Author: Robert Hyatt
Date: 12:59:59 12/01/03
Go up one level in this thread
On December 01, 2003 at 14:41:35, Dieter Buerssner wrote:
>On December 01, 2003 at 12:39:13, Robert Hyatt wrote:
>
>>The server is the final arbiter with respect to time, draws, wins and losses.
>
>One might note, that ICC is also no perfect arbiter. I just played this little
>game:
Yeah, but which would you rather have? ICC or Jaap? At least every time
you have a KN you will be flagged for insufficient material, It won't change
from year to year. :)
Everyone has been aware of that. I've even put a catch in crafty to look
for just that case, as my "draw" detection doesn't kill me there since the
search result (mate) is found before the terminal position (evaluation = draw)
happens.
I've never caught ICC zapping me there, which is not surprising.
I looked at my logs and the time is definitely given to me in
hundredths of a second, but with the last 2 digits always zero. I can't
imagine that being a problem except near the end of a zero increment game.
And I don't recall ever losing even a 1 0 game on time unless I had
introduced some sort of bug.
>
>[Event "ICC w20 1 5 u"]
>[Site "Internet Chess Club"]
>[Date "2003.12.01"]
>[Round "-"]
>[White "guest773"]
>[Black "guest1644"]
>[Result "1/2-1/2"]
>[ICCResult "Black ran out of time and White has no material to mate"]
>[Time "14:13:22"]
>[TimeControl "60+5"]
>[SetUp "1"]
>[FEN "8/8/8/4N3/8/7p/8/5K1k w KQkq - 0 1"]
>
>1. Ng4 {Black ran out of time and White has no material to mate} 1/2-1/2
>
>Note, that white has a forced mate. I doubt, that it will happen often in real
>games, but I remember one game, where my engine could have won with a single
>knight (it preferred to draw, unfortunately).
>
>One other thing, I recognized with ICC - it reports rounded times to WB (rounded
>to the second). So, you may have 8.6 seconds on the clock (which for example is
>displayed correctly in Blitzin), but WB reports 9 seconds to the engine. When
>the engine does not use a sufficient saftey margin, this can easily yield in a
>loss on time. If ICC decides to report only seconds to WB, it should truncate
>the time, not round it. I investigated some WB-logs (which has time stamps),
>where the time between receiving the move from the opponent, and the time
>between WB sent the move back to ICC was 0.3 seconds less than the time left
>reported by WB - game ended in time forfeit (timestamp was running).
>
>Regards,
>Dieter
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.