Last week, the human world chess champion lost a six game match to a computer "Deep Fritz". This is the second time that a reigning world chess champion has lost to a machine. The last time it was Garry Kasparov (he lost to Deep Blue), this time it is Vladimir Kramnik. But last time when Kasparov lost to Deep Blue, the machine was a super computer with custom hardware for calculation. This time it was a much lowly workstation(4 processor Intel) machine. In addition, Kramnik had full access to the final version of the program a month in advance! During the game, Kramnik could also see the opening book used by Fritz, which literally ensured he never came out bad out of the opening. Yet he lost by a margin 4-2.
Let us do a more qualitative analysis of the games and come out with a score.
Game 1, Kramnik outplayed the machine and looked to have a winning ending. So 1-0! Actual result 1/2-1/2.
Game 2, Kramnik again outplayed the machine, got a very good position, but the game was always a draw. Kramnik missed a mate in 1 and lost. So 1/2-1/2. Actual result 0-1.
Game 3, Kramnik defends a difficult position against the machine, the game was a difficult draw for Kramnik. So 1/2-1/2. Actual result 1/2-1/2.
Game 4, same story as game 3. So 1/2-1/2. Actual result 1/2-1/2.
Game 5, an eaqual game with chances for both sides. The draw was a logical result. So 1/2-1/2. Actual result 1/2-1/2.
Game 6, Kramnik is outplayed by Fritz. Nothing more to say about the game. So 0-1. Actual result 0-1.
So qualitatively, the match was drawn. But if we proceed a bit further into this superficial analysis, Kramnik played an open sicilian in the last game. This is akin to chess suicide against a computer. He played this since his blunder in Game 2, cost him a point and he had to win the last game. Kudos to Kramnik, for playing this way for atleast a chance, than playing a quiet Petroff. If Kramnik were leading the match going into the last game, he would have definitely settled for a quiet queenless ending, with a draw being the most probable result.
So is the window of oppurtunity for a human to beat a computer closed? Is this the end of human computer matches. I don't think it is. This match has been a great success in terms of a scientific experiment. We should certainly have more such matches, preferably with a diverse set of human players with different playing styles. Maybe, the machine needs to be handicapped a bit in future. This can be a bit embarassing for the human GM's, but this is the reality. With the way hardware, and chess software are progressing, humans will not be able to win a single game against the machine. Some form of handicaps are required to make the matches more interesting, and to give the human player a chance!
