Author: Robert Hyatt
Date: 21:19:50 11/28/01
Go up one level in this thread
On November 28, 2001 at 18:48:39, Tom Kerrigan wrote: >On November 28, 2001 at 13:47:11, Robert Hyatt wrote: > >>On November 28, 2001 at 03:38:59, Tom Kerrigan wrote: >> >>>On November 27, 2001 at 12:25:01, Robert Hyatt wrote: >>> >>>>Deep mistakes are _common_ from a computer's perspective. You might think >>>>they are "tactically supreme" but there are _lots_ of endgame positions where >>>>they are clueless. And any time you reach such a position, where either it is >>>>(commonly) a long-range kingside attack (or less commonly) a deep endgame >>>>tactic, you get random reports. >>>> >>>>I simply personlly don't like anything "random" when somebody/something is >>>>supposed to be telling me where I went wrong. >>> >>>Well, without any more information than your gut feelings, I will continue to >>>believe that the depth of mistakes is evenly distributed. >>> >>>-Tom >> >> >>I believe that is what _I_ said. Some are one move beyond the search >>horizon, some are 20+ moves beyond. Isn't that "distributed". And the > >Distribution between ~13 and infinity is not "even." Duh. > >-Tom Duh always strengthens your argument? think about this. Evenly distributed between 1 and 20 means what? an average of 2? Or an average of _ten_. duh... indeed... But the duh is on the "other foot"... And that is being gracious, because as I said, the purpose of back-to-front is not to find problems one or two moves sooner, but much more than that. So maybe 5 to 20 is a better sample, with 20 being a rather low upper bound.
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.