Home • Learn to Play Chess • Improve Your Game • Chess History • Chess for Fun • Chess Blog
Home
•
Learn to Play Chess
•
Improve Your Game
•
Chess History
•
Chess for Fun
•
Chess Blog
| Kasparov - X3D Fritz, New York, 2003 | |||||||||||||||||||||||||||||||||||||||||
| Explaining positional play in a game between Kasparov and a computer is not an easy task. | |||||||||||||||||||||||||||||||||||||||||
|
(November 2003) In our two previous articles on Positional Play (see the link box at the bottom of this article) we looked at pawn structure and piece placement. These were introductions to the subject to present standard terminology: here are 'isolated doubled Pawns', while over here we have a 'Bishop pair'. During the course of any chess game positional themes come and go. The players exchange them in almost the same way that they exchange pieces. 'If I give up the Bishop pair, my opponent gets weak Pawns around the King. But then the Rooks can attack my King by doubling on that open g-file. I think my Knight can prevent that from happening by going to f5. First I'll have to maneuver it to e3 or g3. How can I do that without allowing an attack on my center.' And so on and so forth.The player who judges these positional exchanges more accurately wins more games. Since every single game is filled with positional themes, we could pick examples at random from just about any source. Let's look at the games from the man-machine match between Garry Kasparov and X3D Fritz, played at New York, November 2003. Explaining positional play in a game between Kasparov and a computer is not easy to do. Some experts might say it's impossible. First, computers are not known for their positional play. Their great strength comes not from weighing positional pros and cons, but from the rapid calculation of millions of variations. Second, Kasparov's play is so subtle and complex that it is often beyond explanation. One reason he is the world's highest rated player is because he judges positional factors better than most other players on the planet. Let's not let that stop us. On to the match!
The four game Kasparov - Fritz match was played less than a year after a six game match between Kasparov and Deep Junior, another world-class chess computer. The Kasparov - Junior match had seen the same venue at the New York Athletic club. A long-time contender in the World Computer Chess Championship (WCCC), Fritz had won the title only once, at the 8th World Championship, Hong Kong 1995. In 2001, it won the right to play World Champion Vladimir Kramnik in the much ballyhooed Brains in Bahrain man-machine contest, by drawing a 24-game match with Junior and then winning two playoff games. Fritz, one of the cornerstones of ChessBase's commercial success, was represented in the latest match by Jeroen van den Belt, Alex Kure, Frans Morsch, Mathias Feist, and Mr. ChessBase himself, Frederic Friedel. According to match sponsor X3D, it ran 'on an Intel Xeon server with four 2.8 GHz processors.' The winner would receive a golden trophy and the title of Man-Machine World Champion. For risking his reputation once more, Kasparov would receive $150.000 for the match, plus a bonus of $50.000 for winning or $25.000 for drawing. Game 1The first game started with the same opening as in game 1 of the Kasparov - Junior match... 1.Nf3 d5 2.c4 c6 3.d4 Nf6 (Slav Defense) 4.Nc3 e6 5.e3 (Semi-Slav Variation) Nbd7 6.Qc2 Bd6 7.g4 Bb4 ...where Junior had played 7...dxc4 in the earlier game. A few moves later the players reached the following position.
White's entire army is in action. Black has developed only three pieces and the Black King is still in the center. If Black tries to win a Pawn with 17...Ndxc5, then 18.Bb5+ Kf8 (18...Bd7 19.Bxc5 Nxc5 20.Nf5 Qf8 21.Nxg7+) would prevent the Black King from castling. Fritz played 17...O-O, sacrificing material to get the King into safety. After 18.Nf5 Qe5 19.c6 bxc6 20.Bxf8, Black could recapture the Bishop with the Knight. Most human players would do this automatically, but the computer played 20...Kxf8, reaching the following position.
White has an extra Rook against which Black has an extra Knight and Pawn. If White tries to win back the Pawn with 21.Qxc6, then Black gets good counterplay with 21...Nb6 22.Qc2 Bd7 or 22...Bb7. Kasparov played 21.Ng3 and in his notes to the game wrote, 'White should trade this dangerous Knight', meaning the Ne4. The game continued 21...Ndc5 22.Nxe4 Nxe4, where we see why Black played 20...Kxf8 instead of 20...Nxf8. The second Knight has replaced the first on e4. White eliminated the second Knight with 23.Bd3 Be6 24.Bxe4 dxe4, and after 25.Rf4 Bd5, the players reached the following position.
The dangerous Knights on e4 are gone, but the Bishop on d5 is a marvel. Note how it guards three Black Pawns, attacking the White Pawn on a2 at same time. Kasparov said, 'My position is better, but very hard to win.' The game was drawn on the 37th move. Kasparov 0.5 - Fritz 0.5 Game 2The second game saw Kasparov play the same solid defense that had given him so much trouble during his World Championship match with Kramnik, London 2000... 1.e4 e5 2.Nf3 Nc6 3.Bb5 (Ruy Lopez) Nf6 (Berlin Defense) ...Now Fritz avoided the main variations with 4.d3. The game continued 4...d6 5.c3 g6 and after a few more moves the players reached the following position.
White has advanced on the Queenside, Black on the Kingside. The Queenside advance threatens nothing in particular, while the Kingside advance is designed to expose White's King. On top of preparing ...g5, the Pawn on h6 prevents Nf3-g5-e6, where the Knight would be well placed in the absence of the light-squared Bishops. The structure of the White and Black Pawns on the Queenside make an unusual circular formation. If either player pushes the b- or c-Pawn, the game would be altered radically. The formation remained for the next 12 moves. The following fateful position was reached after Black's 31st move.
Here Fritz played 32.Qb4. The move has two objectives:
Kasparov momentarily overlooked the second objective and played 32...Rg7?, which lost immediately to 33.Rxe5. He struggled for a few moves, but resigned on the 39th move. A better move would have been 32...Rg8, neutralizing both of White's objectives from the previous move:
Also good was 32...Rc8. After either ...Rg8 or ...Rc8, Black would not have stood worse and would have had some winning chances. Kasparov noted that 32...Bf4 would have allowed the 'promising' exchange sacrifice 33.Rxf4!? exf4 34.Re6 Qg5 35.Nd2. Kasparov 0.5 - Fritz 1.5 Game 3By a different move order the opening repeated game 1... 1.Nf3 Nf6 2.c4 e6 3.Nc3 d5 4.d4 c6 5.e3 a6 ...where 5...Nbd7 had been played instead. Kasparov closed the position with 6.c5, leaving Fritz clueless. The following position was reached on move 14.
Here the most important positional element is the chain of Pawns spanning the b-, c-, d-, and e-files. The chain cuts the board in two, with White controlling the Kingside and Black the Queenside. There is only one good plan for both sides : attacking the enemy Pawn chain at its base. White will win the Pawn on a5 and advance the a-Pawn to a6. Black will advance the f-Pawn to f4. Instead of a sensible continuation like 14...Ne8 15.Rb1 f5 16.g3, the computer played 14...Bd6, hoping for 15.cxd6? Nxb6, winning the Queen. It probably took Kasparov less than two nanoseconds to see this threat and he continued 15.Rb1. Now the computer retreated 15...Be7, having handed its opponent two free moves. Kasparov won the a-Pawn with 16.Nxa5 and kept to the scripted plan with the advance of his own a-Pawn. Fritz continued to play without a plan and the game reached the following position.
The White King has reached safety on the Queenside. Should Black eventually find the right plan involving the f-Pawn, the White King is well protected by the Pawn chain and by its own pieces, Now Kasparov returned the a-Pawn with 29.a6 bxa6, aimed his pieces at the weak Pawn on a6, and reached the following position.
Note how little the Pawn structure has changed since the diagram after White's 14th move. The Fritz handlers resigned for their machine, although many observers would have enjoyed seeing Kasparov crush his electronic adversary. Kasparov 1.5 - Fritz 1.5 Game 4For the final game in this too-short match, with the score tied at a win, a loss, and a draw for each player, Fritz's operators switched from 1.e4 to 1.d4... 1.d4 d5 2.c4 dxc4 (Queen's Gambit Accepted) 3.Nf3 e6 4.e3 a6 5.Bxc4 c5 ...The game followed a known path until the following position was reached.
Kasparov could have continued 13...Nxd5 14.Rad1 Nxf4 15.Rxd8 Rxd8, as he had played in a game during a blitz match against Kramnik, Moscow 2001. Although he had won the game against Kramnik, 'Playing this position minus the Queen in the last game was too risky', he said later. Instead he continued with the cautious 13...exd5. After 14.Rad1 Be6 15.Qxb7 Bd6 16.Bg5 Rb8 17.Qxa6 Rxb2, the players reached the following position.
Fritz continued 18.Bxf6, the first original move of the game. The remaining minor pieces were eliminated with 18...Qxf6 19.Qxd6 Qxc3 20.Nd4 Rxa2 21.Nxe6 fxe6 and the game petered out to a draw. Kasparov 2.0 - Fritz 2.0
In an interview on WorldChessRating.com, Kasparov said,
This match was very similar to that held in January. I think the human dominated both times. The human had much more opportunities, whereas the defeats were the result of terrible mistakes caused by stress. This time the computer failed to outplay me. I had the initiative on my side throughout the match. I am satisfied with the course of the games. The final outcome depends on me -- if not for the blunders, I would not have lost a single game. The world now awaits the next chapter in the series of exciting man-machine contests.
|
|||||||||||||||||||||||||||||||||||||||||
 |
|||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||
