Chess notes
By Harold Dondis and Patrick Wolff
May 30, 2011
There have been a number of intriguing chess-playing machines that were fraudulent contraptions. The first one was “The Turk,’’ constructed in 1770 by Wolfgang von Kempelen for the benefit of the Empress Maria Theresa of Austria-Hungary. It won most of its games in which it was used over 84 years — even defeating Benjamin Franklin and Napoleon Bonaparte. It took the great French player Philidor to defeat it. Edgar Allan Poe guessed correctly that an accomplished player was hidden inside, though he did not guess the exact arrangement correctly. A second chess machine using a player hidden inside was Ajeeb, constructed by Charles Hooper in 1868, starring, among others, the hidden Harry Nelson Pillsbury of Somerville. Ajeeb is said to have defeated Theodore Roosevelt and Harry Houdini. The Turk and Ajeeb were improved upon by the creation in 1876 of Mephisto, which was remotely operated by the accomplished player Isidor Gunsberg.
In the 20th century along came the chess computer, certain to beat human beings due to its ability to look at 200 million moves per second and run them out on an analogous chessboard. Grandmasters were naive in their attempts to beat it, although of course they were paid for the attempt.
The latest in omnipotent human chess achievers to be reported is Andriy Slyusarchuk, an amateur player and medical doctor who allegedly defeated the world champion chess program Rybka4 in a two-game blindfold public demonstration in his home country of Ukraine. The event was widely cited in news reports. Slyusarchuk claims to be a genius, able to memorize 20,000 books. He said he had studied 3,000 books on computer chess in preparation for the match with Rybka. Conceding that Rybka was not set at maximum playing strength, we do not need to be Edgar Allan Poe to question the validity of this demonstration.
We cannot swallow the 3,000 books claim but in any event it would not come close to achieving all variations necessary to beat a computer. Most books do not go far into solving chess openings. A computer does not need them. What is most likely is that our new genius used a won game and put the computer in a lost position where it would make the desired moves. Still, Rybka sometimes changes its recommendations, and there must have been some way to control it. Or maybe there is some other explanation.
We must conclude that the so-called Memory Man, who did not know how to castle or move a knight, is the latest in ingenious chess frauds.
Source: http://www.boston.com
The most likely explanation for this is that he memorized some of the losses by the program to another computer program, and simply replayed those against Rybka. If he wanted to validate his winning, I would suggest that he allow a randomly selected a five year old checkers player be allowed to make Rybka’s move every 10 moves and see how he fares then.