RE: [Bug-gnubg] RE:Why is odd ply equity always lower?

[Ian Shaw]

> -----Original Message-----
> From: address@hidden
> [mailto:address@hidden Behalf Of Joern
> Thyssen
> Sent: 10 June 2003 13:32
> To: GNU Backgammon Bugs
> Subject: Re: [Bug-gnubg] RE:Why is odd ply equity always lower?
> 
> 
> On Tue, Jun 10, 2003 at 01:11:52PM +0100, Ian Shaw wrote
> > 
> > Since a 1-ply eval is just 21 0-ply evals, it appears that GnuBg is
> > consistently overvaluing the equity for the side on roll.
> 
> Not necessarily. gnubg may just be overestimating one side. 
> For example,
> it may overestimate the side playing against a holding game. The 1-ply
> evaluation is averages of equities for the player having the
> high-anchor, so it's not symmetric unless it's a mutual holding game.
> 
> For any of your positions, try with the opponent on roll, and 
> calculate
> the 0-ply, 1-ply, 2-ply, etc. equities. If they exhibit the 
> same pattern
> then you may have a point.
> 
> Jørn
> 
I've just tried 7 assorted but common positions (money games), changing only 
who is on roll,  and haven't found any evidence to support the theory.

5 point holding game    bLuBAQO6eQ4GAA  1 & 3 ply consistently think roller's 
position is better than 0 & two ply do.

55 blitz after 54               4PPgQUAzTvABMA  1 & 3 ply consistently think 
roller's position is better than 0 & two ply do.

100 - 96 pip race               2N1tAADY9m4AAA  No pattern.

90 - 132 race           YNvdAQC4uzsAAA  No pattern.

Full prime vs 2-point   2N0NABhsdzcAAA  No pattern.

4 point holding game    Nnc3AADY3Q0ABg  Racer's position improves until 3-ply.

Mutual Holding game     sN2GAQOY3YYBBg  3-ply most pessimistic, 4-ply most 
optimistic

--Ian