--- In [email protected], Trevor Halsall <[EMAIL PROTECTED]> 
wrote:
>
> 
> To get a result you can depend on, you generally need at least a 
1000 simulations.  This of course takes some time, so may not be the 
best use of limited time.
>  
> Basically, the more plies the better - but as has been pointed out, 
more plies means more simulations are needed, particularly the closer 
you are to the start of a game.
>  
Thanks to Kevin, Chris and Trevor for your replies.

I'm not so sure about the 1000 iterations minimum. From my previous 
experience of simming, you generally have a rough indication after 
about 500 iterations about which are the best two or three moves and 
there is often a clear, obvious best move at that point which doesn't 
need any further analysis. Yes, where there are several options, 1000 
or more can be helpful, I grant that.

I think I'm more clear on the number of plies needed after reading 
these responses, so thanks for the guidance. I'm half way through 
simming the WPC games and Quackle has so far informed me that I 
embarrassingly missed OrEGANO and WINGOVER as well as the infinitely 
trickier EXCORIATE and EPIFAUNAL.

Thanks again,

Phil R.

Reply via email to