Author: Dieter Buerssner
Date: 08:45:05 08/27/01
Go up one level in this thread
On August 26, 2001 at 18:05:45, Uri Blass wrote: >On August 26, 2001 at 16:30:44, Dieter Buerssner wrote: >> >>When is the clock started? At move 1 or when the first engine is out of book? Thanks to Michael for answering my question. >I think that at move 1 is the fair way to play games. > >The engine may calculate for a short time even when it is in book in order to >avoid possible stupid blunders in the book or in order to choose the book move >that it plays. I agree, that when an engine behaves as you suggest, it would be the only fair way. At the tournament I participated, none of my opponents behaved like this. They all moved instantly, while in the book. I think, the intention of the rule to start the clock only when the first program is out of book under these cirumstances is clear. Operator skill does play a much smaller role. Especially, when both move instantly, and you have to write down your moves, a skilled operator can save quite some time. (For example, there is no time to think about, what you have to do next). Besides what you have mentioned, another small point of unfairness might be, when an engine starts to ponder, while it is still in book. I am not sure, how common this is. >I believe that playing moves that were not checked manually only because they >are in book is a mistake without some blunder checking is a mistake. > >I do not suggest to use a long time to analyze them but I saw in the past cases >when programs did stupid blunders that could be avoided by a search of less than >1 second because of a bad book and even if these cases are only 1 out of 1000 it >is better to care that they will never happen in tournament time control games >if the price is 1 second of search per move. I think, wasting one second is no problem at all. But then, how to filter out obvious blunders? Just when after one second the score is below say -3 or -5. I believe, one might find a considerable ammount of positions in books, where this is the case, and still the position is unclear of even winning. Perhaps you remember one game of Junior - Yace, that was discussed here some time ago. Both engines agreed, that the position was worse than -3 after out of book (in a long time control game on a fast computer). Still the position was a surprising forced win for the other side (this was the conclusion of the discussion here) Your comment "a good choice of book moves", or similar. The only case, I could see, where such a fast check of book moves is totally uncritical, is when a forced mate is found. This however proably won't happen often. 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.