Abstract
We advocate the study of a straightforward variant of chess, called twilight chess. It can be defined in a single sentence: pieces may be removed from the board and replaced later. The idea is to consider a variant rather close to standard chess, but at the same time more difficult (due to the combinatorial explosion) for computers. Our motivation is to stimulate research of new programming techniques. There are two advantages. First, only slight modifications are needed to adapt existing chess engines. Second, twilight chess is easily learnt by human players who can reuse most of their chess knowledge to master this variant. Twilight chess is thus a good candidate for a strategic game, closely related to chess, in which humans could be ahead of computers. Preliminary studies, based on the adaptation of F
Get full access to this article
View all access options for this article.
