reviewed by Keith Smillie
In August 1994, the computer program Chinook, written by a team headed by Jonathan Schaeffer of the Department of Computing Science at the University of Alberta, won the World Man-Machine Checkers Championship at a contest held at the Computer Museum in Boston. Although this was the culmination of six years of intense effort on the part of the Chinook team, it was nevertheless a bittersweet victory. The Chinook team had won by the resignation for health reasons of the reigning world champion, the almost legendary checkers champion Marion Tinsley, a 67-year-old mathematics professor and devout born-again Christian. A return match was discussed but never held, since eight months later, Tinsley was dead of cancer.
Schaeffer had played chess since childhood and entered his first tournament at 14. He had been involved in the design and implementation of computer chess programs since his undergraduate days at the University of Toronto, and he had completed both his Msc and PhD degrees in this area. In 1988, he was readying his chess program Phoenix for the World Chess Computer Championship in Edmonton, Canada, the following year. However, the lunch-time comment "Jonathan, whatever happened to computer checkers?" from two of his colleagues aroused his interest in the game, and he immediately began to read what was available on the subject and to implement his own ideas. When Phoenix failed to live up to his expectations in the chess championship and finished "in the middle of the pack," he began work in earnest on computer checkers.
One Jump Ahead is Professor Schaeffer's intensely personal account of the work of the Chinook team of academics, technical staff, and graduate students who wrote, extended, and refined their checkers program. One Jump Ahead is also an account of their obsession to beat Tinsley. (One measure of just how personal an account he has written is indicated by the 40-some references in the index to his wife, Stephanie, and about half that number to their young daughter, Rebecca). Through his eyes, we experience vicariously the joys and sorrows, the exhilaration of success and the disappointment of failure, the effects of intense prolonged periods of work, and the concomitant stresses placed on family, colleagues, and friends. The informal and at times casual style of writing, while not expected and possibly not desired in most professional writing, helps increase the realism of the book. Although this is primarily a book about the human aspect of the game of checkers and checker- playing programs, the technical aspects are not neglected, although they never get in the way of the story. The limited history of checker programs is covered well, both in the text and the references, and an excellent account is given of Arthur Samuel and his pioneering work.
For those persons who do not live in the western part of Canada or the United States, we might remark that a "Chinook," according to the American Heritage Dictionary, is "a warm, dry wind that descends from the eastern slopes of the Rocky Mountains causing a rapid rise in temperature." One Jump Ahead may be read with pleasure by computer and game-playing specialists and also by those with no background in either subject. It might be recommended as required reading for those persons contemplating a scientific career in academia and who would like a glimpse of what scientific research, with all the warts and pimples, can be like. Let us hope that Professor Schaeffer's book will have the same wide appeal as did Tracy Kidder's The Soul of a New Machine, which was in some ways a model in the early 1980s.
University of Alberta