Author: Dave Gomboc
Date: 01:35:48 08/18/99
Go up one level in this thread
On August 18, 1999 at 03:18:35, Shep wrote: >On August 17, 1999 at 22:22:25, James T. Walker wrote: > > >>Pardon me, but your answer is absolutely silly. There is no "Operator time" in >>auto232 games. CSTal-2 keeps it's own time and ran out of time on it's own >>clock. I would never score a win for one program because the other program said >>it ran out of time. In auto232 games the time lost is always charged to the >>opponent but there is NO claim of victory by the other computer. They continue >>to play the game untill the end or human intervention stops them. I stopped the >>game because CSTal-2 by it's own clock ran over the time control by about 20 >>seconds. This is a common problem with CSTal-2 and one that is most hated by >>me. It loses games on ICC for the same reason. It knows what the time is but >>continues to search anyway. If you have never seen this happen then you have >>not seen many games by CSTal-2. This is the first time I have seen it at 40/2 >>but I have only played two games at 40/2. Usually it only does this when the >>game is a close struggle. This is one of the few times I've seen it do this >>when easily winning. >>Jim Walker > >I had one case of CST violating the time control at 40/120 myself so far, but >didn't care too much. > >Here's what Chris W. himself told me about this issue: > >"If I play CSTal under tournament conditions, I usually give it a few percent > less time than the tourney. > 40 in 120, I'ld do at 40 in 110 for example. > I never designed it to play hyper accurate to the time controller, not > thinking this would be an issue for sensible people." >--- >Shep That's an interesting perspective, though it's not one that's going to score too many points with chessplayers who claim flags when their opponents don't make time control. Dave
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.