Candidates Quarterfinals, Day 3: Three More Draws, Plus a Gelfand Win
There's no want of effort; in fact, it's because the players are really keyed up that it's so difficult to grind out a win. Only Gelfand managed to pull it off today, in what at least superficially looks like a pretty incredible game. Mamedyarov played 6.Bc4 against Gelfand's Najdorf and threw the kitchen sink at Black's king. Old wisdom says that the way to meet an attack on the flank is by counter-attacking in the center, and Gelfand did that; still, putting the general wisdom into practice is not as easy as it looks! Gelfand took material when it was offered, and didn't hesitate to sac it back when needed. Altogether it led to a peculiar material imbalance, and by the end of the game (on move 39) Gelfand had six(!) pawns (over)compensating for a rook.
The other games were less dramatic. Radjabov and Kramnik once again produced the least interesting battle, but again, it's not from want of effort. Radjabov tested Kramnik's Lasker Defense in a more mainline system. It wasn't that testing, though: Kramnik simply followed the very nice idea displayed in van Wely-Kasimdzhanov, Khanty-Mansiysk 2010, found eight further moves (quite possibly already done at home, beforehand), and then Rajdabov forced a perpetual.
Kamsky-Topalov was another 6.a4 Najdorf, but rather than playing 6...Nc6 followed by 7...e6 Topalov opted for a more classical Najdorf approach with 6...e5. After 7.Nf3 Be7 8.Bg5 White had a tempo-up version of the 7.Be3 e5 8.Nf3 system we saw in game one of the Mamedyarov-Gelfand match, but Topalov never seemed to be in any serious danger.
That cannot be said for Grischuk, who also drew his game with Black. Aronian was pressing for a long time in a Russian Gruenfeld, and eventually won a pawn. Unfortunately, that only occurred in a rook ending, and as everyone "knows", all rook endings are drawn.
With one classical game left to play, all four matches are still undecided. Kamsky and Gelfand are both up a game, and will advance to the semis with any kind of positive result. Kramnik-Radjabov and Grischuk-Aronian are tied.
Games, with my comments, will be posted later today - stay tuned!
Refuting the 4.Bf4 Variation of the Exchange French
On the ChessCafe website this week, there's an article by Abby Marshall presenting ("advocating" might be a bit too strong) the line 1.e4 e6 2.d4 d5 3.exd5 exd5 4.Bf4 as an interesting try for White. The basic idea is to follow up with Qd2, Nc3, Bd3, Nge2, 0-0-0 (in some order or other) followed by a kingside advance with h3 or f3 followed by g4, aiming to attack Black's king. Granted, practically no one plays this way, but why shouldn't White be able to torture Black with a variation that can be learned in anywhere from 30 seconds to ten minutes, depending on how rigorous one wants to be?
Still, I was in a contrarian mood this morning, and so after five-ten minutes of rigorous analysis at the breakfast table this morning, I refuted the aforementioned attacking plan:
1.e4 e6 2.d4 d5 3.exd5 exd5 4.Bf4
Now that we know White's plan, let's see if we can thwart it.
4...Nc6!
This is already unpleasant for White. 5.Bd3 hangs the d-pawn, and the various ways of pre-defending the pawn are awkward. If 5.c3, the knight can't go there and the queenside castled position will be loose. If 5.Nf3, then the knight is on the wrong square and 5...Bg4 is good for Black. 5.Nge2 locks in the Bf1, so perhaps the best move is
5.h3*
This is a move White wants to make anyway, helping restrict Black's light-squared bishop, giving the Bf4 a cubbyhole in some lines and preparing a subsequent Nf3.
5...Bf5
Preventing Bd3, at least if White has any dreams of setting up an attack. If White has any ambition, it's time for
6.g4(?),
but this is busted by
6...Be4!
The principled try now is
7.f3,
but after
7...Qh4+ 8.Ke2 Bg6** 9.Bxc7 Kd7!
the wrath of heck is about to fall on White's head in a mighty way.
10.Bf4 Re8+ 11.Be3 Bd6
Threatening ...Bf4. How's that White attacking plan faring, anyway?
12.Kd2 Rxe3! 13.Kxe3 Nh6! 14.a3
Not inspiring, but hoping to run the king to the queenside without getting whacked by a Black piece on b4. But now the end is in sight:
14...Bf4+! 15.Kxf4 Nf5!
Threatening 16...Qg3#
16.Ne2 Qh6+ 17.g5 Qh4#
A compelling refutation, don't you think?
The foregoing isn't meant seriously as an analysis, but it's not a complete joke, either. 4.Bf4 isn't a terrible move and there is a plan, but it's easy to spin a narrative where Victim (always the other guy in these stories) reads Hero's script (that's our guy, who will of course be us once we try the line), follows it, and gets butchered like a pig. But aren't there reasons why 4.Bf4 is almost never played? It can't seriously be the case that the only two realistic options are that Black gets smashed or scrapes out a draw if he plays like a genius, but one might get that impression from articles in this genre. Accordingly, the line above serves as a counter-narrative, showing just how easy it is to reverse the roles with just a little ingenuity.
My point isn't to condemn 4.Bf4 (though I think it's pretty harmless) or that one should always avoid funny sidelines. Rather, it's a reminder of the old saw that if something seems too good to be true, it probably is. We all make mistakes and there's always more to know, but it beggars belief to think that this variation could be the magic key to unlock the French. It has its drop of poison, but if one then thinks prophylactically and asks the question, "How can I interfere with White's plan?", one will find reasonable counter-plans that help return 4.Bf4 to retirement.
Relatedly, this is not meant as a criticism of Marshall, either, except perhaps in one respect. First the praise: it is an interesting idea that she brings to our attention - a clear point in her favor. Further, she does not claim that White gets an advantage - more points there for objectivity. On the other hand, once she has committed herself to the narrative that this variation might be a good way (and not just a practical or interesting way) to avoid mainstream theory, then objectivity can start slipping away, in two directions.
First, there's the danger of at least implicitly suggesting to the reader that White is for choice. She doesn't explicitly say this, but the reader could certainly get this impression. The article repeatedly suggests that the line is good - there are encouraging phrases ("could be a good idea", "could turn venomous", etc.) and good results for White (White wins both main games and the majority of the decisive games in the notes). Of course it's not as effective from a marketing perspective to present 4.Bf4 as a little-known move that gives White nothing against any of several reasonable plans but could work against an unprepared opponent. Not as effective...but true..
Second, and this is at least as important, one generally doesn't try as hard to find good moves for the "opponent" once one has identified with one side or the other. The approach (4...Nc6, with the idea of "zugzwanging" White out of the basic attacking plan) I came up with after a few moments at the kitchen table, half awake and eating breakfast is entirely reasonable, has been played before, gives Black a serious plus score in the database and is pretty easy to find for a player of Marshall's caliber - if only she looks for it!
So I'll reiterate a point I've made on this blog many times before. Dear readers, do not believe in get-rich quick schemes, not even in chess. You are more than welcome to play whatever you want to: chess if supposed to be fun and no one appointed me Chess Openings Pope. (And if they did my first act would be to excommunicate myself.) But please bear in mind that there's a good reason masters and grandmasters play 3.Nc3, 3.Nd2 and 3.e5 against the French the overwhelming percentage of the time. 3.exd5 exd5 4.Bf4 doesn't lose, is fine for variety, may be a useful surprise weapon and so on, but both common sense and a bit of thought (with or without a computer) indicate that this cannot be a serious challenge to the French Defense.
* Marshall suggests 5.Qd2, but after 5...Nf6 (rather than 5...Bd6, which isn't bad either) Black renews the quasi-zugzwang. After 6.Nc3 Bb4 it's hard to see the attraction of White's position - White can maintain equality, but nothing more, after 7.Qe3+ Be6 or 7.f3 Qe7+.
** 8...Bxc2 is even better, and isn't bad on moves 10 or 11 either. Still, if my breakfasting self had noticed that simple tactic you'd have missed out on the fabulous finish.