Kriegspiel on ICC
This is to let you know that the Internet Chess Club (http://www.chessclub.com) supports kriegspiel (and has for about 18 months) and automates the referee's task. The ICC rules are pretty close to the "Cincinnati style" rules you give. In fact, I am considering taking the time to fix the ICC to match those Cincinnati rules, at least with regard to pawn tries (ICC gives the number of available pawn-capturing moves, but it would be a simple to just say whether you have any or not instead.) I might consider adding the British rules also, but I find asking "any" every move annoying that I have trouble believing that anybody would prefer those rules.
A major sticky point in the computer implementation (and indeed, for a human referee) is when to say "no" and when to say "nonsense". E.g. what if Black is in file check and tries to move a piece (other than the king) to a different file than the one the king is on? That could be looked for and called "nonsense". Or if you're not in check, and you try Ra1a3 and told "no", can you then try Ra1a4 and be told "no" again. There are all kinds of inferences that the referee might decide you should have made! It's kind of interesting to try to make inferences from hearing the "no" calls when your opponent is trying to move, but that can of worms seems large. The ICC currently just says "illegal move" to the mover (with that wording, because some of the old clients like xboard match that pattern), and nothing at all to the opponent, and the mover's clock is running the whole time.
Written by Bert Enderton (fishbait on ICC); introduction by Hans Bodlaender.
WWW page created: July 8, 1997.