Re: [Bug-gnubg] GNU backgame eval + bug

[Jim Segrave]

On Sun 06 Oct 2002 (14:49 -0300), Albert Silver wrote:
> >From a post in Gammonoline with Position ID: 2LYBCDPOewMAAA
> ---------------------------------------------------------
> On SW 4's best setting, 3ply/precise. 360 rollouts give a cubeless
> equity of .565 +/- .023. (I'll report back when 1296 are finished in a
> few days as time permits) 
> This position is a double and a clear take. The market will be clearly
> lost if the 6 point is cleared. 
> Anyhow..lets look at some bot cubless evaluations on their highest
> settings (except GNU where I will look at different plies) 
> SW 4 .549 WOW..close to the result like it is supposed to be. 
> SW 3 .773 WOW..it incorrectly thinks this is a BIG pass. 
> JF .760 and also passing incorrectly and underrating this fine backgame.
> 
> GNU is acting funny as there's a bug in the cubeless equity
> presentation(it doesn't add in G and BGs) so I'll just calculate it. 0
> ply .573 , 1 ply .819 . 2 ply .636, 3 ply .825 , 4 ply .607. 
> Apparently GNU 4 ply is seeing enough hitting variations but why is 0
> ply right on and 1 and 3 ply way off ? Anyhow, from playing around with
> GNU I will state that I think it could use some more training in
> backgame aspects. GNU certainly has the potential to become the world's
> premier BG bot. 
> Note that SW 4 1ply thinks this is .504 cubeless which is overvalueing
> the backgame which I am seeing, at times, IMHO. 
> ..neilkaz.. 

And the folloup from Kit:

  Yes, I have seen this oscillation effect often, and it can be
  irritating. It can also happen in a truncated rollout. The result
  after 7 rolls can vary quite a bit from what it would be after 8 (or
  6) rolls because of the oscillation effect -- keep in mind that we
  are talking about an evalution.

  The solution for the rollouts: When doing a truncated rollout (say 7
  rolls deep), take the AVERAGE of the equity estimate after 6 rolls
  and 7 rolls. This would require very little reprogramming, and would
  eliminate the oscillation effect and give us more accurate
  results. GNU people -- are you listening?

  Kit 

Anyone have any opinions on whether this is a valid approach or just a
bandage to cover up this sort of inconsistancy? It certainly would
avoid some confusion for users, on the other hand, the very fact that
the equity is oscillating is a sign that either gnubg is weak here or
the position is one with a very unstable equity.

Should we assume it's an converging series which oscillates around the
terminal value and try to extrapolate from all the previous plies how
the convergence is actually working? (I'm not volunteering).


-- 
Jim Segrave           address@hidden