Author: Chris Whittington
Date: 07:20:07 10/14/97
Go up one level in this thread
On October 14, 1997 at 09:35:57, Ulrich Tuerke wrote: >On October 14, 1997 at 09:00:13, Chris Whittington wrote: > >> >>On October 14, 1997 at 08:38:41, Robert Hyatt wrote: >> >>> >>> >>>I responded to Tony also. I suggested the following modification: >>>Since >>>it is not uncommon for the operator to fudge the clock by some finite >>>amount to allow time for forgetting to hit the clock or for technical >>>difficulties, I suggest that at the beginning of the match, the two >>>operators >>>simply inform each other of the "fudge" (IE 5 minutes) and that this be >>>enforced for *all* clock updates for that program. IE if we say our >>>fudge >>>is 5 minutes, then each correction must bring the clock back to the real >>>clock - 5 minutes, no exceptions. >>> >>>In the case of Crafty, I have this built in. so that the operator >>>doesn't >>>have to enter a false time, we simply set an operator overhead which >>>Crafty >>>dutifully removes from the clock before calculating target times. But I >>>don't >>>see why we would object if our opponent hasn't done this and wants to >>>"lie" >>>to the program, just so he consistently lies by exactly the same amount >>>for >>>each clock update... >> >>Any fudge allowance opens up the possibility of cheating in some way you >>haven't yet dreamed of. >> >>programs should enter only the REAL time (+/- 30 secs at most). >> >>Any desired fudge can be done within the program. Most do some sort of >>operator fudge anyway, as you point out with Crafty. If they don't, well >>they should do; just like they should save the game as PGN and show the >>move on the screen and and and. >> >>Chris Whittington >> > >I would prefer Bob's solution. Why not agree on a constant operator time >prior to start of the game. This time has to be subtracted from the >remaining time on the physical clock; and that is the time to be >entered. This is a unique recipe where I do not see the chance for >manipulation. >To enter the remaining time +/- max 30 seconds on the other hand would >leave some space to manipulate. I say +/- 30 secs because the wooden chess clocks can't be read to greater accuracy; but you can turn the 30 secs into 0 secs if that helps. > Furthermore, there are programs not >accounting for an operator time (as mine); entering the remaining time >without a reserve would soon lead to the next deviation for these >programs (and put a lot of stress on the operators). Its very easy to mod your source code for the time controller. Just take 5 minutes or whatever you want from the internal clock time before you use it for calculating your time per move allowance ...... ? Chris Whittington > >Regards, Uli >
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.