Author: Robert Hyatt
Date: 12:15:56 01/31/03
Go up one level in this thread
On January 31, 2003 at 14:29:05, Sune Fischer wrote: >On January 31, 2003 at 13:19:44, Robert Hyatt wrote: > >>On January 31, 2003 at 06:01:26, Sune Fischer wrote: >> >>>I've noticed people (eg. Crafty) don't reset the 50 move counter when ep. flags >>>and castle flags are destroyed. >>> >>>I wonder if that is correct, can one really say nothing has happened for 50 >>>moves if flags has been destroyed? >>>I would bet this is a "bug". >>> >>>Anyone know what the FIDE rules say about this? >>> >>>-S. >> >> >>The fide rule doesn't have to say anything. If the ep status changes, then a >>pawn _must_ have >>been moved. :) That would have reset the counter immediately. > >True, I was adressing the _next_ move, the one that destroys the ep flag. >Ie. if one should keep from incrementing the counter on this the next move after >the pawn move (it must be 0 because a pawn just moved). I suspect it was not mentioned for that very reason. The _next_ move would either be an EP capture or the EP status would clear, if the pawn advanced two squares and passed an enemy pawn in doing so.... It would hardly change anything. > >The point would be, that according to the repetition there is a difference; If >you return to this position down the tree it isn't the same because the flag is >no longer set. Yes. It would be possible to reach the same position three times, with no repetition, twice where you could castle, once where you could not. But I simply loop back using the 50 move counter as I hash castling status so it won't be a false draw anyway, only a bit more effort (longer loop). > >This means (not that anyone is doing it I think), that you can't detect 50 move >draws by tracking back to the last "irreversable" move, since not _all_ >irreversable moves will reset the counter! Right... > >However you only need to do the repetition check back to the last irreversable >move. So in affect you need two counters here, one for 50 move rule and one for >irreversable moves (for the repetition)! >Not that it really matters much, the keys shouldn't collide just because you >search back a few more moves than needed. Right. A matter of efficiency. The rarity of that happening is not worth the extra work in maintaining two counters... > >Quite a mess, but I guess the rules are easier for the players this way. > >-S. > >>Castling is irrelevant. The rule says "no capture or pawn move for 50 moves." >>It doesn't >>mention castling status. EP is included since to have an EP capture you had to >>move a pawn >>and that resets the counter.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.