# Checkers Programs

Checkers programs are artificial intelligence systems designed to play checkers (draughts), with Chinook being the most notable example as the first computer program to win a world championship title against humans and later solve the game completely.

Checkers programs are computer software designed to play the board game checkers, also known as draughts. These programs use algorithmic techniques to evaluate positions, search for optimal moves, and compete against human players or other computers. The most prominent checkers program is Chinook, developed at the [University of Alberta](https://www.wikiprompt.org/wiki/university-of-toronto) (though the project was based there, not Toronto) by a team led by Jonathan Schaeffer, which achieved historic milestones in both competitive play and game-solving.

Chinook was developed between 1989 and 2007 by a team at the University of Alberta, consisting of Jonathan Schaeffer, Rob Lake, Paul Lu, Martin Bryant, and Norman Treloar. Unlike modern approaches that rely on [machine learning](https://www.wikiprompt.org/wiki/machine-learning), Chinook's knowledge was entirely hand-coded by its creators, using traditional [artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence) techniques such as search algorithms and evaluation functions. This distinction highlights the evolution of AI from rule-based systems to data-driven methods.

## Man vs. Machine World Champion

In 1990, Chinook earned the right to compete in the human World Championship by finishing second to Marion Tinsley in the US Nationals. Initially, the American Checkers Federation and English Draughts Association opposed computer participation in human championships. When Tinsley resigned his title in protest, these organizations created the new title Man vs. Machine World Championship, allowing the competition to proceed. Tinsley won the first match with four wins, two losses, and 33 draws.

A 1994 rematch resulted in Chinook being declared the Man-Machine World Champion after six drawn games, as Tinsley withdrew due to pancreatic cancer. This made Chinook the first computer program to win a world championship title against humans, though it never defeated Tinsley, who was significantly superior to even his closest peers. In 1995, Chinook defended its title against Don Lafferty in a 32-game match, winning 1-0 with 31 draws. After this, Schaeffer decided to stop competitive play and focus on solving checkers entirely.

## Algorithm

Chinook's algorithm comprised four main components: an opening book containing moves from grandmaster games, a deep search algorithm for exploring possible moves, a move evaluation function, and an end-game database covering all positions with eight pieces or fewer. The evaluation function was a linear handcrafted function considering features such as piece count, king count, trapped kings, turn, and runaway checkers (pieces with unimpeded paths to become kings). All knowledge was programmed by the creators rather than learned, contrasting with later AI systems like [DeepMind](https://www.wikiprompt.org/wiki/google-deepmind)'s approaches.

## Timeline and Solving Checkers

Key milestones in Chinook's development include: 1997, when Schaeffer published the book "One Jump Ahead: Challenging Human Supremacy in Checkers" (updated in November 2008); May 24, 2003, when the 10-piece database (5 pieces per side) was completed; and August 2, 2004, when the opening "White Doctor" (10-14 22-18 12-16) was proven a draw. Further openings were proven draws on January 18, 2006 (09-13 21-17 05-09) and April 18, 2006 (09-13 22-17 13-22).

On March 10, 2007, Schaeffer announced at the ACM SIGCSE conference that a final solution was expected within 3-5 months. On July 19, 2007, the journal Science published "Checkers Is Solved," proving that with perfect play, the best outcome is a draw. This made checkers the first major board game to be solved completely, following earlier work on simpler games. At the time of its competitive peak, Chinook was rated at 2814 Elo.

## Legacy

Chinook's success demonstrated the power of classical AI techniques, including deep search and handcrafted evaluation, before the rise of [deep learning](https://www.wikiprompt.org/wiki/deep-learning). Its solution of checkers remains a landmark achievement in game AI, comparable to [chess computers](https://www.wikiprompt.org/wiki/chess-computer) like Deep Blue. While modern checkers programs might use [neural networks](https://www.wikiprompt.org/wiki/neural-network), Chinook's approach influenced subsequent research in game theory and combinatorial game solving, and its methods are still studied in AI courses at institutions like [Carnegie Mellon University](https://www.wikiprompt.org/wiki/carnegie-mellon-university) and [MIT CSAIL](https://www.wikiprompt.org/wiki/mit-csail).

---
Source: https://www.wikiprompt.org/wiki/checkers-programs
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-13T03:54:53.071562+00:00