Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: PB-ON vs PB-OFF (final results)

Author: Jeremiah Penery

Date: 13:53:04 10/16/99

Go up one level in this thread


On October 16, 1999 at 11:55:16, Ratko V Tomic wrote:

>> Even after seeing your equations, I'm still not sure how you can
>> determine the pondering percentage by the winning percentage.
>
>If you know the rating difference you can compute total/effective thinking time
>difference or time ratio (using D=C*(R-1), using value C=100 rating points gain
>for doubling the thinking time). Once you have full thinking times ratio, and
>you know the time controls each _nominally_ got, you will notice that the
>effective thinking time ratio (as obtained from rating difference) is greater
>than the nominal time controls ratio, e.g. in the 1st match both had equal
>nominal times, but the time ratio from performance difference was 1.78, i.e. the
>pondering Rebel had more actual/effective time. Hence you can compute how much
>extra time the pondering version got to reach the times ratio expected from the
>performance ratings difference. That extra time can only come from the pondering
> time gain, and that's how you get G, the gain hit rate. (Obviously, in a fairly
>small sample as this you have a significant statistical error marging, so take
>the numbers 80% with a grain of salt.)
>
>
>>If pondering Rebel had win 100% of the games, would you say it pondered
>>with 100% accuracy?
>
>No. That could be only due to a statistical fluctuation in a 100 (or any finite
>number) game match. For example, if instead of different time controls you used
>depth D and D+1 versions, the D+1 will guess correctly the move of D version in
>every single move, 100% guess rate.


I don't think this is true.  I think the D+1 will ponder different moves more
than 20% of the time, or maybe more.  Even D vs. D would have about 20%
pondering difference.  This is because in D vs. D, say one searches to 8 ply,
then begins to ponder.  The next one searches to 8 ply, but it's already one ply
ahead of the other, so effectively it was like a 9-ply search from the first
one.  Since at any given ply depth, the probability of an engine changing its
best move is around 20% (see "DarkThought goes deep" and "Crafty goes Deep" from
JICCA.  Unfortunately, I don't have real references to these, but I believe
they're on DarkThought's webpage.), there will be a 20% pondering difference.
With D vs. D+1 this will be even more.


> Yet it will lose 20-30 percent of games.
>Guessing what the opponent will play doesn't mean it will find a correct answer
>to that move.

But obviously it does most of the time.

>>(And were you not the one saying that a shallower
>>search can produce ultimately better moves in
>>many cases? In these cases, a correct pondering would _worsen_ the result. :)
>>
>
>It's not something I made up.

Yes, I see that, but if this happened as often as you seem to say, the ~80%
pondering accuracy of Rebel vs. Rebel wouldn't have such a large affect on the
match outcome.



This page took 0.03 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.