In 1999, Garry Kasparov agreed to play a game of chess, via the Internet, against the entire rest of the world in consultation, with the World Team moves to be decided by majority vote. The host and promoter of the match was the MSN Gaming Zone.
Prior to the game, Kasparov was considered a prohibitive favorite. Not only was he reigning World Champion, and playing with the advantage of the white pieces, there also had been previous examples of the mediocre competition provided by majority Internet voting. For example, Anatoly Karpov had taken the black pieces against the rest of the world earlier that year, and won effortlessly. However, contrary to all expectations, the game turned into a scintillating brawl, with deep tactical and strategic ideas surfacing to the very end. After sweating through intricate complications for four months, Kasparov admitted that he had never expended as much efort any game in his life, and declared it to be the "greatest game in the history of chess".
The World Team did have several advantages which were innovative for an Internet game. First, four young chess stars were selected by MSN to suggest moves for the World Team. They were, in order of FIDE rating, Etienne Bacrot, Irina Krush, Florin Felecan, and Elisabeth Paehtz. Also, grandmaster Danny King, recruited to provide a running commentary, effectively acted as a fifth move recommendation. Second, the moves were slowed down to a pace of one move per player every two days. That is to say, Kasparov had 24 hours to consider each of his moves, and the World Team had 24 hours to respond, etc. Third, MSN provided a bulletin board for the discussion of the team's moves. It was hoped that these advantages would collectively allow for true consultation, and raise the level of play.
Kasparov played the first move on June 21, his normal king pawn opening, and the World Team voted by a 41% plurality to meet him on his home turf with the Sicilian Defense.
1.e4 c5 2.Nf3 d6 3.Bb5+
This was first surprise of the game. Kasparov typically plays the more ambitious 3.d4 in this position, immediately opening the game. His actual move is more likely to result in a closed game with only a strategic pull for White. Kasparov apologized to the World Team for this move, but excused himself in light of his upcoming match for the Chess World Championship against Viswanathan Anand in 2000. Presumably he had prepared some innovations in his main lines, and didn't want to reveal them in advance.
3...Bd7 4.Bxd7+ Qxd7 5.c4
Kasparov solidifies his hold on the d5 square by advancing the c-pawn before developing his other knight, which will also attack d5. The World Team responds by contesting control of the d4 square.
5...Nc6 6.Nc3 Nf6 7.0-0 g6
The World Team opts to fianchetto the bishop, further contesting the dark central squares. Kasparov immediately breaks up the center with his queen's pawn, before the black bishop can come to bear.
8.d4 cxd4 9.Nxd4 Bg6 10.Nde2
The center was getting too hot for the knight; the World team was threatening a discovered attack by moving the knight away from f6. Exchanging knights on c6 would be silly for Kasparov, as it would bring a black pawn to c6, giving the World Team greater control of d5; instead a retreat is in order. Kasparov's move was correct according to opening theory of the time, but it has since fallen out of favor, precisely because of what occurred next in this game. If White wants to achieve the strategic aims of this opening line, i.e. cramping Black's position without allowing counterplay, then a different move order is necessary.
10...Qe6!
This fantastic move was a novelty by the World Team, i.e. a move which had never before been played in a recorded game. Krush had discovered and analyzed the move, and enlisted Paehtz to recommend it as well, to give it a better chance of winning the vote. Their combined recommendation, plus much discussion on the bulletin boards, was enough to gain it 53% of the vote.
A few notes about World Team cooperation are in order here. First, MSN requested after this move that the four analysts not coordinate with each other, perhaps to ensure a greater variety of recommendations. The analysts worked in isolation from each other thereafter.
Second, it was clear from a look at the voting results that, although the World Team was managing to pick theoretically correct moves, many rank amateurs were voting as well. Demonstrably bad moves were garnering a significant percentage of the votes; even worse, on move 12, about 2.4% of the voters chose a move which didn't get the World Team out of check! (MSN declined to release the numbers, but apparently two or three thousand people were voting on each move.)
Third, it was clear from a glance at the discussion board that the World Team was not coordinating well with itself. Typical posts were of the form, "My suggested move is brilliant, and if you don't vote for it, you are an idiot!". Much more energy was being spent on flame wars than on analysis.
Fourth, of the analysts only Krush was even attempting to listen to the discussion board, and weed out the bad analysis from the good. As it turned out, her patient efforts in an environment which appeared almost hopelessly hostile gradually earned admiration and respect. Through her postings of analysis, the cooperation on the World Team gradually became more real than ostensible.
The tenth move was a turning point for the World Team, not only because it increased Krush's stature and energized the rabble, but because it blew the position on the board wide open. The black queen forks Kasparov's central pawns; they can't both be saved. Counterattacking with 11.Qb3 is met by 11... 0-0 12.Qxb7 Rfc8, and the World Team will win back a central pawn with a favorable game. Kasparov is instead forced to enter the maelstrom with the next several moves.
11.Nd5 Qxe4 12.Nc7+ Kd7 13.Nxa8 Qxc4 14.Nb6+ axb6
After forced moves on both sides, Kasparov makes a desperado move with his knight to double the World Team's pawns. Materially the game is still even, with a knight and two pawns balancing a rook. Positionally, the World Team has the disadvantages of doubled pawns and a centralized king, but the advantage of a lead in development and a central pawn mass. With no central pawns, Kasparov can't easily expose the black king. In the judgement of most commentators, the World Team is at least equal, and it may even be Kasparov who is fighting uphill.
Kasparov rose to the challenge with an excellent move picked from alternatives which would have let the World Team take a strong initiative. For example, it is tempting to harass the black queen and possibly finachetto the white bishop with 15.b3, but this invites the World Team to switch wings and initiate a kingside attack with 15...Qh4. Or, to blindly follow the rule "never move a piece twice in the opening when you can develop another piece" with 15.Be3 would allow the World Team to play 15...Nd5, bringing the knight to where it most fervently wishes to be. Kasparov's actual move contests d5, somewhat blunts the effect of the black bishop on g7, and retains a compact, flexible position.
15.Nc3!
On the fifteenth move the World Team hotly debated a number of promising alternatives, including 15...e6 (still contesting d5), 15...d5 (occupying d5 outright!), 15...Ne4 (trading off Kasparov's best-placed piece), 15...Rd8 (intending to get the king to the wing and mobilize the central pawns), 15...Ra8 (pressuring the queenside and threatening a rook lift via a5) and 15...b5 (threatening to dislodge the white knight and pressure the queenside). The plethora of strong options available to the World Team was reflected in the analysts' recommending four different moves. The results of the vote were illuminating.
By this point in the game, a consortium of Russian grandmasters, The GM School had taken an active interest in the game, and was posting recommendations and analysis on their Web site. Yet although this weighty group recommended 15...b5, along with Paehtz, it came in a distant second with 15% of the vote. In first place was Krush's recommendation of 15...Ra8 with 48% of the vote.
15...Ra8
The results of the vote were a reflection of the increasing coordination of the World Team. Krush was maintaining an analysis tree, and continually updating with all the suggestions and refutations from the bulletin board. Not only did the analysis tree allow the World Team to work with less duplication of effort, it served as a standing, detailed argument for the correctness of the recommended move.
In short, Krush was faciliting two tasks simultaneously: not only discovering a good move, but building a consensus that it was indeed a good move. Given that Krush had become the center of all the cooperative effort of the World Team (as opposed to individual effort, however heroic), even players of much greater strength than Krush began sharing their ideas with her, so that she would incorporate them into her analysis. In particular, Alexander Khalifman of the GM school struck up a constructive correspondence with her.
Kasparov, for his part, continued to play inspired chess.
16.a4!
This beautiful move is a near-refutation of the World Team's 15th move, far stronger than the alternative straightforward developing moves. For starters, it stops the black rook lift cold: 16...Ra5 is now met by 17.Nb5!, paralyzing the black queenside. Simultaneously, Kasparov is threatening a rook lift of his own via Ra3, which is decisive in several continuations. Finally it fixes the doubled b-pawns and makes them future targets. The game remains very unclear and dynamic, but it now appears to be the World Team which is fighting to retain its balance.
Again for the 16th move the four analysts made four different recommendations. This time Krush's suggestion of 16...Ne4 garnered 50% of the vote to 14% for 16...Nd4 in second place.
16...Ne4 17.Nxe4 Qxe4 18.Qb3
The World Team forces Kasparov to trade off his only piece that isn't on the back rank, and simultaneously unmasks the action of the g7 bishop. Kasparov responds with a queen fork of the black pawns on b6 and f7. The loss of a pawn appears unavoidable, but Black can gain some counterplay. The bulletin board debate raged between playing 18...e6 19.Qxb6 Nd4, to make sure it was the doubled pawn which went missing, or the more aggressive immediate 18...Nd4, allowing 19.Qxf7. Khalifman, however, found the extremely subtle move 18...f5, and after chewing it over, the bulletin board was for the most part convinced.
The strength of that consensus was tested when the other three analysts unanimously recommended 18...Nd4. The vote came out with 43% in favor of Krush's reccomendation of 18...f5, and 35% in favor of the otherwise unanimous recommendation of 18...Nd4. At this point there was some loud grumbling on the bulletin board that Krush had "taken over the game". And indeed, her recommendations were selected every single move from the 10th to the 50th. But the situation was absolutely not one of Krush manipulating the masses. She patiently listened to everyone on the bulletin board, and incorporated all suggested analysis in her analysis tree. Krush's genuis was not so much in excellent analysis as in enabling cooperation of the masses, and voicing the consensus created thereby.
18...f5!
The World Team now offers Kasparov the b6 pawn, but for a price. After 19.Qxb6 Nd4, the World Team would have dual threats of Nc2 and Ra6, ensuring very active play for the pawn. If instead Kasparov continues developing with 19.Be3, the World Team can offer a queen trade with 19...Qb4, and bank on the central pawn mass to be quite strong in any endgame. But rather than these, Kasparov once again finds a powerful continuation: a developing move with stronger attacking possibilities.
19.Bg5
Kasparov, up against much stiffer resistance from the World Team than he had imagined possible, began to drop hints that he was effectively playing against the GM School, and not against the Internet as a whole, but move 19 (among others) debunked that theory. The GM school recommended 19...Qd4, while the bulletin board found a flaw in their analysis, and generally favored 19...Qb4 as being more forcing. Furthermore, for much of the game, a few top grandmasters of the GM School were busy with other commitments, and the World Team analysis was driven instead by a handful of dedicated IM's and FM's, along with dozens of amateurs exploring and double-checking countless lines with strong chess software.
On this particular move, the voting was further complicated by a strong faction in favor of 19...Nd4, with a counter-attack, and this was the recommendation of both Felecan and Paehtz. The winning total of 35% for 19...Qb4 was the lowest winning total for any move of the game, perhaps because the move essentially forced Kasparov to intensify his attack on the kingside. A queen trade is strategically out of the question for White, as it leads to a favorable endgame for Black.
19...Qb4 20.Qf7
It is tempting for Black to grab a pawn and protect the bishop with 20...Qxb2, and dare the white rooks to occupy whichever files they will, but the World Team's king sits rather less comfortably than before. After much debate on the bulletin board, no clear refutation of the pawn grab was discovered, but it was still deemed too risky. Bacrot, Felecan, Paehtz, and King independently agreed, and by a large margin the World Team decided to protect the bishop and close the e-file with
20...Be5
Kasparov of course did not fall for the simple trap of grabbing the h-pawn, because 21.Qxh7 Rh8 (skewering Kasparov's queen and h-pawn) 22.Qxg6 Bxh2+ 23.Kh1 Qg4 will win at least a piece for the World team. Instead he opts for a simple defensive move which restores the threat to plunder the kingside.
21.h3
A few World Team members favored shoring up the kingside with 21...Rh8, effectively admitting that the 15th move was a mistake. To defend now would leave Black with a very passive position, and all sorts of difficulties would ensue. Instead, the World Team opted to play actively, exchanging its kingside pawns for Kasparov's queenside pawns. This line shows that the pressure the black rook exerted on the queenside was not illusory, and Kasparov's reply on move 16, albeit brilliant, did create a weakness.
21...Rxa4 22.Rxa4 Qxa4 23.Qxh7 Bxb2 24.Qxg6 Qe4 25.Qf7 Bd4
The dust has settled, and the material is still even, with a rook balancing a knight and two pawns. With a pair of rooks exchanged, and neither side having pawn levers to use against the enemy king, both kings are safe enough that direct attacks are unlikely. Therefore, although the queens remain on the board, the game starts to take the character of an endgame, with the struggle to promote a pawn rising to the foreground. Indeed, Kasparov could immediately begin marching his h-pawn forward, and the World Team would have difficulty restraining it. On the other hand, the World team's b-pawn would be able to advance equally quickly, making the position very double-edged. Rather than launching the race at once, Kasparov makes a subtle move to tie down the World Team into a more passive position.
26.Qb3
Kasparov hits at the weak b-pawn, and prepares Be3. The World Team would not be happy trading bishops at this point, and considered the consolidating move 26...Bc5 so that 27.Be3 could be met with 27...Nd4. However, Kasparov has the even deeper threat of first using his queen to help his rook into play with 26...Bc5 27. Qb1! The World Team can't afford a queen exchange which would bring the white rook to life, but moving away the queen allows 28.Re1, and suddenly the white pieces are coordinating very well.
In keeping with its play throughout the game, the World Team found a sharp, active alternative in 26...f4, which extensive analysis showed to be at least as good as 26...Bc5. However, Krush's recommendation on behalf of the bulletin board once against stood alone against the unanimous recommendations of the other three analysts. In a hair-raising vote, 26...f4 edged out 26...Bc5 by a margin of 42.61% to 42.14%.
26...f4!
The World Team blocks off Kasparov's bishop from its natural post on e3, and threatens to generate an attack on the white king after all. Now 27.Qb1 can be met by 27...Bxf2+, 27.Qd1 can be met by 27...f3, and starting the pawn race with 27.h4 is met by 27...Ne5 with attacking play for the World Team. Kasparov instead opts for a simple and natural move.
27.Qf7
Moving the queen to the square it just came from only superficiallly appears to lose time. In actuality, the World Team will have to use a move to defend the f-pawn, while the queen indirectly supports white's h-pawn to advance, and puts the brakes on the World Team's threat of advancing the pawn further to f3. The pawn race is now Kasparov's best chance to play for advantage.
27...Be5 28.h4 b5 29.h5 Qc4
The World Team could not afford to blindly keep racing the b-pawn forward, as the white queen still guards the b3 square, which Black would have to lose a tempo to guard before advancing again. The actual move does not lose a tempo, because Kasparov cannot afford to trade queens in a way that undoubles the black pawns and gives the World Team a central steamroller for the endgame. The alternative 29...Qe2 may also hold the fort for Black, by offering to exchange the white h-pawn for the black f-pawn. However, as part of the trade, the bishops would also come off, and none of the four analysts was prepared to trade the World Team's lovely bishop just yet.
Some of the bulletin board analysis focused on Kasparov's possible reply 30.Qf8, keeping the queens on and threatening to harass the black king from behind. However, computer checking of many lines found no advantage for White in this strategy, and in fact revealed chances for White to press too hard in a complex position and stumble into disadvantage. Kasparov elected to force a queen trade, break free his imprisoned bishop, open the f-file for his rook, and create connected passed pawns in a pure endgame. The World Team's responses were essentially forced.
30.Qf5+ Qe6 31.Qxe6+ Kxe6 32.g3 fxg3 33.fxg3
Despite the reduced material, the position remains sharp due to the presence of six(!) passed pawns. The World Team now has the option of snatching Kasparov's g-pawn, losing two tempi in the queening race. After the sequence 33...Bxg3 34.h6 Be5 35.h7 Bg7 36.Rf8 b4 37.h8Q Bxh8 38.Rxh8 an extremely unbalanced endgame would ensue, with Kasparov having a rook and bishop versus the World Team's knight and four pawns. The central position of the black king may be just enough to hold a draw for the World Team, but none of the four analysts trusted the position enough to recommend it. Instead the World Team opted for counterplay, as usual, this time by a vote of 72%.
33...b4 34.Bf5
Kasparov's offer to trade bishops caught the bulletin board entirely off guard. It had been assumed that Kasparov would try to bring his king into the center to restrain the black pawns, and the World Team gave deep thought to 34.Kf2 Kf5!. After Kasparov's actual move, it would be suicide for the World Team to give up its precious bishop, but opinions differed as to whether to hunker down with with 34...Bh8 or keep active with 34...Bd4+. Kasparov admitted afterwards that he could find no advantage after 34...Bh8, but since only Danny King seemed to give it serious consideration, and all four analysts felt more comfortable with the more active move, 34...Bd4+ overwhelmingly won the vote.
34...Bd4+
The World Team had hastily put together proposed defenses against either of Kasparov's king advances 35.Kg2 and 35.Kh2. In one of the former lines, the black knight threatens to usher home the b-pawn and returns to the kingside just barely in time to stop the pawn, delivering a check from f4 on the way. In one of the latter lines it turned out to be critical that the black bishop could attack the white king from e5. But Kasparov stunned everyone (including the GM School) with an incredible move:
35.Kh1!
Although it would seem crazy to move the king away from the action into a corner where it can neither support the white pawns toward queening, nor delay the black pawns from queening, this move puts the World Team in a serious predicament.
But perhaps even greater than the effect of this move on the position was its effect on the psyche of the bulletin board. For the second straight move, Kasparov had avoided almost all of the World Team's preparation without positionally conceding anything in the process. A small number of people had been rude and abusive the entire game, and Krush had graciously accepted their analysis while ignoring their incivility, but the tenuous nature of the World Team's position emboldened the complainers. For example, there were shrill cries that the 33rd move had lost the game for the World Team.
Possibly even more annoying, from Krush's perspective, were the ego-driven participants who contributed their analysis to the World Team only under the condition that their genuis be recognized in comments in the analysis tree. Krush had already credited every good move idea for black to whomever posted the analysis first on the bulletin board, and didn't attach her own name to any of her own freely contributed analysis, but people demanded further that they be given credit for entire complexes of moves, including Kasparov's good ideas, presumably by virtue of having posted them to the bulletin board before Kasparov thought of them. As the World Team began to panic in a dangerous position, the flames, insults, and petty bickering reached heights not seen since the initial dozen moves of the game.
Eventually Krush got fed up with the pressure of providing courteous service to the World Team while enduring constant criticism and invective. She drastically scaled back her participation on the bulletin boards. She continued to maintain an analysis tree, and eventually, at the begging of many World Team members, re-assumed a somewhat more active leadership role, but the unity and effectiveness of the World Team was definitely strained for a while, and perhaps permanently damaged.
In the scramble following Kasparov's 35th move, no one on the World Team noticed that 35...Ne5 is probably enough to hold the draw, and somewhat fatalistically opted for pushing the b-pawn.
35...b3 36.g4
Kasparov has connected passed pawns which can support each other, whereas the World Team needs the knight (or possibly even the king) to laboriously move into position to usher the black b-pawn to queening. Furthermore, should the black bishop move, the white rook can slide over to g1 where it supports the g-pawn from behind while still keeping an eye on the b1 queening square, an additional subtle point of Kasparov's 35th move. Finally, by temporarily controlling the dark squares with his bishop and the light squares with his pawns (which 36.h6 wouldn't have done) Kasparov keeps the black king from advancing to f5, which in some lines is sufficient to blockade the pawns.
The bulletin board was near despair at this point, having convinced itself that 36...b2 loses to 37.g5 Nb4 38.g6 Nd3 39.h6, and then 39...Nxf4 is not check due to Kasparov's 35th move, and therefore fails to hold the draw. Similarly an immediate 36...Nb4 merely transposes to the above line and loses. The only move for which some lines seemed still possibly drawn was 36...Kd5, which Krush duly recommended, but Bacrot and Felecan recommended 36...b2, while Paehtz favored 36...Nb4. This created yet another razor-thin vote, with 36...Kd5 scoring 37.69% of the vote winning over 36...b2 with 37.11%.
36...Kd5!
This game, which had started in June, now spilled over into September, longer than anyone had anticipated. Kasparov, however, had grown sufficiently confident in his position that he let it leak that he would shortly make an announcement about the game, presumably an announcement of a forced win. The black pieces do not seem up to the task of both holding off the white pawns and pushing through the black pawn, while the white rook is working effectively through threats alone, without even moving for the present.
37.g5
Kasparov's two passers will apparently require two black pieces to restrain, but the knight can't cross over via e5 because Kasparov will simply exchange the bishop for it. Therefore the bulletin board turned its attention to 37...e5, driving away the white bishop and clearing the way for Ne7. But Kasparov has a devilish reply in 38.Bc1! His pawns are so strong in this line that he can afford to sacrifice his bishop for Black's b-pawn, particularly since the black bishop is temporarily cut off from the h8 queening square. Also, in many lines where Black doesn't force the white bishop to give itself up, it can reverse field with Ba3, tying the black king to the defense of the d6 pawn.
With its back against the wall, the World Team found the only saving move.
37...e6
This move opens e7 so the black knight can cross over, but keeps open the a1-h8 diagonal for the black bishop. Kasparov might have thought that 38.Rd1 was winning in this line. Certainly many participants on the bulletin board thought so! But an exhaustive analysis shows that the World Team has the resources to hold on, at times by the narrowest of margins, if it responds with 38...Ke4. The power of the centralized black king vis-a-vis the white king off in the corner comes into play, showing that even brilliant chess moves have minor disadvantages.
Instead of trying the complexities of the 38.Rd1 line, Kasparov called off his announcement and began to force the World Team into an ending in which both sides get a new queen, and the outcome is still very unclear.
38.h6 Ne7 39.Rd1 e5 40.Be3 Kc4 41.Bxd4 exd4
The World Team's moves are all essentially forced. Kasparov's main chance to deviate would have been with 40.Bc1, but then 40...Ke6 appears to hold. Although the black bishop is at that time temporarily cut off from holding back the pawns, the black king is temporarily not cut off, and an extra tempo in such positions makes all the difference. Kasparov, when making his 38th move, had in all probability already elected to force the World Team's responses all the way to move 50.
As the game became more and more strategic, with endgame plans replacing middle-game tactics, voting for a good move seemed increasingly beyond the ability of casual World Team members. For example, on the 41st move Kc4 is a crucial tempo in the World Team's fight to promote the b-pawn, and it was recommended by all the analysts and Danny King, but it recieved a mere 79% of the vote. Earlier forced moves had gotten as much as 98% of the vote. Perhaps this reflected the fact that many weak players got hints from their home computers, which are notoriously poor at endgames.
The World Team's pawns, although they are ugly, are now a sufficiently potent threat that Kasparov's rook can't take the double role of guarding the home rank and forcing through the passed pawns against black's knight. Therefore the white king must come out of its corner at last.
42.Kg2 b2 43.Kf3 Kc3 44.h7
Kasparov could have made the pawn advance on either of the two previous moves as well, but it would have merely transposed, with the World Team responding as it did in the actual game:
44...Ng6 45.Ke4 Kc2 46.Rh1
The World Team needs to advance the d-pawn as well as the b-pawn, in order to get a second queen after Kasparov sacrifices his rook for the first one. Against the recommendations of the analysts, nine percent of the World Team voted for the horrible blunder of immediately queening the b-pawn, which loses after 46...b1Q 47.Rxb1 Kxb1 48.Kxd4. It was merely to give the World Team this opportunity to blunder that Kasparov marched his king to the center instead of straight forward. The white king will arrive to assist the white pawns just as quickly with diagonal moves as with straight ones, and the diagonal moves give it incidental threats in the center.
46...d3 47.Kf5
The analysts were again unanimous, this time recommending immediate queening. Yet 15% of the voters were tempted to try to hang onto the knight a few moves longer with 47...Nh8. This would have led to a lost endgame after 48.g6 d2 49.g7 d1Q 50.Rxd1 Kxd1 51.gxh8Q b1Q+, when Black can't engineer a perpetual check.
47...b1Q 48.Rxb1 Kxb1 49.Kxg6 d2 50.h8Q d1Q
The race to queen a pawn is over, and it ends in a tie! The position is not equal, though. Kasparov's far-advanced g-pawn is an imminent threat to queen, while the World Team's pawns might be liabilities as much as assets, since they give the white king a modicum of shelter from threatened checks. The general plan of defense for the World Team is to place its king favorably, so that Kasparov can't threaten a deadly queen trade, and then determinedly check Kasparov's king to prevent him from queening the g-pawn. In some lines the World Team gains counterplay by advancing its own pawns, but this is a secondary strategy.
The World Team had relied heavily on computer analysis for much of the game, but at this point the forward-searching chess engines began to produce worthless suggestions. This type of position is seldom understood by computers, except by endgame tablebases. As of October 1999, however, there were no seven-piece endgame tablebase, and seven pieces remain in the position. After the game was over, IM Ken Regan constructed a specialized tablebase for the purpose of fully understanding this endgame, and was able to prove that the postion after the World Team's 50th move was drawn with best play on both sides. Both Kasparov and the bulletin board suspected the position was drawn, but as the further course of the game proved, no one fully understood the position at the time.
Some World Team members tied to gain insight from the position by consulting state-of-the-art five piece tablebases, with the black pawns missing, and were encouraged to find the position dead drawn. Unfortunately for the World Team, the extra black pawns complicate the position enormously. For example, analysis positions arose which were theoretically drawn with both black pawns in place, but which would be won for White if exactly one black pawn weren't there, and then again drawn with both black pawns gone.
51.Qh7
This fine move threatens to gain Kasparov time by advancing the king with discovered check. The bulletin board and Krush came up with the response 51...Ka1, which Kasparov later said he had considered dead drawn, and the subsequent tablebases confirmed to be so. But Felecan recommended 51...d5 and Paehtz favored 51...b5, while Bacrot sat out the turn. For the first time in 40 moves, Krush's recommendation was not selected, recieving only 34% for 51...Ka1 to 39% for 51...b5.
There are several reasons the vote might have gone as it did. For starters, endgames are hard for amateurs, and even for fairly accomplished players, which makes it easier to go astray. Secondly, computers are little help. Thirdly, the World Team had done well with counter-attack the entire game, eschewing purely defensive moves whenever there was a more active alternative. It was in keeping with the fighting spirit of the World Team to bid for a promotion of its own, rather than merely hoping to hinder Kasparov's promotion.
Unfortunately, an uglier possibility emerged when someone bragged to the bulletin board that he had tricked MSN into letting him vote multiple times for 51...b5. He went so far as to tell everyone how he had done it. Whether or not he had actually added unfair votes, and if so, whether or not the number of votes added was enough to tip the balance, the ballot-stuffing method he outlined was indeed workable, as several bulletin board members verified on the 59th move, after Kasparov had already secured the victory.
51...b5?! 52.Kf6+
The bulletin board and Krush now concluded that 52...Kc1 offered the best chances of holding a draw, and tablebases later verified it was both sufficient and necessary. However, with Bacrot recommending 52...Ka1 and both Felecan and Paehtz favoring 52...Kb2, the latter move eked out a victory with 42%. The idea is to use the king to support the b-pawn towards promotion, but it doesn't quite work.
52...Kb2?
Now tablebases show that Kasparov has a forced win after 53.Qe4, but the win is so deep that he failed to spot it. This endgame position was beyond the realm of previous endgame theory, and even the World Champion did not quite grasp all the thread. Instead he played the aggressive but inadequate
53.Qh2+?