ICGA Journal: The Journal of the Computer Games Community

It is argued that an extremely important requirement for doing an effective selective search is explicitly distinguishing different degrees of reliability of evaluations arising in the search. A suitable mechanism for administering the information for multiple levels of reliability is given, and comparisons with other algorithms are made.

This paper draws certain conclusions about how long it may take before the strongest brute-force chess programs are capable of winning a match against the human World Champion.