Thursday, November 11, 2010

Janowsky - Capablanca

I’ve chosen the game Janowsky vs. Capablanca played in the Rice Memorial Tournament, New York, 1916, to look at in this post. This game is quite interesting and somewhat controversial because of it’s opening and because of the way it has been annotated in the past. It also has some instructional value.

First, the opening. There was a big brouhaha on one forum recently centering around a statement made by GM Alex Yermolinsky when he stated there was no theory on this line. Technically there is, but it is what I call negative theory because it all goes to prove that 4…Bf5 is a bad move and as Yermo said, no GM would play it. There was a lot of nitpicking and hair splitting over Yermo’s comments…all useless drivel because his final conclusion, based on his GM’s understanding, has been confirmed. I’ll give the analysis of IM Tim Taylor on this opening so as not to have to include them in the game itself.

Notes by IM Timothy Taylor: … here’s an example where “opening knowledge” is paramount, where the game only lasts 21 moves, and more than that – the game was really decided on move 4! I recommend printing out this page and playing over Taylor's analysis.

Taylor - Ruden
U.S. Open, 1977 (first round)
Slav Defense
1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Nf3
This is a standard position in the Slav Defense, which has been reached literally thousands of times. The normal move is 4...dxc4 and then, for example, 5.a4 Bf5 6.e3 e6 7.Bxc4 Bb4 is a main line Slav. Also perfectly playable for Black are 4…e6, aiming for the Meran, or the currently fashionable 4…a6. Instead, my opponent plays …
4…Bf5?
And right away, I figure the game is over, and I’ve won it! Why so optimistic? Because I know this move is bad!This requires some explanation, and bear with me, as this is the most important note in this article. You may be asking me, “How can a quite reasonable looking developing move be ‘bad’?” Before I answer that, let me riff a bit on the difference between “bad” and “risky.”

Let’s say, in my next tournament, my first round game (Taylor-NN) starts out like this: 1.d4 d5 2.c4 e5!? I say to myself, the Albin Counter-Gambit! It’s quite a risky opening – but it’s played from time to time, with success, by high class Grandmasters like Morosevich and Nakamura. I’ve played it myself! Yes, Black takes a big chance, sacrificing a center pawn in the opening, but he also gets (you know my favorite word!) play!

You certainly can’t (and I certainly wouldn’t) consider playing the Albin to be a mistake. On the contrary, I would be doubly alert, knowing my opponent was out to attack me from move two!

Now let’s consider the played 4…Bf5. Yes, this move was played by Capablanca (against Janowski at the Rice Memorial tournament, 1916, game 30 in MY CHESS CAREER). The wily Cuban, unwilling to admit a gross blunder on move 4, starts his annotations on move 6 (!!) after Janowski has already missed his chance to refute Capa’s play! Janowski continued 5.Qb3 Qb6 6.Qxb6? and Capa used the now open queenside files to create one of his famous positional masterpieces – but his fourth move was still a blunder!

After Capa’s lucky escape, the writing appeared on the wall, in the game Johner-Nisson from 1920. There White played the correct 5.cxd5 cxd5 6.Qb3, and killed Black.

Check out some statistics: I looked this line up on Chessbase.com. The first seven games played with this line (White playing correctly, starting with Johner-Nilsson) give us six White wins, and Black scored one measly draw. Check out the present day: put in the same position after 4…Bf5, and see how players did against White opponents rated over 2450 – there’s that writing on the wall again: seven games, six white wins, one measly draw!! In other words, if Black plays 4…Bf5, he has a statistical chance of drawing about 7% of the time! Take a look at it from another angle: is this (4…Bf5) a rare but risky move that GMs try from time to time? Let’s go back to Chessbase – and the answer is …NO!!

If you put in a minimum rating for Black of 2450 (more or less GM strength), and ask it to find someone, anyone, who played 4…Bf5 in this position, you get exactly one game in all of modern historical cyber records, namely Ligterink-Vasquez. Sad to say, the unfortunate Mr. Vasquez was hammered as follows: 5.cxd5 cxd5 6.Qb3 Bc8 (making amends but too late!) 7.Bf4 a6 8.Ne5 and White had a big advantage and won easily. I gather that Vasquez never repeated the experiment!

My point is this: there are risky but playable moves that give you a real chance to “mix it up” and create problems for your opponent. Even very strong players will take these risks from time to time. And then there are moves like 4…Bf5, which lead by force to “playless” positions in which only one side, in this case White, has everything, and Black can only suffer and pray for a draw. A further point: Why do I know this? Even in 1977, well before computers and Chessbase, I knew exactly how bad my opponent’s move was – and the reason I knew was simple: I was the stronger player! I had done my opening analysis; I had studied books on the Slav Defense; I had read the footnotes as to why 4…Bf5 was bad, and I remembered them. And so I, as GM Chris Ward would say later, stepped into “move order punishment mode” and took over the game.

5.cxd5
As noted above, 5.Qb3 is inaccurate and allows Black to stay in the game with Capablanca’s 5...Qb6, though even then White could keep an edge with 6.c5 instead of Janowski’s 6.Qxb6?!.
5...cxd5
The alternative 5...Nxd5 is different but not better. White’s advantage after 5…Nxd5 is clear for two reasons: the obvious one is the center pawn majority. The second is a gain in time: Note that the structure is now the same as a main line Slav – but White is an important tempo up, as, unlike in the line 4…dxc4 5.a4 given above (which has been proven to be playable for Black), here White has dispensed with the a pawn move, and so has an extra development tempo. This fact is very important in some of the tactical variations that follow.

After 5...Nxd5 the simplest is 6.e3 (Also 6.Nd2, immediately enforcing e4, is good for White, but not 6.Qb3 Qb6 and Black might sneak back into Janowski-Capablanca!) and then 6...e6 7.Bd3. Simply put, White is going to take over the center, and Black has no good way to stop him—visualize a few variations:

A. 7…Bxd3 (Simplest, Black just rolls over and plays dead.) 8.Qxd3 Be7 9.0–0 0–0 10.e4, ±, Moskalenko-Tare, Agios Nikolaos, 1997.
B. 7...Bb4 (Black flails at c3 to no effect) 8.0–0 (White is too cool to be concerned) 8…Nxc3 (for 8…Bxc3 see B1) 9.bxc3 Bxc3 10.Rb1 b6 11.Bxf5 exf5 12.Qc2, ±.
B1. 8...Bxc3 9.Bxf5 exf5 (9...Bxb2 10.Bxb2 exf5 11.Ba3, ±, as the Black King is caught in the center) 10.bxc3 Nxc3 11.Qb3 Nb5 [11...Ne2+ 12.Kh1 Nxc1 – avoiding 12...Qb6? 13.Bb2 – 13.Qxb7, ±] 12.a4 Nd6 13.Ba3, ±, for example 13...b6 14.a5 Ne4 15.Rfc1 b5 16.d5 and Black is getting killed.
C. 7...Qa5 (Another flailing swing at c3, equally frustrated!) 8.Bxf5 Nxc3 [8...exf5 9.0–0 Be7 (9...Nxc3 10.bxc3 Qxc3 11.Rb1 b6 12.Ne5 Be7 13.Bd2 Qa3 14.Qc2±) 10.Qc2 g6 11.e4 and again White’s central and development advantages prove a clear White plus – Black is not going to get out of this alive!] 9.bxc3 Qxf5 (9...Qxc3+?? 10.Bd2 wins for White) 10.Qb3 b6 11.0–0 Be7 12.Ba3 and Black has a “play-less” position, while White has numerous possibilities, especially of advancing the center pawn majority, e.g. 12...Qf6 13.e4, ±.

All these variations show the simple strength of White’s position: a central pawn majority and a lead in development.

6.Qb3
This move, in this sequence, is the exact reason 4…Bf5 is bad. The b7 pawn is under attack, and there is no good way to defend it, given that the d-pawn is also under heavy pressure.

6…Qb6
There’s no good advice; this unsound pawn sacrifice is no worse than any other move. We have already seen the retrograde 6…Bc8, dispatched by Ligterink above.
Other alternatives:
A. 6...b6 (weakens the a4-e8 diagonal) 7.e4


(Sharp and ultimately crushing, but even 7.Bg5!? Nbd7 [7...e6 8.e4 dxe4 9.Bb5+ is game over] 8.Nxd5, ±, is good enough) 7...dxe4

Seventh move alternatives are:
a) 7...Nxe4 8.Nxe4 Bxe4 (8...dxe4 9.Ne5 Be6 10.Bb5+ Nd7 11.d5 Bf5 12.g4 a6 13.Bc6 f6 14.gxf5 fxe5 15.Bxa8 and Black is lost) 9.Bb5+ Nd7 10.Ne5 Bf5 11.Qxd5 Be6 12.Bxd7+ when Black must resign;
b) 7...Bxe4 8.Bb5+ Nbd7 (8...Nfd7 9.Nxe4 dxe4 10.Ng5 e6 11.d5 is crushing) 9.Nxe4 Nxe4 (9...dxe4 10.Ne5 e6 11.Bg5 is also hopeless for Black) 10.Ne5 Nef6 11.Qf3 and the threat of 12.Bxd7+ Nxd7 13.Qxf7 mate is more than Black can handle.

Since neither of these helps, let’s go back to the main line after 7...dxe4: White continues with 8.Ne5 Be6 (8...e6 9.Bb5+ Nfd7 10.g4 wins material) 9.d5 Bd7 (9...Bxd5 10.Bb5+ Nbd7 11.Nxd5) 10.Bg5 h6 11.d6 e6 12.Nb5, ±, e.g. 12...Na6 13.Bxf6 gxf6 (13...Qxf6 14.Nxd7 Kxd7 15.Nxa7) 14.Nxd7 Kxd7 (14...Qxd7 15.Qa4) 15.Qc3 Nc5 (15...Rc8 16.Qa3) 16.b4 Nd3+ 17.Bxd3 exd3 18.Qxd3 a6 19.Rc1 with a winning attack.
B. 6...Nbd7 7.Nxd5, ±;
C. 6...e6 (the surprising choice of my materialistic Mr. Fritz, who evidently saw blown circuits in all straight defenses!) 7.Qxb7 Nbd7 8.Bf4 and even the computer admits he has nothing for the pawn.

Now back to the game where even some very prominent annotators annotated by result and missed some interesting alternatives!
7.Nxd5
White noshes on a center pawn
7...Qxb3
The books give the more frequently played 7...Nxd5 8.Qxd5 e6 9.Qb3 Qxb3 10.axb3 Nc6 (10...Bc2 11.Bd2 [11.Nd2!? is possible, while 11.e3 Bxb3 12.Bb5+ gives White a simple plus] 11...Bxb3 12.e4, ±, Torre-Gotthilf, Moscow 1925) 11.e3 Bd6 (11...Nb4 12.Bb5+ Kd8 13.Ke2 a6 14.Bd2 Nc2 15.Ra4, ±) 12.Bd2 and White is better, e.g. 12… Ke7 13.Bc4 Be4 14.Ke2 Rac8 15.Rhc1 a6 16.g3 h6 17.Bc3 Bd5 18.Nd2 f5 19.f3 b5 20.Bd3 b4 21.e4 and the center pawns prove their worth: White’s advantage is close to decisive.
8.Nxf6+ gxf6
An amusing variation is 8...exf6 9.axb3 Nc6 10.e3 Bd6 11.Bc4 a6 12.0–0 Ke7 13.e4 Bxe4 14.Re1 f5 15.Ng5 Nxd4 16.Nxf7 Nc2 17.Bg5+ Kd7 18.Rxe4 fxe4 19.Rd1 and wins; all this is not forced, but shows White has plenty of play in addition to the extra pawn.
9.axb3 Nc6 10.e3 Nb4
Or 10...Rg8 11.Bd2 a6 12.d5, ±.
11.Bb5+ Kd8 12.0–0 a6 13.Bd2 e6
Here and on the next move Black has the opportunity to stir up a little trouble with 13...Nc2. Of course White stands clearly better after 14.Ra5 Be4 15.Bc4 but at least Black’s light-squared Bishop does not get shut out, and he can hope for some activity with his Knight.
14.Rfc1 h5
Still, 14...Nc2 is a better try.
15.Ne1! and White went on to win.

And now on to the game:

No comments:

Post a Comment