Most people expected Garry Kasparov to win, but Vladimir Kramnik had prepared a nasty opening surprise for the match and it helped him win the title.

Editor’s Note: This is part of a series of articles that we began before the recent World Championship match in New York City. The earlier articles in this series can be found hereherehere, here, and here.

The World Championship match between Garry Kasparov and Vladimir Kramnik was not only the first of the millennium, it marked the shift into the modern era of how the game is played.

The match took place against a backdrop of division and discord in the chess world. The World Chess Federation, also known as FIDE, did not recognize the match as one for the title because Kasparov had split with the federation in 1993 in a disagreement about the organization of another title match – that one with Nigel Short of England. Despite FIDE’s refusal to sanction the Kasparov-Kramnik match, there was little doubt in the public’s mind that Kasparov was the true champion.

There were also questions about why Kramnik, a former student of Kasparov’s, was playing for the title. There had been an official challenger’s match in 1998 between Kramnik and Alexei Shirov of Latvia – and Shirov had won. But there were problems finding sponsors, so a Kasparov-Shirov match never materialized.

In the end, Kasparov decided that Kramnik was the best challenger because he was No. 2 in the world. The logic was inescapable and nobody could accuse Kasparov of trying to hold on to his title by ducking his best opponent (which was a more common practice in the early 20th century when the champions chose their challengers). Shirov would have been a lot easier opponent for Kasparov, particularly considering Kasparov’s substantial plus score against him. 

Once the match was set, Kramnik prepared for it rigorously. He trained and trained and when he showed up in London, where the match was held (under the auspices of the sponsor, a company called Braingames), he was sporting a new, short haircut and had lost weight.

Most importantly, he decided to adopt a defensive style of play that was clearly inspired by the best computer programs at the time. As Kramnik put it in a 2003 interview with Frankfurter Allgemeine Zeitung, “You can’t play the same way today as people did ten years ago. I admire Kasparov’s imaginative attacking victories from the ‘80s and ‘90s, but when you check them with a computer, in every other game the machine accepts the sacrifice, defends, and wins.”

The style wasn’t very exciting, but it also helped Kramnik avoid a single defeat against arguably the world’s best chess player. It isn’t easy to play defensively against an attacking genius like Kasparov who was always extremely well prepared, so Kramnik had to come up with an innovative strategy in the opening. He did. He decided to use the Berlin Defense (also known as the Berlin endgame and even, in jest and admiration, the Berlin Wall).

I can’t think of another World Championship match where such a stunningly new concept debuted. For a long time before this match, the Berlin was considered to be weird and dubious for Black. But Kramnik, and his team of seconds, discovered a goldmine of extremely rich strategic ideas.

In the first game, Kramnik completely neutralized Kasparov’s aggressive mood with some clever maneuvers in the opening. It was amazing to see Kramnik play Nf5-Ne7-Ng6 to improve the placement of his knight. It is a plan that has been used many times since then, but it certainly isn’t intuitively obvious. Kasparov probably felt that the maneuver was dubious, but he couldn’t create any problems for Black:

Kasparov, Garry vs. Kramnik, Vladimir
World Championship | London | Round 1 | 08 Oct 2000 | ECO: C67 | 1/2-1/2
Ne7! 14. Ne2 Ng6 White can't really target the Black knight on g6. And the knight helps to keep the White pawns at bay.
15. Ne1 h5 16. Nd3 c5 17. c4 a5 A typical idea in the Berlin stuff. The move forces White to create weaknesses on the queenside.
18. a4 h4 19. Nc3 Be6 20. Nd5 White's maneuvers haven't done much for him, while Black has really improved his position. Now he can just shuffle his pieces and White cannot make any progress.
20... Kb7 21. Ne3 Rh5 22. Bc3 Re8 23. Rd2 Kc8 24. f4 Ne7 25. Nf2 Nf5

In Game 2, Kramnik again demonstrated nice preparation to get an edge against Kasparov. Kasparov nearly managed to hold a draw, but Kramnik’s excellent technique kept up the pressure until Kasparov blundered and lost.

Kramnik, Vladimir vs. Kasparov, Garry
World Championship | London | Round 2 | 10 Oct 2000 | ECO: D85 | 1-0
gxf4 The opposite-colored bishops usually lead to a drawish endgame. Even though Kasparov is close to having position he can easily draw, there are still a few pitfalls.
27. e6! Winning a pawn because after
27... fxe6 28. Rxe6 The pawn on a6 can't be protected because of the threat of discovered check. Though White has won a pawn, the game is far from over. Indeed, with best play, Kasparov should be able to draw. But Kramnik keeps the pressure up, torturing Kasparov, who must defend his slightly worse endgame. Over the next few moves, Kramnik slowly improves his position:
28... Kg7 29. Rxa6 Rf5 30. Be4 Re5 31. f3 Re7 Moving the Black rook to a7 seems like the correct way to defend against the advance of the a-pawn.
32. a4 Ra7 33. Rb6 Be5 34. Rb4 Rd7!? Black could have also kept the rook on a7 to keep the White a-pawn in check.
35. Kg2 Rd2+ 36. Kh3 h5 37. Rb5 Kf6 38. a5 Ra2 39. Rb6+ Ke7?? Just before the time control ends, Kasparov cracks.
39... Kg7 40. a6 Bd4 41. Rd6 Be3 White can continue trying to build pressure for a while more, but it is difficult to see how White will ever get to play a7. And his king is stuck on h3 as well for the foreseeable future.  )
40. Bd5! Ra2 is attacked - and Re6+ wins the e5 Bishop.
40. Bd5 Re2 Now the pawn can't be stopped:
41. a6 Bd4 42. a7!  )

In the next game, Kramnik showed another deep and stunning idea in the Berlin. The idea is well-known now but was rather unthinkable then:

Kasparov, Garry vs. Kramnik, Vladimir
World Championship | London | Round 3 | 12 Oct 2000 | ECO: C67 | 1/2-1/2
17. Rfe1 White seems to have a commanding position. He has control of the center and excellently placed pieces, while Black is barely developed. But it is very hard for White to improve his position. Kramnik's great insight in the Berlin Defense was that even in such a passive endgame, Black's position is very solid and there are ways for him to improve it. Here he presents another creative idea:
17... Rg8! Black defends the g-pawn which would be vulnerable if White plays e6. He also has the option of eventually pushing g5, but he is in no hurry. The main problem for White is that, other than playing e6, he doesn't have another plan to improve his position.
18. Nf4!? Continuing to try and play e6. The knight on f4 makes e6 a stronger threat.
18... g5!? This creates a clear weakness on f6 and White can try to get his knight there. But as Kramnik demonstrates, it is not a big deal.
19. Nh5 Rg6 Keeping e6 under control.
19... Bxf3 20. gxf3 White's structure is spoiled, but the bishop is really important for Black's defense. Trading the bishop would allow moves like Rd7, etc.  )
20. Nf6 Bg7 21. Rd3 Bxf3!? This wasn't necessary, but Kramnik sees a way to simplify into a drawish endgame:
22. Rxf3 Bxf6 23. exf6 Nc6 24. Rd3 Rf8! Again excellent defense by Kramnik. White has active rooks, but it is not clear how he can create pressure, while Black is ready to regroup.
25. Re4!? White is trying to create a pawn majority on the kingside with f4 or h4. This would usually be a good idea, but, as will become apparent, having the pawn majority will not help White very much.
25... Kc8 26. f4 gxf4 27. Rxf4 Re8! Creating counterplay.
28. Bc3 Re2 29. Rf2 Re4 30. Rh3 White seems to be better because of his kingside majority, but Black has a typical idea in the Berlin up his sleeve:
30... a5! 31. Rh5 a4 It is very annoying to deal with this pawn.
32. bxa4 Rxc4 33. Bd2 Rxa4 Now the position is clearly very double-edged because White's queenside has collapsed. The rest of the game was quite messy, but, in the end, it ended in a draw.

In the next game, Kasparov tried to change his opening with Black, but again got into serious trouble. However, Kramnik let him escape in a tense endgame:

Kramnik, Vladimir vs. Kasparov, Garry
World Championship | London | Round 4 | 14 Oct 2000 | ECO: D27 | 1/2-1/2
Re5 55. a6+? After this the knight on c7 and pawn on a6 get entangled. White is unable to keep the knight on c7 for long to continue to defend the pawn. Instead, White had a nice winning maneuver:
55. Nd5+! Ka6 56. Nb4+! The point is Black cannot play Kxa5 without allowing Nc6+, winning the rook.
56... Kb5 57. Rf4! Again an unexpected move. White's position doesn't look any more stable than in the game, but after a6 on the next move, the pawn would be untouchable. That is because Black can't force White to move the knight on b4 because he also has to deal with the passed pawn on a6.
57... g5 58. Rd4  )
55... Kb6! 56. Rxg7 Ra5 Now all of White's pieces are stuck and Black was able to draw quite easily:
57. Kd2 Ra1 58. Kc2 Rh1 59. Kb2 Rh8 60. Kb3 Rc8 61. a7 Kxa7

Kasparov was completely frustrated at this point, so in Game 5 he avoided 1.e4. It was an incredible achievement for Kramnik’s preparation. Kasparov switched to 1.c4 but got no advantage. Indeed, Kramnik perhaps even had a small edge when they agreed to a draw after move 24.

Kramnik surprised Kasparov in the opening in Game 6, got a substantial edge and pushed Kasparov to the brink, but the game ended in a draw.

Kasparov tried 1.c4 again in Game 7 and did even worse than in Game 5. He offered a draw after just 14 moves. It was amazing that Kasparov had posed no problems for Black in four games in which he had White.

In Game 8, Kasparov, who had Black, played aggressively and finally got Kramnik in trouble, but precise defense by Kramnik allowed him to escape:

Kramnik, Vladimir vs. Kasparov, Garry
World Championship | London | Round 8 | 21 Oct 2000 | ECO: E32 | 1/2-1/2
18. dxc5 Black is better developed, but it isn't clear if that's important in the current position. Kasparov makes the best use of his development edge by playing:
18... f5! White's center is surprisingly weak and vulnerable.
19. cxb6 axb6 20. Ne2
20. exf5 exf5 And the open e-file will create enormous problems for White.  )
20... fxe4 21. fxe4 Bxe4 22. O-O Kramnik attempts to save the game by giving up his extra pawn, but he still has to deal with the much more active Black pieces. Kasparov keeps the pressure on in the next few moves:
22... Rd2 23. Nc3 Bb7 24. b4 Rf8 25. Ra2 Rxa2 26. Nxa2 Nd5 27. Bd4 Ra8 28. Nc3! Exchanging the knights improves the defensive side's chances in opposite-colored bishop endgames as it is easier to create a blockade of any passed pawns.
28. Bb2 Ne3! 29. Rf2 Nc4 And the pawn on a3 will fall.  )
28. Rf3 Would have held on to the pawns for a while longer. But there was a tricky idea that White would have needed to find:
28... Nxb4?! 29. Rg3! After 29. Nxb4 Bxf3 30. gxf3 Rxa3, Black would be a bit better and could make White suffer quite a bit.
29... Nxa2 30. Rxg7+ Kf8 31. Rxb7 And now White is better. Despite that variation, 28. Rf3, seems like a shaky move, and Kramnik's desire to exchange the knights and steer toward an opposite-colored bishoped endgame feels very practical.  )
28... Nxc3 29. Bxc3 Rxa3 30. Bd4 b5 31. Rf4! A forcing move that quickly clarifies the situation. Anything else would have led to long-term suffering for White, and there is always a chance of making a mistake as happened in Game 2 when Kasparov collapsed in an opposite-color bishops and rooks endgame.
31... Rd3 32. Rg4 g5 33. h4 Kf7 34. hxg5 hxg5 35. Kf2 Rd2+
35... Kg6 36. Be3  )
36. Ke3 Rxg2 White forced Black to exchange the rooks in the last few moves or else the pawn on g5 would have fallen - and now, even though Black has two extra pawns, it is quite easy for White to draw.
37. Rxg2 Bxg2 38. Be5 There is no way Black will be able to advance his g-pawn past the g4 square, or the e-pawn past e6 because he can't break White's blockade of the dark squares.
38. Be5 Kg6 39. Bc7 Kf5 40. Kd4 Kg4 41. Ke3 Kh3 42. Be5 g4 43. Kf2 so on.  )

Kasparov took another crack at the Berlin in Game 9, but Kramnik demonstrated how many resources the opening had by choosing a different setup than in his first two games with the system. It clearly foiled any preparation Kasparov’s team had done in the previous week. So, once again, Kasparov failed to get any advantage:

Kasparov, Garry vs. Kramnik, Vladimir
World Championship | London | Round 9 | 22 Oct 2000 | ECO: C67 | 1/2-1/2
9. Nc3 This time Kramnik uses a slightly different setup. In the previous games Kramnik, had moved his king to the queenside, but now he is willing to keep it in the center.
9... h6!? A new move compared to Bd7, which Kramnik played before.
10. Rd1+ Ke8 11. h3 a5 12. Bf4 Be6! Stopping White from playing e6.
13. g4 Ne7 14. Nd4 Nd5 Allowing the exchange of the light-squared bishop is usually not the best idea for Black in the Berlin, but allowing Nxe6 would be fine in this case. After 15. Nxe6 fxe6, Blacck's pawn structure is actually a bit better than it was because it is easier for him to check White's kingside pawn majority as the breakthrough after f4 and eventually f5 can be slowed by g6.
15. Nce2 Bc5! 16. Nxe6 fxe6 17. c4 Nb6 18. b3 a4 As in earlier games in the match, the thrust a5-a4 helps Kramnik create counterplay.
19. Bd2 Kf7 20. Bc3 Rhd8 Black is too active to have much to worry about from Whites kingside pawn majority. Kramnik faced no problems in the rest of the game:
21. Rxd8 Rxd8 22. Kg2 Rd3 23. Rc1 g5 24. Rc2 axb3 25. axb3 Nd7 26. Ra2 Be7 27. Ra7 Nc5 28. f3 Nxb3 29. Rxb7 Nc1 30. Nxc1 Rxc3 31. Ne2 Rc2 32. Kf1 Rxc4 33. Rxc7

Kasparov’s frustration was clearly building. In the next game, he made an early tactical mistake and went down quickly with Black, giving Kramnik a two-point lead:

Kramnik, Vladimir vs. Kasparov, Garry
World Championship | London | Round 10 | 24 Oct 2000 | ECO: E54 | 1-0
Be7 14. Bxf6! Nxf6 15. Bxe6! fxe6
15... Rc7 Was still perhaps fine for Black, but it required some very precise calculations:
16. Ng5 Qxd4! 17. Nxf7 Bc5! And the threat with Qxf2+ gives Black enough counterplay.  )
16. Qxe6+ Kh8 17. Qxe7 Bxf3! 18. gxf3 Qxd4 It seems like Black has enough compensation to be able to draw, particularly as the White king could be in danger, too But it is actually Black's king that has more to worry about as is soon clear.
19. Nb5! Qxb2 20. Rxc8 Rxc8 21. Nd6 Nf7+ will create some serious threats.
21... Rb8 22. Nf7+ Kg8 23. Qe6! Rf8 24. Nd8+ Kh8 25. Qe7 It is stunning how quickly things ended after Nb5!
25. Qe7 Kg8 26. Ne6 Rf7 27. Qd8+ And Black will be checkmated.  )

Surprisingly, in the next game, Kramnik avoided the Berlin. Perhaps he was a bit bored with it. But, once again, he showed that he had prepared extremely well. He played an old variation arising from the Ruy Lopez that was considered better for White, and equalized with no problem. The game ended in another draw and another lost opportunity for Kasparov with White.

In Game 12, Kasparov played quite provocatively with Black and obtained a tangible advantage. But Kramnik held on, with a little bit of help from a couple of miscues by Kasparov.

Game 13 was another Berlin endgame and Kramnik again made some subtle changes. Kasparov’s frustration was again clear as he offered a draw after just 14 moves despite being two points behind in the match.

The match was all-but-over. Though the last two games were hard-fought, they were essentially formalities.

It was amazing to see Kasparov completely neutralized when he had White. Such a thing had never happened to him, or really to any champion. In the aftermath of the match, Kramnik’s achievement made the Berlin fashionable again. The opening has since been studied to death, and quite often the same boring draws are played again and again. Unfortunately, that makes it harder to appreciate just how incredible Kramnik’s conceptual ideas were at the time.


Parimarjan Negi is an Indian grandmaster who is the second-youngest ever to earn the title (at 13 years 4 months and 22 days). Ranked No. 77 in the world, he is a junior at Stanford University. He can be found on Twitter at @parimarjan.