Author: Don Dailey
Date: 20:26:27 12/08/97
Go up one level in this thread
>Hi Bob, > >I wrote a little program to look at the time distributions by move >number of using 1/30 vs 1/25. Based on this, I think 1/25 does >make more sense, especially with move predictions and such. > >This chart assumes the Dutch time control of game in 90 minutes. >Surprising to me is that at move 40, there is only a small difference >in the two numbers. > >-- Don > > > MOVE 1/30 1/25 > ---- ------ ------ > 1 180.0 216.0 > 2 174.0 207.4 > 3 168.2 199.1 > 4 162.6 191.1 > 5 157.2 183.5 > 10 132.7 149.6 > 15 112.0 122.0 > 20 94.5 99.5 > 25 79.8 81.1 > 30 67.3 66.1 > 35 56.8 53.9 > 40 48.0 44.0 > 45 40.5 35.8 > 50 34.2 29.2 > 55 28.9 23.8 > 60 24.4 19.4 > 65 20.6 15.8 > 70 17.4 12.9 > 75 14.6 10.5 > 80 12.4 8.6 > 85 10.4 7.0 > 90 8.8 5.7 > 95 7.4 4.7 > 100 6.3 3.8 To further expand on this data: What I like about 1/25 over 1/30 is the fact that up to about move 30, where both methods are using about the same amount of time, the version with 1/25 has spent significantly more time on each move, ie is playing better chess if you will. But even up to move 40 or so, the difference is small, and the odds of the game being over by this time are getting fairly high. I think I agree with your earlier comment that it might even make sense to be more aggressive than this. Especially if you are using the "dumb" version (without my enhancement ideas :-) and considering the fact that tooting and book moves will tend to extend you a few more moves into this table. -- Don
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.