The Chess Mind

...at least sometimes.

Alexei Shirov is not only one of the most exciting players around, he's flat-out one of the greatest players of our era, period. He can calculate, he can improvise, he can innovate - you name it, he can do it.

But although he can prepare with the best of them, he failed to do so - in spades - in his game yesterday with British GM Peter Wells. Wells, like many elite British players, is a fan of the Trompowsky (1.d4 Nf6 2.Bg5), and in fact recently published a widely acclaimed book on that very opening (Winning with the Trompowsky, Batsford 2003).

Wells, with White, defeated Shirov in 13 moves. Brilliant preparation? No. A clever novelty? No. A new trap? No again. Shirov followed a game between two well-known grandmasters - Tromp specialist Hodgson and Dutch GM van der Wiel for 11 moves (at which point he was probably lost), disimproved on van der Wiel's play on move 12, and resigned after Wells' 13th move.

Of course GMs are human and it's impossible even for them to keep up with everything, but the funny part is that Hodgson-van der Wiel is one of the main games in Wells' book, analyzed in significant detail!

So prep up for your opponents, and if you think you're going to play a GM who has written a book on an opening you'll play into, you might want to buy it beforehand. Conversely, Shirov's example shows that even legends of the game can go awry in slightly offbeat openings; in principle, there's no reason why a well-prepared 2000+ player couldn't have won the same game Wells did. (Granted, he'd have taken longer to resign.)

Be prepared!

(Click here for the game.)

In the FIDE World Championships in San Luis last year, the game Kasimjanov-Anand (round 4) started as follows

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 Ng4 7.Bg5 h6 8.Bh4 g5 9.Bg3 Bg7 10.h3 Ne5 11.Nf5 Bxf5 12.exf5 Nbc6 13.Nd5 e6 14.fxe6 fxe6 15.Ne3 O-O 16.Be2 Qe7 17.O-O Rad8

And now, I bring you Mr. Kasimjanov, courtesy of New in Chess 2005/8:

Now we've come to the position that I had studied in serious depth together with Darmen Sadvakasov and Daniel Fridman, who helped me in my preparation for this world championship. At some point Sadvakasov set up this position and said: 'This is an important position. It's important for the whole evaluation of the plan with 11.Nf5 after 6...Ng4.' To me it looked like a lost position for Black. White has two bishops, a better king and a better structure. But he insisted that things were very tricky and that once the knight moves to f4 (via g6), as happened in the game Dolmatov-Sakaev, Aeroflot Open 2003, things are completely unclear. For days and days we analyzed this position and we tried all kinds of ideas. I tried c3, Bd3, Bb1, Qc2 and mate, but it doesn't work quite so easily. I tried to play positionally with Re1, Bf1 and c4, but nothing was working - which was amazing given White's trumps. The black knight would come to f4, the other one to e5, he'd play ...b5 and ...Qb7, and suddenly I was under pressure with White!

And then there were some training games against Grischuk played on ICC - three-minute games. (Maybe they should be studied better. I have played quite a number of theoretically important games on ICC in the last few months, but who cares, who would watch them?) Well, actually Fridman played 'with my very close obsrevance'. I played c3, Bd3, Bb1 and then there came ...d5, ...d4, and he crushed me on the d-file. Then I was wondering if maybe the pawn should stay on f5, so that the knight cannot go to g6 and f4. An interesting idea, but Grischuk played ...Qe7, ...Rad8, and then when at some point I took on e6, he took back with the queen! And then the d-pawn and the f-pawn started rolling, and once again he tore me apart. Then, having finished this match we looked at each other and asked ourselves the question: what if the knight never gets to g6? We played the bishop to h5, and it was immediately clear to me that this was the refutation of the whole set-up. [Page 53]

18.Bh5!

A serious improvement compared to the Dolmatov-Sakaev game. Black has no useful moves. After executing this move I looked at Vishy and I knew that he was lost. He has a good sense of danger and after 18.Bh5 he knew that the danger had arrived. Black has no counterplay at all. [Page 54]

(To replay the game - and Dolmatov-Sakaev - with my quick, reasonably accurate but not particularly perspicacious commentary, done during the event, click here.)

I find this account extremely interesting on several levels.

First, his novelty, contrary to what one might be inclined to think about GM preparation, wasn't the product of switching on the computer and awaiting and interacting with its results. (That said, Fritz 9 has it at #1 or #2 just about immediately, and after all other normal moves Black's reply is 18...Ng6; Shredder 9 is less hip to the ...Ng6 idea, but it too is immediately fond of 18.Bh5. Yet neither program evaluates it as meaningfully better than 18.Re1 - if better at all - and until one realizes the power of ...Ng6-f4, it's far from obvious that Kasimjanov's move is anything special.)

Second, the sense that a position is somehow important is one that gets developed over time and is, I think, more prominent in stronger players. It took them a long time to sniff out the right approach, but Sadvakasov's original intuition was right.

Third, their great strength and insightful intuition notwithstanding, they didn't just find the right move immediately. Thus we too should be patient in our own searchings, and should also realize that GMs have to figure things out, just like the rest of us.

Fourth, this flash of insight confirms what researchers into human creativity have long known: these "aha!" moments aren't generally picked out of the air, as if by magic, but result from one steeping oneself in the ins and outs of a given problem. They knew where to look, they tried various approaches, and only then did the answer "suddenly" appear.

In short, you're much more likely to be creative if you work hard first!