RE: [Bug-gnubg] Beaver bug

[Albert Silver]

Did you look at the stats it is producing? I saw it alternate from
favorite to lost in each ply. At ply 0 it says it has 7.9% chances to
win but at 1-ply it says it has 67.8%. Now *that's* odd.

                                                Albert

> -----Original Message-----
> From: address@hidden [mailto:bug-gnubg-
> address@hidden On Behalf Of Christopher D. Yep
> Sent: Thursday, July 31, 2003 12:21 AM
> To: address@hidden
> Subject: [Bug-gnubg] Beaver bug
> 
> Sometimes gnubg doesn't report beavers correctly.  For example,
consider
> the following extreme position:
> 
>      GNU Backgammon  Position ID: Pp8PAAAAAPh8Pg
>                      Match ID   : cAkAAAAAAAAA
>      +24-23-22-21-20-19------18-17-16-15-14-13-+  O: gnubg
>      | X  O  X  O  X  O |   |                  |  0 points
>      | X  O  X  O  X  O |   |                  |
>      | X  O  X  O  X  O |   |                  |
>      | X  O  X  O  X  O |   |                  |
>      | X  O  X  O  X  O |   |                  |
>      |                  |BAR|                  |v (Cube: 1)
>      |                  |   |                  |
>      |                  |   |                  |
>      |                  |   |                  |
>      |                  |   |                  |  On roll
>      |                  |   |                  |  0 points
>      +-1--2--3--4--5--6-------7--8--9-10-11-12-+  X: Chris
> Pip counts: O 60, X 330
> 
> X is playing an extreme 15-checker backgame.  X owns the cube and is
on
> roll.
> 
> 0-ply, 1-ply, 2-ply, and 4-ply analysis (no reduced evaluations) all
state
> that X is a cubeless underdog.  Therefore, 0-ply, 1-ply, 2-ply, and
4-ply
> evaluations should definitely be no-redouble/beaver.  However gnubg
says
> it's no-redouble/take at 0-ply, 2-ply, and 4-ply.  At 1-ply it
correctly
> says that it's a no-redouble/beaver.
> 
> I'm using gnubg for Windows version 0.14-devel 1.1227 030727 (build
Jul 27
> 2003).
> 
> Here is gnubg's display for 0-ply through 4-ply (for both the hint and
> evaluations windows):
> 
> 0-ply:
> 
> Cube analysis
> 0-ply cubeless equity  -2.120
>      7.9%   0.0%   0.0% -  92.1%  87.0%  40.8%
> Cubeful equities:
> 1. No double            -2.093
> 2. Double, pass         +1.000  ( +3.093)
> 3. Double, take         -4.297  ( -2.204)
> Proper cube action: No redouble, take (41.6%)
> 
> 1-ply:
> 
> Cube analysis
> 1-ply cubeless equity  -0.012
>     67.8%   0.2%   0.0% -  32.2%  26.6%  10.4%
> Cubeful equities:
> 1. No double            +0.218
> 2. Double, pass         +1.000  ( +0.782)
> 3. Double, take         -0.243  ( -0.462)
> Proper cube action: No redouble, beaver (37.1%)
> 
> 2-ply:
> 
> Cube analysis
> 2-ply cubeless equity  -2.091
>      7.5%   0.0%   0.0% -  92.5%  86.5%  37.5%
> Cubeful equities:
> 1. No double            -2.066
> 2. Double, pass         +1.000  ( +3.066)
> 3. Double, take         -4.237  ( -2.172)
> Proper cube action: No redouble, take (41.5%)
> 
> 3-ply [note: the bug doesn't apply here; the interpretation of
> no-redouble/take is correct since .139 > 0]:
> 
> Cube analysis
> 3-ply cubeless equity  +0.159
>     73.7%   0.1%   0.0% -  26.3%  22.0%   9.6%
> Cubeful equities:
> 1. No double            +0.414
> 2. Double, pass         +1.000  ( +0.586)
> 3. Double, take         +0.139  ( -0.275)
> Proper cube action: No redouble, take (32.0%)
> 
> 4-ply:
> 
> Cube analysis
> 4-ply cubeless equity  -2.027
>      6.9%   0.0%   0.0% -  93.1%  85.1%  31.3%
> Cubeful equities:
> 1. No double            -2.004
> 2. Double, pass         +1.000  ( +3.004)
> 3. Double, take         -4.107  ( -2.104)
> Proper cube action: No redouble, take (41.2%)
> 
> Chris
> 
> 
> 
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/bug-gnubg