Re: [Bug-gnubg] I'm a bit confused

[Jim Segrave]

On Tue 29 Jul 2003 (16:19 -0300), Albert Silver wrote:
> Forgive me if this is a stupid question, but I'm having trouble reading
> the stats here. I rolled out the following position from a match:
> 
> GNU Backgammon  Position ID: 8ObEAShwNwJgSw
>                  Match ID   : cAkpA8AAYAAA
>  +13-14-15-16-17-18------19-20-21-22-23-24-+     O: Jon Royset
>  |       O     O    |   | O  O  X  X  X    |     12 points
>  |             O    |   | O  O  X  X       |     
>  |             O    |   |    O             |     
>  |                  |   |    O             |     
>  |                  |   |                  |    
> v|                  |BAR|                  |     25 point match (Cube: 1)
>  |                  |   |                  |    
>  |                  |   |                  |     
>  | O                |   | X  X             |     
>  | O              X |   | X  X             |     Rolled 22
>  | O     X        X | X | X  X        O  O |     12 points
>  +12-11-10--9--8--7-------6--5--4--3--2--1-+     X: Moshe Tissona
> 
> The rollout said:
> 
> 1. Rollout     bar/21 23/21 6/4      Eq.:  -0.176
>   43.5%  10.2%   0.5% -  56.5%  24.6%   1.3% CL  -0.287 CF  -0.176
> [  0.3%   0.2%   0.1% -   0.3%   0.3%   0.2% CL   0.009 CF   0.028]
> Full cubeful rollout with var.redn.
> 1296 games, Mersenne Twister dice gen. with seed 1175248512 and
> quasi-random dice
> Play: 0-ply cubeful [expert]
> Cube: 0-ply cubeful [expert]
> 2. Rollout     bar/21 23/21 5/3      Eq.:  -0.297 ( -0.121)
>   41.7%   9.7%   0.5% -  58.3%  24.8%   1.0% CL  -0.328 CF  -0.297
> [  0.3%   0.2%   0.1% -   0.3%   0.3%   0.1% CL   0.009 CF   0.024]
> Full cubeful rollout with var.redn.
> 1296 games, Mersenne Twister dice gen. with seed 1175248512 and
> quasi-random dice
> Play: 0-ply cubeful [expert]
> Cube: 0-ply cubeful [expert]
> 3. Rollout          bar/21 6/2*      Eq.:  -0.323 ( -0.147)
>   43.4%  11.3%   0.6% -  56.6%  23.2%   1.1% CL  -0.259 CF  -0.323
> [  0.3%   0.2%   0.1% -   0.3%   0.3%   0.1% CL   0.008 CF   0.021]
> Full cubeful rollout with var.redn.
> 1296 games, Mersenne Twister dice gen. with seed 1175248512 and
> quasi-random dice
> Play: 0-ply cubeful [expert]
> Cube: 0-ply cubeful [expert]
> 
> I'm trying to understand why move 3 is a -0.147 blunder when its wins
> are only 0.1% less, it scores more gammons, and loses less gammons and
> backgammons. 

All I can guess is that during the rollout, more of the losses and/or
gammon losses were with the cube turned.

-- 
Jim Segrave           address@hidden