Not Letting Chess Engines Distract You From What's Important
NN (sorry, I misplaced his name!) asked:
You describe the 1931 Mitchel - Nimzowitsch game, illustrated at Diagram 138 on page 217 of the Third Edition, as one of your favorite positions. You mention that the “greedy” variation 10.Bxa7 is to be avoided on account of Black’s response 10…Nf4 11.Rxd5 Rxd5, followed by 12…Ne2+ 13.Kh1 Qxh2+ 14.Kxh2 Rh5 mate. I followed this analysis through with the help of ChessBase and note that 12.Re1 appears to be a complete answer to this mating threat, and therefore makes sound the capture of the a7 pawn. I recognize that computer analysis is sometimes blind to subtleties of the position, but I have considered the variation carefully and can see no flaw with the answer 12.Re1. White could also play 12.Bb8 to discourage the Qxh2+, but this is not as good as 12.Re1.
First off, let’s look at the example in question:
In this position White played 24.Rfd1, but it turns out that 24.Bxa7! was indeed best. Then 24…Nf4 25.Rxd5 Rxd5 26.Be3?? Ne2+ 27.Kh1 Qxh2+! 28.Kxh2 Rh5 is mate. However, you pointed out that White could successfully defend with 26.Re1. True enough!
The book was written in 1993, and as is so often the case with pre-Fritz creations, analysis was often shoddy. Back then, my only concern was to make every line as instructive as possible, so I leapt at the chance to show the pretty mating pattern that occurs after 26.Be3. But, times changed. Nowadays an author can’t get away with this because everyone owns a chess engine, so he needs to mix instructive intent with pristine analysis.
But, I have a question for you, and for everyone else whose engine notices an analytical error in any of the classic books (they all have them): were you so busy watching the random numbers produced by Fritz/Rybka/Shredder (+0.21, +0.035, -0.19, etc.), and so busy staring at the cascade of moves the engine threw your way, that you “forgot” to notice the game’s beauty and/or instructive value?
I’ve talked to many people who notice an error in an Alekhine book, or in mine, or in one of Kasparov’s books, and it’s not unusual to find that the engine distracted them from the game’s point. And that’s a terrible shame. In the student’s zeal to find flaws in analysis, he has missed the game's lessons (which are usually concept driven) and thus missed a chance to learn something. The simple fact is that concepts and patterns will make you better, but raw analysis that flows by like a tidal wave will leave you just where you started – dreaming of getting better and wondering why it’s not happening.
A few months ago I was going through various pre-1800 games (it’s fun!) and noticed a Philidor contest that showcased his amazing understanding of pawns – it was a work of art. Everything he did was dripping in deep instructive content, yet I also noticed that most of the analysis was completely wrong. But, who cares? It would be nice if everything Philidor did was analytically perfect (and if the analysts were also perfect), but if you clutch onto the old “Only flawless games are worthwhile” paradigm, then you’ll be missing out on many (in fact, most) of the greatest works of chess art.
I want to thank Mr. NN for allowing me to point out this “lack of improvement due to computer worship” mentality that’s infecting so many amateurs nowadays (of course, Mr. NN might be improving by leaps and bounds, and he might well have loved this Nimzowitsch game with all his heart – I’m not directing any of this at him personally). I also want to thank him for reminding me that chess.com readers will learn a lot from this example (a riveting Bishop vs. Knight battle), and hopefully some will find it as beautiful (and instructive) as I did (and still do). I’ll use a lot of the prose from the old 3rd edition of How to Reassess Your Chess.
W. Mitchel - A. Nimzowitsch, Bern 1931
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 c5 5.Nge2 cxd4 6.exd4 d5 7.a3 Bxc3+ 8.Nxc3 dxc4 9.Bxc4 0-0 10.0-0 Nc6 11.d5 exd5 12.Nxd5 Be6 13.Nxf6+ Qxf6 14.Bxe6 Qxe6
I must confess that this is one of my favorite positions. Though it appears to be a hopelessly boring game that is heading for a draw, it is actually a fantastic illustration of a battle for one little square (revolving around the minor pieces) waged by almost every piece from both armies.
Let’s break the position down in an attempt to understand it: with the pawn structures almost identical, many pieces traded off, and no real weakness in either camp, things would seem to be pretty equal. To see if this assessment is true though, we must carefully weigh the only two imbalances that exist here – namely black’s temporary lead in development and the very important difference of Bishop vs. Knight.
As stated earlier in the book, if a Bishop vs. Knight is the only imbalance or the major imbalance in a position, then you must strive to make your individual piece superior to the one your opponent has. In this case we have a wide-open position – the type of thing that usually favors a Bishop. If white’s bishop can get to c3, it will dominate the Knight and would then show its superiority over the unfortunate horse. Clearly, Black cannot allow this! Additionally, Black must find an advanced support point for his Knight. How is this possible? No central pawns exist that can be used to anchor the beast. This is where willpower comes into play. Conceptually, Black knows what he needs to do. Now he must insist on it becoming a reality! Black’s plan is as follows:
His fantasy square for his Knight is d3. It is clear that the Knight will be extremely powerful on this post; but if he is to keep it there he will need the aid of his army, since no pawns can help. To accomplish this, Black will need to place his Rooks on the d-file and move the Knight from c6 to e5 to d3. This plan will be aided by black’s lead in development. Black knows that a development lead is only temporary and he wishes to convert it into a permanent advantage, namely a superior Knight versus a less active Bishop. This leaves us with the unusual case of Black playing to win neither a pawn nor space. The apple of his eye in this case is a seemingly useless, disembodied square!
White’s first move, a big mistake, allows Black to bring his plans to fruition.
Making immediate use of his development lead. Since White can neither move his Queen nor take black’s, he is forced to place his Bishop on d2 and into an unfortunate pin.
Black devotes his whole existence to the control of d3. A big mistake would be 16…Qd5? 17.Bc3 when white’s minor piece would be the one to claim dominance.
Heading for the Promised Land.
A blunder would be 18.Rc5?? Nf3+ winning the loose Rook.
Forced since 19.Re2? Nxc1 20.Qxf5 Nxe2+ 21.Kf1 Rxd2 wins easily for Black.
19…Qg4 20.Rcd1 Qe2
Black’s Knight is fantastic! It blocks the d-file and also prevents White from placing Rooks on the e-file and c-file. However, great care must now be used. Black must not think that he has won the fight – the battle for d3 has only just begun!
Many players would be tempted to play something like 21…Rac8. This would be very pretty, but it would have nothing to do with black’s plan of dominating d3. By playing his Rook to d5 he prepares to give added (and much needed) support to d3. Also note how the d5-Rook can swing over to the kingside in some variations. Naturally a light square was chosen; it is always a good idea to avoid possible contact with white’s Bishop.
Under the present circumstances, White decides to play his Bishop to e3, where it prevents the Knight from eventually going to f4 with chances of a kingside attack.
22…Rad8 23.Rd2 Qh5
It turns out that 23…Qg4! was more accurate since 24.Bxa7? loses to 24…Nf4 25.f3 Qg5 26.Rc2 Rd2 27.Rxd2 Rxd2 28.Rf2 Nh3+ and Black wins. After 23…Qg4! 24.Rfd1 b6 25.f3 Qg6 we reach the exact same position as the game.
Missing his chance. He should have snapped on a7: 24.Bxa7! Nf4 25.Rxd5 Qxd5 (25…Rxd5? 26.Re1) 26.f3 Qd2 27.Bf2 Nd3 with approximately equal chances.
Black leaves his Knight hanging and instead guards his pawn! Has he gone mad? No, this is based on an X-Ray combination. If 25.Rxd3?? then 25….Qxd1+! 26.Qxd1 Rxd3 27.Qf1 Rd1 wins outright.
A very rare situation has arisen. Both sides are devoting all their forces to the capture of this one little d3-square! White knows that if the Knight can be made to move, then his Bishop will take over as the superior minor piece.
26.Qc2 h6 27.Kf1
The battle for d3 is reaching a climax. White has decided to bring his King to e2 and force the Knight to move away! However, White must first guard the pawn on g2 by g2-g3. This will obviously weaken his kingside pawn structure. What we are going to end up with is a case of “trading advantages”. Black is willing to give up his superior minor piece and move it away if White ends up with an insecure King.
26…Kh7 28.Qc3 R8d6
Black takes his time and calmly improves the position of his pieces.
29.b4 h5 30.g3 Qf5 31.Ke2
White has finally succeeded in chasing away the irksome Knight, but only at the cost of a weakened King position.
Not 32.Rxd5 Qxf3+ followed by 33…Rxd5.
32…Rxd2+ 33.Rxd2 Rc6!
Since black’s chances are now based on attacking the white King, it is logical for him to keep as many pieces on the board as possible.
Avoiding the blunder 34.Qxe5?? Re6.
Amazing! The minor pieces’ roles are reversed! Black’s Knight is now clearly inferior to white’s Bishop, but white’s King has no safe home to go to.
35…Qh3 36.Ke3 Qf1 37.Rd1 Qg2 38.Qd2 Qh3 39.Qd5 Rc2 40.Rd2
40.Qxf7 Qxh2 gives Black a winning attack.
40…Rc3+ 41.Rd3 Rc1 42.Qxf7 Qxh2 43.Qf5 Rc4
Threatening to mate by …Qg1+.
44.Rd1 Rc3+ 45.Kd4
The obvious 45.Rd3 loses to 45…Qg1+ 46.Ke4 Rxd3 47.Qxh5+ (47.Kxd3 Qb1+ wins the Queen) 47…Kg8 48.Kxd3 Qd1+ winning the Bishop.
45…Rxa3 46.Kd5 Rc3 47.Ke6 Qe2+ 48.Kf7 Rc7+, 0-1. It’s mate after 49.Bxc7 Qe7 mate.
A close study of this game will prove illuminating for a vast range of players. Don’t let your computer distract you from its enormous instructive content!