Kasparov vs. The World was a unique experiment, one that is not likely to be repeated again anytime soon. Almost certainly, Kasparov would not do it, and also as likely, Microsoft would not sponsor it.

The primary reason for this is what a fiasco the endgame turned out to be.

The game was set up as follows. The game was 'moderated' by Grandmaster Danny King, who mainly provided some ongoing analyses of the positions. The World was black; Kasparov was white and so went first. The World's moves were decided by a plurality of votes submitted to the web site (hosted on the Microsoft Gaming Zone). To help the World Team, Microsoft had four analysts; these teenage chess experts submitted their recommendations for the next move so that the World Team would have some guidance.

The analysts were Irina Krush, Elisabeth Pähtz, Florin Felecan, and Etienne Bacrot. Of these, Krush took a particular interest in the game, spending many hours with her trainers analyzing each position. She was helped along greatly by analyses done by the World Team and posted on the official World Team message boards -- analyses done there were analyses that she and her team did not have to do.

The message boards were another unique aspect of this experiment. They were a hotbed of activity from almost the very beginning, as new lines of play were constantly being analyzed. Reams and reams of chess notation passed through the boards, along with a disappointing number of flames and trash talk. The boards served as a kind of multi-processor supercomputer, because different lines could be split up and analyzed by different people on their own computers at the same time. This allowed the World Team to go farther down the various lines of play than any single individual would have been able to do. Irina Krush is known to have read the boards, and rumors were rampant that she posted as well, under a pseudonym. There was also suspicion that Kasparov or his assistants read the boards, which would have been cheating.

After Kasparov's move 10, three of the analysts proposed standard responses to the opening moves -- there are only so many ways to open a chess match. But Krush suggested Qe6, an uncommon move that is now referred to as the "World variation." At the time, she admitted it was a gamble -- if it wasn't, it would have been used more often. But then, The World was playing Garry Kasparov -- what did it have to lose? If The World played a mundane game of chess, Kasparaov would cream it just as he did most of his opponents.

But it was here that problems emerged. Along with their suggestions, the analysts usually posted a brief description of what the move would accomplish and where the game was likely to go next. For move 10, Krush supported her suggestion with extensive analysis of the lines to come. The fact that Krush posted this support along with her move was enough to sway most of the voters -- never mind the fact that most of them had no idea whether or not the analysis was correct, or even how to read it in the first place.

You see, the majority of World Team members were not chess experts. The vast majority did not read the message boards. All they saw were the analysts' suggestions, and from what they saw, it seemed as if only Irina Krush put any effort into the move -- so Qe6 won.

King and Kasparov immediately noted the move as a novelty -- Kasparov later admitted he felt he was playing for a draw after move 10 -- and from then on it was Krush's game. The game took a turn there, and only Krush had done the analysis necessary to keep the line going in the right direction. The other analysts posted their best guesses at the best moves, but Krush and the World Team message boards analyzed every position to death.

The message boards' involvement kept the game from turning into Kasparov vs. Krush, but many members of the World Team were unhappy with the dominance of Krush's ideas. They derided the World Team as sheep, blindly following Krush's suggestions without analyzing the positions themselves. For the most part, this was true; most of the World Team was unable to analyze positions on their own, but there were plenty of message board posters who agreed with her ideas.

According to King, The World (led by Krush) played very agressively after move 10, keeping Kasparov surprisingly on the defensive. Kasparov had the opportunity to force a draw around move 28, but he is a notoriously competitive person, and he chose to pursue the win. King says that 35 Kh1 was a surprise move by Kasparov that was probably the turning point of the game; from here on in, it was The World fighting for the draw instead of Kasparov.

Thanks to the exhaustive analyses by Krush's team and the folks on the message boards, The World made all the right moves and stayed in the game, but things started to break down in the low 50s. There arose a good amount of dissension among the regular posters about what to do next, and the analysts also disagreed. Krush's posted analyses (long since moved off onto her own web site for space considerations!) had grown so long that most stopped looking at them. Yet this was the crucial point of the match -- and no one could agree on what moves to take!

The result was that different analysts' suggestions were taken each move, and so no one coherent plan of play was executed -- it was a mishmash of different ideas, and so failed.

Then came move 58.

Krush and the message board posters had devised a last-ditch plan. It was a long shot, but Kasparov basically had the win at this point. There hadn't been a lot of time to analyze the position, but it appeared from all angles that it was the only hope. On October 13, 1999, Krush waited for Kasparov's move 58 to be e-mailed to her so that she could submit her own suggestion.

The e-mail, she claims, did not come on time, and her own suggestion also was not received by Microsoft in time to be posted on the site. Move 58 -- of all moves, the World's last possible chance at salvaging a draw -- and Irina Krush's suggestion was not posted. Since the other analysts were clueless about Krush's plan, their moves were basically marking time until the inevitable checkmate. And although Krush's planned move got a fair number of votes, it did not win a plurality.

The World Team message boards exploded with accusations that Microsoft had kept Krush's suggestion off the web page to get the game over with and give Kasparov the win. While the truth may never be known, it was an eerily suspicious coincidence.

In protest over losing their last chance at getting a draw, the World Team message board readers decided to resign -- except there was no resignation option! Microsoft had failed to anticipate the need for a Resign or Offer Draw choice on the submission form. After Kasparov's move 59, then, the protesters voted for the clearly losing move Qe1 (sacrificing the Queen for no gain).

Qe1 won a plurality of the votes, but Microsoft claimed that the 'ballot' had been stuffed, so to speak. Their FAQ on the endgame (http://www.zone.com/kasparov/endgamefaq.asp) notes that ballot stuffing had been occurring from the start of the game, but it had never affected the result. Nor had it ever been on such a large scale.

It seems likely that some unscrupulous World Team members did indeed cheat to ensure that Qe1 would be chosen, but it was justified in their minds by Microsoft's failure to post Krush's last-ditch suggestion for move 58. Regardless, Microsoft invalidated votes for Qe1, and Kb2 won instead.

Microsoft allowed a Resign option starting with move 60, although it took three plays for the World Team to choose it. The analysts -- Bacrot in particular -- bascially gave up after move 59, reflecting just how committed they were to The World's success (even Krush lost heart after her move 58 plan wasn't used), and the game ended with Kasparov predicting checkmate in 25 moves.

Kasparov claims now that move 58 didn't matter; he's analyzed Krush's proposed move and found that he still would have won. Certain members of the World Team still disagree, as they claim to have found lines that ensure a draw.

The game was excellent and the system worked fine for fifty moves or so. But as the endgame approached, things fell apart, probably due to the too-many-cooks problem. In the endgame, precise maneuvering is vital, and with four different analysts offering suggestions, it was inevitable that the World Team would take a wrong turn somewhere.

A rematch would be interesting, but some changes would have to be made. The trick would be to allow the World Team to have a coherent, single plan (particularly in the endgame) without taking away its ability to choose from among different suggestions. If the other three analysts had dropped out (or even better, supported Krush) after move 50, it's possible The World might have gotten a draw -- but then the endgame would have been Kasparov vs. Krush, not vs. The World. True, Krush was supported by the analyses on the message boards, but it was she who was directing the play.

All in all, it was a wonderful experiment. Krush gained some fame for herself, particularly thanks to (10 ... Qe6). Many regulars on the Message Boards formed friendships and found regular chess partners. The Message Board regulars even created a giant online thank-you card for Irina Krush, to thank her for the hours and hours of hard work she obviously put into the game. But all this greatness was marred by the events at the end, the details of which may never be known for certain.

See (http://www.zone.com/kasparov/) for Microsoft's official web site on the match.