With no recent games, I figured I’d fill in some of the blanks with older games. Hopefully I can find enough games to keep this “series” going …
Back in 2000, I beat my first FIDE 2600+ opponent at the Koltanowski Memorial in San Francisco. I had started out pretty well at the event: I drew in the first round with black against GM Shulman (where I had the better of the draw), followed by 2.5/3 against non-GM opponents. In round 5, I was paired with Ehlvest. At the time, he was about 2630 FIDE and in the top 50 in the world (not his peak as a top-5 player by rating, but still, I’d take either one!).
I had played Ehlvest earlier in the year, also as Black, and he managed to engineer a miraculous escape in my time pressure:
1r4kr/2q2p2/5Pp1/p2p3P/1P2np1Q/5N2/2P4P/R5RK b - - 0 35)
I was tempted with a little tactic to win White’s queen, and played 35…Ng3+?. If 36.Rxg3, then White is just down an exchange with no compensation after 36…fxg3, while 36.Kg2 Rxh5 37.Qg4 Qxc2+ leads to checkmate.
Thus, Ehlvest had to play 36.hxg3, and after 36…Rxh5 37.Qxh5 gxh5 38.gxf4+ Kf8, I thought I was on top – White only has a rook and knight for the queen, and while both kings are exposed, I didn’t see so many threats for him.
Unfortunately, after 39.Ne5! Qc3!? (39…Rxb4 is a better try, but should lead to a drawn rook and pawn endgame after 40.Rg7) 40.Kh2!, it’s a forced draw!
1r3k2/5p2/5P2/p2pN2p/1P3P2/2q5/2P4K/R5R1 b - - 0 40)
After 40…Qxc2+ 41.Rg2 Qf5 42.Nd7+!!, Black can’t take the knight because of 43.Rag1. Thus, I had to take a perpetual after 42…Ke8 43.Nxb8 Qxf4+ 44.Kh1 Qh4+ 45.Rh2 Qe3+ 46.Rg2 Qh4+ (White can’t play Kg1 ever because of …Qd4+).
Instead, going back to the original position above, 35…Qxc2! would have won, as there isn’t a good way to deal with the threats of …Nf2+ or …Ng3+ anymore.
In our second tussle, it was another French Defense, but instead of 3.Nc3 and a MacCutcheon Variation from the above game, Ehlvest played the Advance Variation. The Advance is a funny variation in that it seems extremely logical for White, but it’s far and away the line against which I’ve had the most success – in classical play, I’ve scored 81% (!!) against the Advance. At the time, though, I had only played the French about 7 times against 2200+ opposition, and my only two losses ever against the Advance came in that time frame, so maybe that’s why Jaan choose it for this game.
r1b1kb1r/pp3ppp/1qn1p3/3pPn2/1P1P4/P3BN2/5PPP/RN1QKB1R b KQkq - 0 9)
The first main crossroads – Ehlvest had just played 9.Be3 (9.Bb2 is the other main branch, and probably slightly more popular overall), and Black has a number of choices. I was already a bit surprised by the variation, and without anything special prepared, I stuck to whatever I had chosen when I first picked the French up in 1999.
The game thus continued 9…g6!? 10.Bd3 Nxe3 11.fxe3 Bh6 12.Qe2. White already has more central and queenside space, and I think his general plan is to strangle the life out of Black’s position. To this end, he has two main ideas he’d like to execute: (1) transfer a knight to c5 (either via d2-b3 or c3-a4) and (2) clamp down on Black’s kingside with g2-g4-g5 to prevent the freeing …f6 break. Since Black can’t actually stop g2-g4, it makes sense to focus on the knight transfer to c5.
Thus, I played 12…Bd7. Now if 13.Nc3, Black can play 13…Ne7. A possible continuation would be 14.0-0 Rc8 15.Rac1 Nf5 16.Bxf5 gxf5 17.Rc2 Qa6!, with a clear advantage for Black, as in Wall-Lputian, 1999.
With one path clearly taken away, Ehlvest played 13.Nbd2, angling to get there via b3. I can’t directly stop Nb3, but I can make it more difficult for him to plant a knight on c5, so I played 13…a5!. Interestingly enough, Ehlvest had this position later in the same tournament, against a much lower rated player, and his opponent played 13…Ne7?. After 14.Nb3 a6 15.Nc5 Bb5 16.Bxb5 axb5 17.g4, White had a huge bind and won pretty easily. Back to my game, after 14.b5 Ne7, White has a big choice:
r3k2r/1p1bnp1p/1q2p1pb/pP1pP3/3P4/P2BPN2/3NQ1PP/R3K2R w KQkq - 0 15)
Now White is at a crossroads. Either he allows Black to fix the a- and b-pawns with …a4 (also taking away any Nb3-c5 plans), or he can play a4 himself but allow Black’s knight to get to f5. If the knight gets to f5, then White will have to exchange there, and after …exf5, White won’t be able to set up a kingside bind with g4-g5. Thus, Black can slowly prepare …f7-f6 (say with …0-0 and …Bg7 and …Re8), when the opening of the e-file makes the e3-pawn a liability. The c5-square is also not a guarantee here, as Black can play …Qa7 and …b6 to try and cover that weakness as well.
More specifically, lines like:
(1) 15.a4 Nf5 16.Bxf5 (16.g4? Nxe3 17.g5 Bxg5! 18.Nxg5 Qxd4 wins: Black threatens …Qxa1+ and …Qh4+) exf5 17.Nb3 Rc8 (Black would often be happy to give up an exchange on c5 too),
(2) 15.a4 Rc8!? 16.Nb3 Qa7, and
(3) 15.Nb3 Nf5 16.g4 a4 (this time d4 is doubly protected, so 16…Nxe3 doesn’t work) 17.Nc5 Qa5+ 18.Qd2? Nxd4! 19.Nxd4 Bxe3 is much better for Black, as the knights are skewered) are not particularly clear.
However, given that Ehlvest repeated this line a couple days later (no immediate transmission of the games back in the stone age, so it’s unlikely his opponent would have had this game), maybe he was angling for one of these lines?
Ehlvest decided to clamp down on f5 and the kingside with 15.g4, but that was probably a mistake as after 15…a4!, I think the tradeoff favors Black. The knight will never get to c5 now, and his a- and b-pawns are fixed as weaknesses. The kingside bind is nice, but it doesn’t do much in and of itself, while now it’s not clear what his minor pieces will be doing. Black, meanwhile, can work away on the c-file, and many endgames will favor him because of the weak pawns that can be attacked by the pair of bishops (from f8 and d7, for example, while the knight might be rerouted to a7 via c8).
After 16.0-0 0-0 17.h4!? Rac8 18.h5, we reached the position in the diagram below:
2r2rk1/1p1bnp1p/1q2p1pb/1P1pP2P/p2P2P1/P2BPN2/3NQ3/R4RK1 b - - 0 18)
I played 18…gxh5! here, really surprising Jaan. It’s an extremely concrete move and I think it only now occurred to him that he wasn’t really attacking on the kingside with his pawn advances.
If White plays 19.gxh5, then 19…Nf5 says “thank you very much” for moving the g4-pawn away. After 20.Bxf5 exf5, Black’s pawns look bad, but the b5-pawn is hanging, the e3-pawn will be weak (after …Rc3), and Black is more likely to do some attacking with …Kh8, …Rg8, …f4, etc.
To this end, Jaan played 19.g5 Bg7 20.Kf2, supporting e3 and clearing the h-file for a heavy piece assault. A nice plan, but it’s a bit slow … After 20…Rc3! 21.Rh1 Nf5, Black is the one making the immediate threats:
5rk1/1p1b1pbp/1q2p3/1P1pPnPp/p2P4/P1rBPN2/3NQK2/R6R w - - 0 22)
White can’t play 22.Rxh5 because of 22…Bxb5!. Then after 23.Bxb5 Rxe3 24.Qf1, Black has 3 main choices, all of which lead to a clear advantage: 24…Nxd4, 24…Bxe5, and 24…Rc3. Of the three, 24…Nxd4 is what I had planned during the game, while 24…Rc3 is the weakest and 24…Bxe3 is the most computer-like (and strongest).
After 24…Nxd4, a couple possibilities would be: 25.Kxe3 Nxf5+ 26.Kf4 Bxe5+!! leads to checkmate, while 25.Bd3 Nf5 sets up all sorts of nasty discoveries that White can’t escape.
Ehlvest tried to cover the soft e3-pawn with 22.Ne1 instead, but then I simply played 22…h4!, and the h-file remains closed! The threat of …Ng3 is almost incidental at this point, as Black is already winning. He tried 23.Qh5, but that left e3-a bit weak – I responded with 23…Bxe5!.
Taking the bishop allows 24…Qxe3+ 25.Kg2 Qxd2+ 26.Be2 Re3, for example, when White’s position has completely collapsed.
The game finished: 24.Kg2 Bg3 (24…Nxe3+ would also have been good enough) 25.Ndf3 Nxe3+ 26.Kg1 Nf5 (keeping things simple) 27.Rd1 e5!, and White finally threw in the towel.
5rk1/1p1b1p1p/1q6/1P1ppnPQ/p2P3p/P1rB1Nb1/8/3RN1KR w - - 0 28)
White is completely dominated. Something like 28.Bxf5 Bxf5 29.Nxh4 Be4 30.N1g2 Rc1! is one picturesque finish. After two bad experiences trying to deal with the French, Jaan has never repeated 1.e4 against me, even in blitz. Sadly, he’s done better with his alternative approaches!