Some Studies in Machine Learning Using the Game of Checkers

Samuel described a program that adjusts its position evaluation from the results of its own games and plays better than its author.

Why it matters

The phrase machine learning appears in print here, alongside a demonstration that a program can outplay the person who wrote it.

The paper describes two mechanisms: memorising evaluated positions, and correcting the weights of the evaluation function from the difference between the value of the current and the next position. The second is what was later named temporal-difference learning. The program played itself to gather experience.

Event record

Event date

July 1959

Timeline date

Event date

Verification

Sources gathered automatically · September 17, 2026

Lines

ID

evt-0088

The July 1959 issue of the IBM Journal of Research and Development.

Sources

primary Some studies in machine learning using the game of checkers
IBM Journal of Research and Development · Published July 1959
archive Some Studies in Machine Learning Using the Game of Checkers (full text)
University of Virginia · Published July 1959

Related events

Earlier

Related A machine that mutates its own program 1958links to this record
Samuel in 1959 teaches a program by adjusting the coefficients of an evaluation from the results of games; Friedberg in 1958 changes the instructions themselves from a success signal. Both appeared in the same journal; neither cites the other in what was read.
Related Dynamic Programming 1957
Updating a position's value from the next position matches Bellman's decomposition.
Extends Samuel's checkers on television 1956
The paper describes the learning mechanism of the program shown on television in 1956.
- Some studies in machine learning using the game of checkers · IBM Journal of Research and Development

Later

Extends Samuel's signature tables 1967links to this record
The paper is titled «II - Recent Progress» and continues the work of 1959: its reference list names the first part (IBM Journal 3, 211-229).
- A. L. Samuel, «Some Studies in Machine Learning Using the Game of Checkers. II - Recent Progress», IBM Journal of Research and Development 11(6), 601-617, November 1967 (DOI 10.1147/rd.116.0601) · International Business Machines Corporation (IBM Journal of Research and Development)
Builds on Actor and critic 1983links to this record
Updating a value from the difference between successive states comes from Samuel's program.
- Neuronlike adaptive elements that can solve difficult learning control problems · IEEE Transactions on Systems, Man, and Cybernetics
Builds on Temporal-difference learning 1988links to this record
It traces its rule directly to what Samuel's checkers program did.
- Learning to Predict by the Methods of Temporal Differences · Machine Learning
Related TD-Gammon 1992links to this record
Self-play is taken to a level Samuel's program never reached.
Related Chinook becomes world champion 1994links to this record
The same game Samuel started on, taken to a world title.