Author: Harald Faber
Date: 05:20:19 07/13/99
Go up one level in this thread
On July 13, 1999 at 07:21:54, Shep wrote:
>>>But given their game scores so far, it is highly unlikely CM6 or CST will end up
>>>on #1...
>>
>>This is your criteria?? No (suposed!) chance for getting no.1 so games are
>>stopped earlier?
>
>Sometimes, yes. The last time we had this discussion was on last year's SC in
>the game Rebel vs. Virtual. After the tourney, I decided it was not worth
>resuming the match (adjudicated a draw) since Rebel wouldn't have changed its
>position and Virtual would only have gotten from #10 to #9 in case of a win.
You astonish me on and on.
AFTER the tourney you decide to continue the interrupted game or not, based on
the ranking and effects a different result woud have???
>>>I interpret successive displays of "+0,00" as the program's way of offering a
>>>draw.
>>
>>Maybe they will play a repitition but since there is none in sight, why not make
>>some more moves?
>
>I came up with this rule when programs didn't have an "offer draw" possibility.
>Maybe it is outdated; I will ponder your arguments for future tournaments.
>Even more so since my next tournament will be all-amateur, where showing "+3" is
>not a guarantee for being able to win... :)
Not only for amateurs.
Best example at this moment is Bobs game he posted Crafty-vs-Loek van Wely.
Look at it and the -4.xx eval!
>But changing the rules while a tournament is running is a no-no for me.
OK.
>>OK, this is your tournament.
>
>I just have to "optimize" some decisions because I cannot play until after
>midnight, and some programs don't allow me to save the game including the clock
>settings, and some wreak havoc if you try to adjust the clocks (Rebel is a good
>example for both).
I know and that is why I set the time control either 40/120+g/60 or if not
possible g/180.
>>>I have been using the draw rule as well as the win rule ("A game is won if both
>>>sides agree that one side is up by +3 or more for 3 moves in a row, or by +5 or
>>>more in the endgame") successfully for years and am quite happy with it. :)
>>
>>Then you missed some games where such games were NOT won.
>
>Yes, but as they always say, that is "not statistically significant". :)
ONE may not be statistical relevant. BTW a 9 round-robin isn't either...
But if you adjust say 3 or more games for a program in the tourney the result
can be VERY different.
>I am already gradually changing this rule because of the improved endgame
>performance of most programs and the endgame tablebase issue.
>Actually, I _have_ played on in some of these cases and have only seen one game
>where Rebel was up +3 against Genius and did not win (but this was a rapid
>game).
I know such strange games/evals.
>>>Besides, don't forget these are _manually played_ games! It gets really boring
>>>for me to watch them play on when both I and the programs know that nothing new
>>>is going to happen.
>>>Shep
>>
>>I know that it is boring and annoying but you CAN get wrong results when
>>stopping games when they have almost just begun...
>
>Agreed. The problem is finding a good compromise. But I will try. :)
>---
>Shep
Fine, finally we agree. :-)
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.