At the start of December I was 1951 - my highest rating in almost a year.
Now I am 1854 and fast approaching my all-time low of 1821 from July 1917.
It has got me thinking about better times, and specifically how to recapture those times.
I do not know if I have the answers, but my musings reminded me of a tournament where everything, or almost everything, went right.
I guess we all have occasional tournaments like that, but, for me, Lausanne 1999 was something really special.
In a way it should not have been special at all as I was feeling off chess when I flew to the French-speaking city on the Swiss shore of Lake Geneva.
The tournament was a nine-round swiss over five days, but I was so uninterested that I did not bother to take my computer or any chess books.
My results soon rekindled my enthusiasm.
Jānis Klovāns (GM 2465) - Spanton (2096)
Lausanne 1999 Round 1
Sicilian Accelerated Dragon
Klovāns' Wikipedia entry shows he was a nine-tine Latvian champion, being awarded his grandmaster title for winning the 1997 world senior championship. He was to win the world seniors again in 1999 and 2001, the latter year also seeing him become a correspondence GM. His elo peaked at 2530 three years before my game against him, and the entry speculates that his position as a career officer in the army of the Soviet Union reduced his chances to travel abroad and earn Fide norms. Klovāns died in October 2010, aged 75.
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Bg7 5.Nc3 Nc6 6.Be3 Nf6 7.Bc4 Qa5 8.0-0 0-0 9.Bb3 b6!?
I was certainly out of theory by now, and probably had been for some time as the Accelerated Dragon was a new part of my repertoire. My notes at the time show the analysis engine Rybka3 "quite likes my choice; Stockfish4 doesn't." Stockfish4's modern descendant, Stockfish10, at first rates ...b6 equally with the normal ...d6, but very quickly agrees with Komodo10 that the move is not good.
10.Nd5!?
9.b6!? first features in ChessBase's 2020 Mega database in 2008, played by a 2238, and again last year, played by an 1196. Both games saw White reply 10.f4 and win.
10...Nxd5 11.exd5 Nxd4 12.Bxd4 Bb7
Stockfish10 and Komodo10 quickly come to prefer my other major thought of 12...e5, when after13.dxe6 dxe6 14.Bxg7 Kxg7 15.Qd4+ my notes say White is better "thanks to his superior bishop and safer king." However, the modern engines reckon Black is fine after 15...e5 or 15...Kg8.
13.Re1 e6
Immediately on moving, I realised White has a possible combination.
14.Bxg7 Kxg7 15.Qd4+ Kg8
 |
| How should White proceed? |
*****
*****
*****
*****
16.Rad1
Klovāns thought for some time here, presumably looking at 16.dxe6 dxe6 17.Rxe6!? Qg5 (after 17...fxe6? 18.Bxe6+ Rf7 19.Bxf7+ Kxf7 20.Qd7+, White emerges two pawns up) 18.g3 fxe6 19.Bxe6 Rf7 20.Qd7 Qf6 21.Qxb7 Qxf2+ 22.Kh1 Raf8, and now 23.Bxf7+ Rxf7 24.Qe4 leaves White up a pawn, but Black has full compensation, according to Stockfish10 and Komodo10. This is a long line, but most of it is forced and I was able to calculate it at the board. What I missed is that White has alternatives at move 23, with the modern engines reckoning 23.Rd1 or 23.Qe4 gives White a slight edge. However, they are not sure about this, and anyway the edge is small, so no wonder Klovāns did not go down this line.
16...Bxd5!?
Swopping off into an ending I hoped to hold.
17.Bxd5 Qxd5 18.Qxd5 exd5 19.Rxd5
In going for the ending, I had judged that 19.Re7 Rfe8 20.Rxd7 Re2 gives Black sufficient counterplay. The modern engines' 20...Rac8 is probably even better.
19...Rac8 20.c3 Rfe8 21.Rxe8+ Rxe8 22.Kf1
 |
| Black's extra pawn-island means White is better, but by how much? |
*****
*****
*****
*****
Komodo10 and Stockfish10 reckon White's edge is worth less than half a pawn (much less in Komodo10's case). However, even though this is a rook-and-pawn ending, I was not expecting to be offered a a draw anytime soon …
22...Re7 23.a4 Kf8 24.a5 Ke8 265axb6 axb6
The number of pawn-islands remains the same, but now Black has two isolanis.
26.Rd6 Re6
Black has no choice but to allow a pawn-ending.
27.Rxe6+ fxe6!
My notes show Fritz5 "narrowly prefers" 28...dxe6, as does Stockfish10 and, given enough time, Komodo10, but I was not about to allow White a farside pawn-majority.
Having typed this, and so given Komodo10 and Stockfish10 more time to consider the position, I am chuffed to find they have come to prefer my move, and so I have upgraded ...fxe6 from !? to !
28.Ke2 d6!?
My thinking here was that the text, as opposed to the more obvious 29...Ke7, would allow me to meet Kd3-c4 with …Kd7-c6.
29.Ke3 Ke7 30.Kf4 Kf6 31.h4 h5?!
Komodo10 and Stockfish10 at first much prefer 31...d5, and if 32.g4, then 32...e5+, but come to view the text as also good enough to hold. Nevertheless, my move makes matters trickier.
32.c4 d5?
Any king retreat is correct. The text loses.
33.cxd5 exd5 34.f3?
Missing the win, which the engines show is to be had by, for example, 34.b4 (34.b3 also seems to win) b5 35.Ke3 Ke5 36.Kd3, when Black will run out of tempi and so lose the d pawn, one line going 36...Kf5 37.Kd4 Ke6 38.f3 Kd6 39.g4 Ke6 40.g5 Kd6 41.f4 Ke6 42.Kc5 Kf5 43.Kxd5 Kxf4 44.Ke6 Kg4 45.Kf6 Kxh4 46.Kxg6 Kg4 47.Kf6 etc.
34...Ke6 35.g4
The push b4 is no longer a major threat as it can be met by, among other moves, ...Kf6. But trickier is 35.Ke3!? when, according to the engines, Black has to find 35...Kf5! as 35...Ke5 sees Black end up in zugzwang again after 36.g3, eg 36...b5 37.b4 Ke6 38.Kd4 Kd6 39.g4, as in the previous note.
35...Kd6 36.gxh5 gxh5 37.Kf5 Kc5 38.f4 Kc4 39.Ke5 d4 40.Ke4 d3
Also drawing is 40...b5 41.f5.
41.Ke3 d2 42.Kxd2 Kd4 43.f5 Ke5 44.Ke3
Or 44.f6 Kxf6 45.Ke3 Ke5=.
44...Kxf5 45.Kf3 Ke5 ½–½
Afterwards, Klovāns complained that the ending had been too simple to create winning chances, but, as we have seen, that was not the case.
No comments:
Post a Comment