 Bxd5 11. Nxe4 Bxe4 12.Nd2 Bc6 13.Nc4

White is doing very well indeed.

9...f5 I don't really believe this, though come to that I don't really believe that by the time I played my reply to it, I'd used 68 of my 120 minutes (until move 40) contemplating ten fairly straightforward moves. I don't know what I was thinking about.

10. Nbd2 Be7? Nor do I know what he was thinking about when playing this, though I suppose it may be his previous move that's really at fault if the knight can't be sustained on e4. What I do know is that after this, White is suddenly a pawn up for no very obvious compensation against a grandmaster.

11. Nxe4 fxe4
12. Qxe4 O-O

13. e3
More than that, White is also going to have a dangerous attack on the castled king. I could see that I might have to abandon king safety myself in order to make it effective, but I was ready to do that. At this point, and for the next half-dozen moves, I thought I was going to win the game.

14. Kd1!

Obviously I could see how this might end in tears - as indeed it did - but I didn't hesitate long before playing it. 14.Nd2 looks natural but after 14...Na6 15.Bd3 g6 White still can't castle because the knight is hanging if he does - while White will have one piece fewer with which to attack the king. Yes, after 14.Kd1 White's king is uncastled in the centre, but what - I thought - are any of Black's pieces doing? And how do the queenside pieces get across to defend the king?

15. Bd3 g6
A critical position. White has a lot of options and even after a spell with a computer I am not at all sure what was correct here. Or for that matter, incorrect.

16. Bd6!? Chatalbashev didn't like this but I think it keeps a large advantage and I'm not convinced there's a better move. He indicated 16.h4 but this doesn't seem to work provided Black plays 16...Na6! rather than 16...Bxb2, the move I, wrongly, thought was the refutation but which loses, totally, to 17.h5 Bxa1 18.hxg6 Bg7 19.gxh7 + Kh8 20.Be5! (the move I missed).

After 16...Na6 however Black seems to survive, some long computer lines going 17.Ne5 exd5 18.cxd5 Rae8 19.Bc4 d6 20.Nc6 Rxe4 21.Nxa5 Rxc4 22.Nxc4 Bxd5 23.Nxd6 Bxb2 24.Rb1 Bg7 with advantage to Black despite White's extra exchange

and 17.h5 exd5 18.cxd5 Rae8 (you can see how 16.h4 perhaps just gives Black the chance to get his pieces over) 19.Qc4 Nb4 20.Qxb3 Nxd3 21.Qxd3 Bxb2 22.hxg6 Bxa1 23.g7 Bxg7 24.Qxh7+ Kf7

and now the computers tells me White has a choice of perpetuals, but no more. Of course, I'd have loved to take a perpetual check against a grandmaster and if I'd been able to analyse that far (which I wasn't) I might even, at this point, have taken it. But as it happens, I think White can do better.

But with what? Maybe 16.Be5, which I considered but rejected as 16...Bxe5 17.Nxe5 Rxf2 gets at White's underbelly. But 17.Qxe5, which I didn't consider, is better, and also gets the queen onto a less vulnerable square. Now there are multiple lines but Black doesn't appear to have better than 17...Na6 18.Qg3 Rf7 19.Ng5 Rg7 20.dxe6 Nb4 21.Be4 dxe6 when 22.Bxb7 Rxb7 23.Ke2 looks safer than taking the e6-pawn, and White should be better.

Then there's 16.Ne5, which the computer likes and carries on liking down another long line: 16...exd5 17.cxd5 Re8 18.Nc4 Rxe4 19.Nxa5 Bxd5 20.Bxe4 Bxe4 and now 21.Nxb3 (but not 21.f3 Bc2+ 22.Kc1 Na6 23.Ra3 c4! 24.Nxc4 Nb4/Nc5 25.Kd2 Nd3 26.Na5 Nxf4 27.exf4 Rb8 and Black is fine) 21...Bxg2 22.Rg1 Bf3+ 23.Kc1 Na6 24.Nd2 Bc6 25.Nc4 Re8

with a position in which it's hard to believe that Black has enough play for the exchange, but finding ways to actually free White's pieces to prove it is not straightforward either. But White, once again, must surely be better.

Then there's the game move, 16.Bd6.

16... Na6 I think I was expecting 16...Rf7 (16...Re8 17.Qg4) after which I couldn't swear what I'd intended to play. 17.Qg4 is actually a draw (which, as per my comment above, I might well have taken if I could see no better) after 17...Bxb2 18.Bxg6 etc given that 17...Na6 transposes into the game, if indeed White has no better than 18.Ne5.

But 17.h4 looks better than 16.h4 did: 17...Na6 (17...exd5?? 18.Qe8+) 18.h5 and now if 18...exd5? 19.Qg4 since the bishop, now on d6, no longer obstructs the queen's path to this square. So Black can no longer get his pieces across so swiftly and that probably means that after 16...Rf7 17.h4 White is winning.

17. Qg4! I never seriously considered taking on f8: I was simply expecting to win by a direct attack, and I was possibly justified in thinking so. Does 17.Bxf8 win? 17...Rxf8 18.Ne5 and now Black seems to have 18...dxe5. You can also try 18...Bxe5 19.Qxe5 Nb4 but White seems to keep the game going with 20.Be2! Nb4 21.Rb1 Rxf2 22.Qg3 Rf8 (22...Rxg2 23.Qb8+ Kg7 24.Qe5+) 23.Rf1 Rxf1+ 24.Bxf1 exd5 25.Qb8+ Kf7 26.Ke2 Qa6 27.Qf4+ Kg8 28.Kf2.

I assure readers I have not forgotten what Larsen said about long variations. Anyway, as suggested above Black seems to do better with 18...dxe5 and now 19.cxd5 Bxe5! (19...Re8 20.Nc6!) 20.Qxe5 Nb4 21.Bc4 Ba6 22.Qc3 Rxf2

and now 23.d6+ Bxc4 24.Qxc4+ is no more than a draw (though once again, that's a draw I'd have been delighted to get). So maybe White does best to spurn the exchange. And after

17...Rf7 he can no longer take it anyway.

18. Ne5? I thought I needed one more piece in the attack! But in retrospect it's incomprehensible that I didn't play 18.dxe6, which looks like it should win. Possibly I was thinking of 18...dxe6 19.Qxe6?? which loses, and very quickly, to 19...Rd8, but 19.Bxg6+ is obvious and good.

Maybe (I simply cannot remember) I thought that after 19...hxg6 20.Qxg6+ Rg7 21.Qxf6 I was in trouble on the d-file, but Black doesn't have any way of getting there and is simply three pawns down. Maybe I missed that unlike 18.Bxg6? Rg7, 18.dxe6 dxe6 19.Bxg6 Rg7? loses, as White takes on e6. Maybe I thought 19...Bxf3 was a deflection but after 20.Qxf3 both Black rooks are hanging and so after 20...Rd8 21.Bxf7+ Kxf7 22.Ke2 Rxd6 23.Rd1 (either rook) Rxd1 24.Qb7+ Be7 25.Rxd1 Qxa4 26.Qe4

Now, whatever the balance between Black's two pieces and White's rook and pawn, the difference between the respective queens and kings is surely the difference between winning and losing.

Who knows? I doubt I was very much troubled by 18...Rg7, after which White has a variety of good moves including 19.Ke2, 19.Be4 (19...Bxe4 20.Qxe4 Re8 21.Be5) or perhaps best of all 19.exd7 after which 19...Qd8 20.Qe6+ Rf7 21.Ne5 Bxe5 22.Bxe5 and given that White is three pawns up, he can expect to bring the point home after, say, 22...Nc7 23.Bxc7 Qxc7 24.Ke2 Rd8 25.Be4 Bxe4 26.Qxe4 and after Black takes on d7, a rook comes to d1 and I do not see what Black has for the two pawns White remains ahead.

That was the moment, I think. Even after missing 18.dxe6, White is probably still better. But that was, I think, a win that just passed by.

18...Rg7 and not 18...Bxe5 19.Bxe5 Rxf2 20.Bc3 winning.

19. dxe6? One move too late, because everything has changed, though I didn't understand this. I still expected to win, and shortly at that, and it wasn't until Black's reply that I realised everything had gone wrong. But now, Black's rook is no longer attacked, and that gives him the chance to bring another piece across and turn the tables.

White should still have the advantage, but a direct win is out of the question. It's hard to pin down his best line (which is the nature of a position in which no direct win is available) so for that reason, I'll just give one possible line: 19.Qf3 Qb6 20.Nc6 Nb4 21.a5 Qa6 22.Nxb4 Qxd6 23.Qxf6 Rf7 24.Qh4 cxb4 25.f3 exd5 26.Qd4

Well, who knows. White plainly has the better pawns. I might fancy this against a lower-rated opponent and with lots of time on the clock. But against a grandmaster and without much time (after move twenty I would have fourteen minutes left) I think not.

19.dxe6, though, is a poor move. But I didn't know that yet. And after I had played it, I was waiting to see how I was going to win, since 19...dxe6 loses immediately, while anything else just seemed to give me captures on d7 and a queen entry on e6. Anything, that is, that I had actually seen.

19...Qb6 I had completely missed this. Missed the move, missed the idea. But suddenly my king and bishops on the d-file looked very awkward and very vulnerable indeed. Although it still seemed to me that I was two pawns up: it wasn't until both those pawns had been taken back that I realised I was lost.

20. Nf7 Assuming that I actually was lost, either here or after the pawns were won back. I must confess to having rather less will to find out whether this position could have been saved, than to find out whether some of the positions above could have been won: and nor did I have time to devote to the problem during the game. I played the game move more out of an instinctive feeling that it kept my pieces defended than anything more analytically-based. Obviously I saw 20.Nxd7, but it felt much looser: it might, however, very well be better, and perhaps easier for White to play than the game, as some forced exchanges clarify the position a little. 20...Rxd7 21.exd7 Qxd6 22.Ke2

Now Black would like to have the b2 pawn but 22...Bxb2 23.Be4 looks OK for White so instead it may be better to put the queen where it inhibits Be4, while still threatening b2. So maybe 22...Qe5 (also possible is 22...Qe7, but let's not overdo it more than we have done already) when 23.Rab1 Rd8 - followed by ...Bc6 - leaves White tied up. So better 23.Rad1 and now 23...Rd8 (23...Qxb2+ gives White a couple of very handy tempi) after which there are a couple of options:

a. 24.Rd2 Bc6 25.Bb1 Nb8 26.Rhd1 Rxd7!

and now White can't take on d7 because the pawn on b2 goes (and maybe the bishop on b1 too) but is he actually losing yet? I would hazard not, theoretically speaking, but assuming a pair of rooks come off I wouldn't back White's remaining rook against those bishops.

b. 24.h4, having a go instead of battening down. 24...Bc5 25.h5 Bxd7 26.Qe4 Qxb2 27.Rd2

with a mess in which further analysis tends to achieve little except the probable production of further diagrams. Suffice to say that White is struggling to hold on, but it's quite possible that he does. Provided he's got plenty of time, which he did not, and a computer. Which he did not.

That said, the game continuation may be no worse.

21. Nh6+ Kh8
22. Qf4
22.Qxe6 looked, and is, too dangerous after 22...Bxb2 (or, indeed, the curious 22...Rf8). But now Black gets the second pawn back and with it, most of the attacking chances that just a few moves ago appeared to belong to White. But not all of them.

22.. Bxb2
23. Ke2!
Not a bad find, especially given that I had just a couple of minutes left.

The trick here is that after 23...Bxa1 24.Rxa1 White forces a draw. Black may have to find 24...Nb4 (24...Qd8 or 24...b2 allow 25.Rad1 when White has chances of an advantage) when White plays 25.Be5 Nxd3 26.Nf7+. Not that I was expecting Black to allow this. Or that I saw all of it.

24. Be5?!
Maybe White should play the other obvious move, 24.Rad1, though my computer wheezed its way to a win for Black after 24...Ba6 (24...Bxg2 25.Rg1 followed by the draw with knight checks) 25.Kf3 (25.Bf8 Rd8 26.Bxg7+ Kxg7 27.Ng4 Qd6) 25...Nxd3 26.Rxd3

and now 26...g5 27.Nf7+ Rxf7 28.Qxf7 Bxc4 29.Rdd1 Qc6+ 30.e4 Bg7 and then 31.h4 b2 appears to win the game. But it took some fiddling about to find that line and maybe a better computer than mine will come up with better for White. Besides, after 24.Be5, Black's other rook gets into play and I think White is definitively lost.

25. Qxe5 Rf8
26. Rad1
Quite likely I should have played the other rook, but it's not important.

26...Bxg2?! As it happens this is not a particularly good move. 26...Qc7 should see White off rather more simply.

27. Be4! Apeing the theme introduced by the other White bishop. Obviously if Black takes on h1 then 28.Rd7 wins.

27...Nc6 Black actually has to find this! But once the only move is found, the game is still won....

28. Bxc6...although Black still has to find the right way. How would you recapture?

* * * * * * * *

* * * * * * * *

28... Bxc6 If you preferred 28...Qxc6 - as did my colleagues later that evening, in the pub - then you need to explain what you're going to do after 29.Nf7+. Chatalbashev did not make this error.

29. Rd6 Bf3+?
30. Kf1?

A shame not to take advantage of some inexact Black moves. After 30.Kd3, I don't even think White is necessarily losing.

Now what? Black, in fact, has only three moves, all with the queen. 30...Qc7 is no longer protecting the b-pawn after 31.Rb1. 30...Qb8 seems better but after 31.Rb1 Bc6 32.Qxc5 (there are alternatives) 32...Bxa4 33.f4

and while Black is now a pawn up, the e6 pawn is liable to drop. Maybe 33...Qc7, but after 34.Qxc7 Rxc7 35.Ng4, White to my eyes is looking good for the draw.

But if 30...Qb7 31.Rb1 Bc6 (his other pieces are stymied) 32.Rxe6 - back level on material - and now 32...Bxa4 33.Rf6 Rxf6 34.Qxf6 Qe7 35.Qa1! (computers, eh?)

and Black doesn't seem able to make progress.

So - last variation! - Black tries 32...Qxc7 instead and after 33.f4 (after 33.Qxc7 Rxc7 Black's bishop is able to make a surprising amount of trouble for the knight, so White gives it the chance to take itself elsewhere) 33..Bxa4 - there doesn't seem to be an effective waiting move - White looks OK after 34.Qxc7 Rxc7 or even more so after 34.Kc3!

and as per the diagrams above this one, it looks to me as if White is likely to hold the draw.

All of which analysis is rendered pointless not just because I played a different move, but by the fact that I had less than a minute to play eleven moves and would almost certainly have failed to make the time control. But, still. I am looking at it, and thinking that even at the death, I didn't have to lose.

Even Black's last move wasn't the best (30...Qb8, 30...Qc7). But it didn't matter.

30...Qb7 and White resigned.

I said after the game that I thought I had been winning, He said he didn't think so, though it was complicated. It was complicated all right. But I still think I might have been winning.

I lost. Chatalbashev went on to win the tournament.

That was the top board game. On board four, Angus was playing Keith.

Not one of our victims at Penarth

Angus' notes are in italics.

My third game against Keith. In the other two, I had White. And in the first of those I'm sure I had an advantage out of the opening (a Caro-Kann) though it was wasted through a series of inaccurancies. Still, the game lasted just short of seven hours.

1. d4 d5
2. c4 c6
3. Nf3 Nf6
4. Qc2 g6
5. Bg5 Bg7

I decided against 5...Bf5 because I didn't like the look of 6.Qb3. However, after 6...Qb6 7.Qxb6 axb6 8.Bxf6 exf6 9. cxd5 the open lines should probably compensate for the pawn weaknesses (9...cxd5 or 9...Bb4+ or 9...Bxb1).

6. Nbd2 O-O Maybe 6... Bf5. [I thought the set-up Black chose was a little passive - Slav-playing ejh]

7. e3 Qb6
8. Bd3 Re8
9. O-O Nbd7
10. Rfd1?!

Possibly not accurate as now Black gets some play: better 10.cxd5 cxd5 11.Bf4

11. dxe5
[I wondered whether 11.cxd5 cxd5 - 11...Nxd5 12.Nc4 - 12.dxe5 Nxe5 13.Nxe5 Rxe5 14.Bf4 Re8 might have been a superior version of the game, as it seems to leave Black a bit more passively placed. But Black can play 12...Ng4 which seems better. It reminds me of some lines of the Chebanenko Slav, albeit with the pawn still on a7 - ejh]

11...Nxe5 11... Ng4!?
12. Nxe5 Rxe5
13. Nf3
[Now if 13.Bf4 - though it is probably a better move- Black plays 13...Re6 and White doesn't have 14.cxd5 (14.c5 and 14.Rab1 are superior) 14...Nxd5 15.Nc4 because now the d-pawn has already been exchanged, Black can go to c5. Which is why the rook goes to e6 rather than the apparently superior e8, since otherwise White could continue 15...Qc5 16.Bd6 - ejh]

14. Bxf6!?
To give Black an IQP. I can't say I minded as it seemed that my pieces would come to good squares. [I wanted to play 14.Rab1 in the pub post mortem, but after 14...Ne4 Black is already on top. As indeed he is after the game move - ejh]

15. cxd5 cxd5
16. Rd2 Be6
17. Be2 Rec8
18. Qd1 Rc5

19. Ne1 White's pieces are in retreat! But might they pop up in better positions, I wondered? Also possible 19. Nd4.

19... Rac8
20. Nd3 R5c7
21. Bg4 Bxg4
22. Qxg4 Rc2
23. Rxc2 Rxc2
24. Qd1 Qc7

25. Rb1 a5
26. g3 Qc6
27. a4 Kg7
28. b3 Qc3
I liked the look of c3 for my queen but Houdini 1.5 tells me that now would have been a good time to play ...d4 when either the pawn is swapped off or, after e4, it becomes passed. Apparently the position is level after the move I played.

29. Nf4

29...d4 29...Rd2!? [Comes to much the same as far as I can see - after 30.Qg4 Qe5 White will probably play 31.Rd1 so the rooks will come off and Black will get in ...d4 - ejh]

30. Nd5 Qd2
31. Qf3 Be5
32. Rd1 Rc1
33. Rxc1 Qxc1+
34. Kg2

34...dxe3?! I wanted to play 34...d3 but it's no good because of 35.Qe4, so I swapped off the pawns, not seeing that I could have left mine where it was. Better is 34...Qc2 - intending ...d3 - 35. exd4 Bxd4 36. Qf4 Qb2 and the position is apparently dead equal. [Or indeed 34...Qb2 or 34...Qd2, though the latter is a bit hairy - ejh]

35. Nxe3 b6? I should have brought back the queen to defend with 35...Qc7. Now White's knight is superior to Black's bishop. [My computer likes 35...Qc6 which looks at first like an invitation to lose a NvB ending, but it might be all right. I'd be reluctant to defend it against a computer though. Or against Keith Arkell - ejh]

36. Nc4 Bd4 [As this seems to lose it's worth hunting round a bit to try to find a way for Black to hold on, which is not easy as the b-pawn cannot be defended (36...Bc7 37.Qb7 Bd8 38.Nd6). I thought Black might have a chance with 36...Qe1 and if White allows the minor pieces to come off - e.g. 37. Qd5 Bc7 38.Qd4+ Kg8 39.Nxb6 Bxb6 40.Qxb6 - then surely Black ought to have good drawing chances]

[but if simply 37.Nxb6 I don't see a way for Black to force off the minor pieces, the retention of which is surely to White's advantage. If we play the long-variation game again we can come up with something like 37...Qb4 38.Qe3 Bd4 39.Nd5 Qc5 40.Qe4 Bxf2 41.Qe5+ Kh6 42.Qf4+ Kg7 43.Qf6+ Kg8 44.Ne7 Kf8 45.Nc6 Kg8 46.Ne5 Bd4 47.Qxf7+ Kh8 48.Nd7 Qc2+]

[and it is not yet obvious to me either how White is making progress in winning the game, or how Black is making progress in drawing it. Ultimately though though I'd be surprised if that were due to the earlier position being drawn, rather than my own indolence in finding the win - ejh]

37. Nd6 And now I saw a problem with my intended ... Qc7.

37...Qd2 [Loses straightaway, but Black doesn't seem to have better than 37...h5 38.Qxf7+ Kh6 39.Qf8+ Bg7 40.Nf7+ Kh7 41.Qd6 and Black can't both defend the b-pawn and keep the queens on - ejh]

38. Qxf7+ Kh6
39. Qf8+ Bg7
40. Nf7+ Kh5
41. Qxg7 Qd5+
42. f3
and I, of course, resigned. I was slightly disappointed as the game had been pretty equal for a long time and I only tripped up when approaching the time control. That said, I don't think I played badly and I made Keith work a little for his win.

He does work a little bit, Keith Arkell. Nobody will be surprised to learn that his was the last game to finish in the final round.

Jonathan's second round game against Marijan Petrov was a little more swift to come to a conclusion and we shall be rather more swift in presenting it here. But do not entirely neglect it...

Not one of our victims at Penarth

Jonathan had White.

1. Nf3 g6
2. d4 Bg7
3. c4 c5
4. e4 Qb6
5. dxc5 Qxc5
6. Nc3 Bxc3+

[There's that bishop-for-knight exchange that Jonathan mentioned last Wednesday, a theme which, Angus points out, also occurred in his game above - ejh]

7. bxc3 Nf6
8. Be3 Qa5
9. Bd4 Nc6
10. Bxf6 exf6
11. Qd2 O-O
12. Bd3 b6
13. Nd4 Ba6

14. Nb5 Ne5
15. a4 Rac8
16. O-O Nxc4
17. Bxc4 Rxc4
18. Qxd7 Rxa4

and White resigned, as he was losing a second pawn after 19.Rab1 Bxb5 20.Rxb5 Qxc3.

Or was he?

[* for the record, I also played, and drew with, WGM Maria Leconte, which game can be found on the playthrough here. But we're talking, in this piece, of the full grandmaster title.]

[Penarth 2012 index]

[Chatalbashev image: Chessbase]
[Arkell image: Ray Morris-Hill]
[Petrov image: Arctic Chess Challenge]

[Thanks to Sean Terry for game playthroughs]

[S&B against the grandmasters: Benasque 2011]


Michael Yeo said...

Thanks for this. I am surprised there is no reference to luck anywhere, especially in your game!

In answer to your last question, 21.Rxb6 would appear to win a pawn back. White is still worse, but the position isn't resignable yet, even against a GM.

ejh said...

Indeed not On which theme a further piece will be appearing...