John Timm vs. Dr. Hans-Otto Herr [B19]
XIX World 3/4 Finals [Notes by Timm]
1.e4c62.d4d53.Nd2dxe44.Nxe4Bf55.Ng3Bg66.h4h67.Nf3Nf6?!When I was learning to play tournament chess in the 1960's, and for many years thereafter, this move was regarded as a crude positional blunder, and the defensive 7 ... Nd7 was regarded as the only good move. A 1980 booklet, "Understanding the Caro-Kann Defense", written by various OTB GMs, explained why: "accuracy is essential .... [After other moves], White [by Ne5] gains too much time by forcing Black's bishop to move again." There were very few games with 7 ... Nf6 before 1995. But by 1998 (this game started in 1999), 7 ... Nf6 was deemed rock-solid and was being played by many top OTB GMs. But I still think it's dubious.
8.Ne5Bh79.Bd3One reason why 7 ... Nf6 became popular in 1995-1998 was that, in this position, White kept trying to gain an advantage by 9 Bc4, and Black kept equalizing by 9 ... e6. By late 1998, it was clear that top OTB GMs were mostly switching to 9 Bd3, but were still not getting any advantage.
9...Bxd310.Qxd3e611.Bd2Nbd712.f4Of course the Knight can't sensibly stay on e5 without this move, but f4 has both advantages (anchoring the Knight on e5) and drawbacks (creating a huge hole on g4, which Black promptly exploits). It is entirely characteristic of modern chess that Black is willing to ignore general principles ("don't lose time by moving the same piece multiple times in the opening") in return for counter-chances in sharp positions (if White didn't have the "free" tempo Ne5, White wouldn't be playing f4 and creating a hole on g4).
16.Qf3 [ Not 16.g4Bxe5!
which can get complicated but favors Black. Now White is threatening g4 again.]
16...h517.g4! [ In March 1999, Anand reached this position against Ivanchuk at Linares and played 17.Rhg1c5!18.g4Bxe519.dxe5 ( 19.fxe5is more complicated but doesn't seem to work either) 19...Nxg420.Ng3f521.exf6Ndxf622.Nxh5Nxh523.Qxg4Rf524.Qe2Qf7
and Black was at least equal. Commentators at the time suggested that 17 g4 was not sound due to either 17 ... hg4: 18 N:g4 N:g4 19 Q:g4 f5 20 Qg2 Nf6 or 17 ... B:e5 18 fe5: N:g4 19 Ng3 g6, in both cases with advantage to Black. However, I didn't believe any of that, and, apparently, neither did Ilescas Cordoba, who, in the only other game I know of that reached this position, played 17 g4! against Karpov at Dos Hermanas in April 1999, won brilliantly, and subsequently published his annotations to the game in virtually every chess magazine in the world. As the current game started in June 1999, I had to assume that Dr. Herr would know about the Ilescas-Karpov game and Ilescas' annotations.]
18.fxe5Nxg419.Nf4!Ilescas' move and much better than the previously-recommended 19 Ng3, which may look more logical because it gains the same tempo on the h-pawn and in addition allows Ne4 targeting weak squares on f6, d6 and g5. But Ilescas figured out the concrete tactical reasons why Nf4 is stonger: f5 is prevented (N:e6) and the g-file remains half-open, allowing a quick Rg1-:g4, with a winning attack in many lines.
20.dxe5Qxe5At this point I decided to deviate from Ilescas-Karpov, even though White had won. There were two reasons for this: (1) I was deeply suspicious of Ilescas' move, and couldn't believe it was the best; and (2) Ilescas himself had suggested an improvement on Karpov's play, resulting in positions which were at most only slightly better for White (see below). I had no doubt my opponent would know about the improvement, I felt that the board position was very favorable for White, I could find no antidote to Ilescas' improvement for Black, and so I decided to play my own move.
21.Rhg1!Ilescas had played [ 21.Bc1!
(his mark), explaining that the basic idea is to protect the first rank and Nf4, trade Roooks, and attack with Q:d1-d8+, possibly in conjunction with Nd3-c5-d7-f8+. This is a deep, non-obvious, and grandmasterly plan, and moreover it was strong enough to defeat one of the best players in the world. But Ilescas himself pointed out a basic flaw in the idea, which is that Black pushes e5-e4-e3 and has just enough play to scrape a draw. Ilescas-Karpov continued 21...Qf522.Rxd8Rxd823.Rd1Rxd124.Qxd1e525.Qd8+Kh726.Nd3e427.Nf4Ne5? ( 27...e3!28.Ne2 ( 28.Qd3Qxd329.cxd3g530.hxg5h431.Kc2Nf232.Bxe3h333.Ne2h234.Ng3h1Q35.Nxh1Nxh136.Bxa7Kg637.Be3Ng338.Kc3Nf539.Bd2f640.gxf6Kxf6=as given by Ilescas) ) 28.Qe7
White has a substantial advantage. Ilescas pointed out 27 ... e3!, after which his only suggestion to get an advantage for White is 28 Ne2 with a slight advantage. However, that position looks equal to me. The long-term strategic point of 21 Rhg1 is that the transition to an endgame with piece against three pawns isn't yet winning, so White keeps the Rooks on and plays a middlegame in which White, with an extra piece, can make progress on the Kingside whereas Black can't push his extra pawns too aggressively without weakening his King position.]
21...Qf5Black's position is very difficult to play. ... f5 is desirable (to anchor the Knight on g4), but it can't be played due to the weaknesses on e6, g6 and h5. 21 ... g6 may be relatively best, but it weakens the long diagonal and, even worse, encourages R:g4, Q:g4 and h5 (the Queen can't defend g4 indefinitely because it can be chased away by a Rook). Without making any obvious mistakes, Black soon lands in a lost position.
22.Qg3Rd4This is a perfectly normal and reasonable move. But Black's position is so desperate that, in addition to 22 ... g6, Black could consider 22 ... e5 23 N:h5 Q:h5 24 Q:g4, Q:g4 25 R:g4, f5. Black is in a depressing ending with only two pawns for a Bishop, but at least the connected passed e and f pawns are moving. As the game goes, Black loses a pawn anyway and the e and f pawns never get moving. [ 22...e523.Nxh5Qxh524.Qxg4Qxg425.Rxg4f5
]
23.Nd3Rfd824.Bc3R4d725.Rdf1Qd526.a3White has achieved everything he wanted from 21 Rhg1. White is improving his position move-by-move, Black's control of the Queen-file means nothing, and Black has nothing useful to do. I don't think it's an exaggeration to say that White is winning.
31.Rf2Qe432.h5The same theme. White is winning, and continues to make incremental improvements to his position. He doesn't want just any winning Bishop versus two pawns ending at this point, he wants to win still more material and/or enter an endgame where he has a clear plan for dealing with the e and f pawns.
