April 10, 2012

Cognitive Science, Computer Science and Chess: Grandmaster Christiansen Visits C.U.

Print More

According to Grandmaster Larry Christiansen, chess is more than a strategy game — it’s a “mental war” involving sharp mental faculties and efficient cognitive processing.

Christiansen gave a simultaneous exhibition at a Cornell Chess Club event on March 30th.   At a simultaneous exhibition a highly ranked chess player plays multiple games at the same time with a number of different players.  In this event, Christiansen faced more than 20 opponents without suffering a single loss.  Prior to the exhibition, Christiansen shared a few secrets of the trade with other avid chess players.

Also present at the event were former Chess Club members Rob Weinberg ’75 and Frank Niro ’74.  The pair explained how chess increases mental acuity and improves logical thinking. Prof. Shimon Edelman, psychology, who was not at the event, explained a connection between chess and cognitive science. Chess benefits the individual because it involves decision-making and forethought  he said. Edelman related chess to the various computational aspects of cognitive science.

“In chess it all boils down to finding the best move, which is constrained by the rules of the game,” he said. According to Edelman, there are many logical faculties necessary in a game of chess. A game of chess begins with 20 possible moves and the number of moves increases significantly as the game progresses. The vast number of possible moves are comparable to the branches of a tree, Edelman said.

“The tree is rooted in the initial situation of 20 possible moves. Once the white piece makes its move, the number of possible moves is likewise affected. Each possible move is represented by a branch of the tree. You have to reason through all possible combinations of your moves and your opponent’s moves in order to get to the leaves of the tree, that is, the final outcome: a win or a loss,” Edelman said.

He said it is impractical to find every possible combination of moves at one particular point in the game because the number is so great. Skilled chess players implement the most successful plays by assessing the efficacy of each move in relation to the position of the other pieces at that particular moment. Players implement these series of moves through the use of short-cuts which are the same as how computers use shortcuts Edelman said.

“Thinking of and executing the move boils down to computing parts of this tree-structured space of possibilities, which in its entirety is too large to compute, and searching through it for optimal moves.”

In addition to the cognitive science aspects of chess there are computer science aspects of the strategy game as well. Computing may be at the foundation of chess in the modern world.

In 1996, IBM supercomputer Deep Blue and World Chess Champion Garry Kasparov played a six-game series of chess matches. Although Kasparov defeated Deep Blue with three out of four wins in the match, Deep Blue’s one victory marked the first time a machine won a chess game against a world chess champion.  A year later, Deep Blue and Kasparov had a rematch where the machine had an overall victory over the world champion.

Chess-playing computer software has improved dramatically since Deep Blue.   In 2009 a chess engine running on a mobile phone was able to reach the Grandmaster level, the highest title, apart from world champion, that a chess player can achieve. This software, called Pocket Fritz 4, was not faster than supercomputers like Deep Blue; it searched 20,000 positions per second compared to Deep Blue’s 200 million positions per second. Its higher performance can be attributed to smarter software instead of increased speed.

“Back in 1996, Deep Blue was the equivalent of Jeopardy’s IBM Watson supercomputer,” Edelman said. “Jeopardy is much harder than chess in the cognitive computational sense, which is illustrated by the fact that you still need a supercomputer to compete at human levels. In chess you no longer need a supercomputer to supersede human ability.”

Edelman also said that it is useful to compare chess to other board games in terms of understanding computational difficulty. “A human will never again beat the strongest computer in chess. A piece of software on a laptop could probably beat most of the human population,” Edelman said.

The only board games where humans still have the advantage over computers are those games with much larger search spaces, like Go. Go is an ancient Chinese board game known for being rich in strategy. The Go board is much larger than a chess board and the units are very uniform. The amount of possibilities are so large that Go computers are not yet up to human ability. But, according to Edelman, “it’s only a matter of time,” before computers catch up.

Niro, who is also a former executive director of the United States Chess Federation,  described how computers have come to redefine how the game of chess is played. Through the Internet, advanced strategies implemented by Grandmasters can be easily accessed by any individual who seeks to utilize the same tactics in their next game. It has revolutionized the way the world views chess.

Despite being more than a thousand years old, chess and the strategies used in playing it are still relevant in modern times. Jasper Wu ’14, president of the Cornell Chess Club, remarked that chess has a variety of practical applications, especially for learning about science.

Chess also involves analytical and strategic thinking. As in scientific inquiry, creativity and foresight are key for success. “Any chess player knows that you have to think ahead. It’s important to make a strategy and think things through,” Wu said.

According to Wu, the value in chess lies in the inevitability of making mistakes.

“It is impossible to never lose a game in chess, just as it is impossible to never make a mistake in life.” Wu said. “Chess teaches kids and players of all ages that making mistakes is a natural part of the game and that learning from those mistakes is what propels you to the next level.” He elaborated that science, like chess, is all about learning from mistakes.”

Original Author: Nasif Islam