Author: Uri Blass
Date: 11:18:27 11/23/01
Go up one level in this thread
On November 23, 2001 at 10:52:58, Robert Hyatt wrote:
>On November 22, 2001 at 23:16:41, Dave Gomboc wrote:
>
>>On November 22, 2001 at 09:42:36, Robert Hyatt wrote:
>>
>>>On November 22, 2001 at 05:27:05, Gordon Rattray wrote:
>>>
>>>>The Fritz GUI analyses games ("Full Analysis") by starting at the end of the
>>>>game and retracting moves. How does this compare to going forwards? Does it
>>>>produce better results?
>>>>
>>>>I think this issue has been discussed before, but my search has failed to find
>>>>anything. Please feel free to forward me a past link if appropriate.
>>>>
>>>>Gordon
>>>
>>>
>>>Here is the idea...
>>>
>>>If you start at the end of the game, you load the hash table with stuff
>>>that will help as you search at earlier moves... with the "idea" that
>>>earlier analysis will be more accurate since it will have access to these
>>>scores.
>>>
>>>It doesn't work however.
>>>
>>>IE pick three points in the game, (a) where a key mistake is made, (b) a
>>>position further into the game, and (c) a position near the end where the
>>>program can see that it is lost. As you search backward, when you reach
>>>(b) the search might well _still_ see that it is lost, because of the persistent
>>>hash entries that help. But when you back up past (b) eventually the
>>>hash entries get replaced, and you "lose the key scores". You don't find the
>>>_real_ place where you screwed up (a), instead the score seems to drop at
>>>(b) which is the wrong place.
>>>
>>>Since neither way finds the actual mistake, I don't like the back-to-front
>>>approach because if you do search front to back you will find the "mistake"
>>>at a different place, which is nothing more than confusing.
>>
>>Overall, the analysis is more accurate when it is performed from back to front.
>>Sure, there are still mistakes, but that's life.
>>
>>You used to recommend the back-to-front order yourself. In fact, I remember you
>>criticising Fritz once for doing it in the forward direction. :-)
>>
>>Dave
>
>
>I don't remember critizing Fritz, but whether I did or not, I decided to try
>it and didn't like it. I want the program to be as consistent as possible when
>it gives me analysis. If it can _prove_ that my move was worse then I want it
>to say so. But if it can't, I don't want it to say so just because it luckily
>kept something critical in the hash.
>
>IE if the _real_ problem is at move (A) and the program can't see it is in
>trouble until move (B), then Crafty will report the problem at B every time,
>until the hardware gets faster and then it will report it at B-1, etc...
>
>With the reverse annotation idea, it will report it at random locations between
>(b) and (a) which I simply didn't like. Change the hash size (bigger or
>smaller) and the point where it spots the problem changes.
Between b and a is better than b because you have a better upper bound for the
move that you made your mistake and if you go to do long analysis with Crafty to
discover the mistake you can at least avoid analyzing some moves after the
random location.
There are also cases when it can see the real problem by going backward.
Nothing is perfect but going backward is better than going forward.
Uri
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.