Monday, 27 October 2014

Chess has been solved, apparently.

"Person invents improved version of chess" is a familiar and yet somehow boring headline that I see quite often these days. The latest version that this headline has been applied to is XYQ4, which seems to deal with the issue of reliance on learning and memory by making everything random.
But what is truly special about this version it that it seems to have been built on an entire erroneous basis. In this article the games creator, Damien Sommer, thinks chess is broken because "Chess is either a draw or White wins. There is no situation where Black wins". Now this claim fails on both a practical level, where Black wins quite a number of chess games that I have witnessed, and on a theoretical one. No such proof exists of the outcome of a perfectly game of chess, and for Sommer (or the reporter) to make a claim like this is difficult to understand. This is even pointed out in the comments to the liked article!
I suspect this variant will go the way of the other versions of "New Chess", attracting a brief level of interest before fading away. But having said that, I have a version you might be interested in ....

1 comment:

Anonymous said...

The argument I believe is that "as chess is a finite game and has no uncertainty, it could be solved backwards. In other words, if players had perfect foresight, it would end up in either a draw or a win for white."

For this reason, from game theory point of view chess is trivial. However, the assumption of perfect foresight is not realistic with chess. A better assumption would be the assumption of bounded rationality.