The Death of Checkers at the Hands of a Machine Called Chinook

Print Story | Email Story
Williamstown - Jonathan Schaeffer, professor and chair of computer science at the University of Alberta, has solved the game of checkers. As the creator of the Chinook project, which began with the goal of designing a computer to win the human World Checkers Championship, Schaeffer will tell his story of the man- versus-machine battle for checkers supremacy on Thursday, Feb. 7. The lecture will begin at 8 p.m. in Wege Auditorium in the Science Quad on the Williams campus. The event is free and open to the public.

In 1989, Schaeffer and his team set out to beat the reigning human checkers champion, Marion Tinsely, not by training themselves but by programming a computer. Because Tinsely was as close to perfection at the game as was humanly possible, the computer had to be perfect. In other words, Schaeffer, or more precisely, his computer program named Chinook, had to "solve" checkers.

It took 18 years to achieve the goal of unraveling the mysteries of checkers, and it was filled with ups and downs.

In 1990, Chinook became the first program in any game to win the right to play for a human World Championship, but in 1992, it lost the championship match. Rebounding from its defeat, Chinook was victorious in 1994, and by 1996, it had become clear that the program was superior to any human and it was "retired."

Despite Chinook's retirement, Schaeffer was not satisfied with his program's dominance. He wanted to understand checkers to its core. Consequentially, for 10 years of post-retirement work, Chinook not only beat its opponents but also beat the game itself.  On April 29, 2007, Schaeffer and his team announced that checkers is now solved.

In essence, Schaeffer and his colleagues at the University of Alberta have proved that Chinook cannot ever lose. The best that an opponent can achieve is a draw. Indeed, they argue that if two opponents play as perfectly as humanly possible, the outcome will always be a draw, despite the 500 billion billion theoretically possible board positions. That astronomical number of possibilities makes checkers the most complex game that has been solved to date.

As Schaeffer told The New York Times, "from my point of view, thank God it's over." Schaeffer now has his eyes set on mastering poker. Will Texas Hold-em be the next, beloved game to be conquered by a machine? Time will tell. Schaeffer admits that even if he "solves" poker, "chess won't be solved in my lifetime."

For those checker enthusiasts out there determined that a machine named Chinook has not spelled the death of human dominance, you can test your skills at http://www.cs.ualberta.ca/~chinook).
If you would like to contribute information on this article, contact us at info@iberkshires.com.

Williamstown Planning Board Narrowing in on Subdivision Bylaw Changes

By Stephen DravisiBerkshires Staff
WILLIAMSTOWN, Mass. — The Planning Board late last month discussed specific features of what it plans to pass as a new subdivision control bylaw this year.
 
The board long has discussed the complex set of regulations as being out of date and cumbersome to both potential developers and the board itself, which has needed to hear requests for waivers of outdated rules for the handful of residential subdivisions that have been proposed in town in recent years.
 
This spring, the town engaged consultants from Northampton's Dodson and Flinker Landscape Architecture and Planning to go through the existing bylaw, compare it to more contemporary regulations in other communities and help craft a revised bylaw.
 
Unlike the zoning bylaw, where amendments require approval of town meeting, the subdivision control bylaw is a creation of the Planning Board, which can make changes on its own after a public hearing process it hopes to complete this year.
 
At a special Planning Board meeting on May 26, Dillon Sussman of Dodson and Flinker and his colleagues walked the board through a dozen different decision points that the board must resolve — either by leaving the bylaw as is or making a change — and offered suggestions based on best practices.
 
All of the issues are technical and ranged from the fundamental, like how the bylaw will define types of subdivisions, to the highly specific, like what turning radii will be required in new streets that are constructed to serve planned developments.
 
One example of a topic that came up in the recent approval of a four-home subdivision off Summer Street is stormwater management.
 
View Full Story

More Williamstown Stories