Re: [Bug-gnubg] Bug#659494: Resigns more points than necessary when fewer points would still result in opponent win

[Joseph Heled]

if the bot does not care about points lost, it should resign, no?

-Joseph

On 15 February 2012 11:23, Michael Petch <address@hidden> wrote:
> As an addendum. Bots often make what seem like unnatural plays compared
> with that of a human. If every decision has the same equity the bot will
> use the one that happens to be at the top of it's decision list (And to
> a human can seem random and odd). If we start going down this slippery
> slope then the user will then ask us why the bot might do this (look at
> what move the bot plays in this case. It doesn't save the gammon):
>
>    GNU Backgammon  Position ID: AwAA2Lo7EAAAAA
>                    Match ID   : 8AlnAaAAAAAA
>    +24-23-22-21-20-19------18-17-16-15-14-13-+  O: mpetch
> OOO | O                |   |                  |  10 points
> OOO | O                |   |                  |
> OOO |                  |   |                  |
>  OO |                  |   |                  |
>  OO |                  |   |                  |
>    |                  |BAR|                  |v 11 point match (Cube: 1)
>    |                  |   |                  |
>    |                  |   |                  |
>    |          X  X  X |   |                  |
>    | X  X     X  X  X |   |                  |  Rolled 61
>    | X  X  X  X  X  X |   |                X |  0 points
>    +-1--2--3--4--5--6-------7--8--9-10-11-12-+  X: gnubg
>
>
>    1. Cubeful 2-ply    12/6 4/3                     MWC:   0.000%
>       0.0000 0.0000 0.0000 - 1.0000 1.0000 0.0000
>        2-ply cubeful prune [world class]
>    2. Cubeful 2-ply    12/6 3/2                     MWC:   0.000% (
> 0.000%)
>       0.0000 0.0000 0.0000 - 1.0000 1.0000 0.0000
>        2-ply cubeful prune [world class]
>    3. Cubeful 2-ply    12/6 2/1                     MWC:   0.000% (
> 0.000%)
>       0.0000 0.0000 0.0000 - 1.0000 1.0000 0.0000
>        2-ply cubeful prune [world class]
>    4. Cubeful 2-ply    12/6 1/off                   MWC:   0.000% (
> 0.000%)
>       0.0000 0.0000 0.0000 - 1.0000 0.0000 0.0000
>        2-ply cubeful prune [world class]
>    5. Cubeful 2-ply    12/5                         MWC:   0.000% (
> 0.000%)
>       0.0000 0.0000 0.0000 - 1.0000 1.0000 0.0000
>        2-ply cubeful prune [world class]
>    6. Cubeful 2-ply    12/6 5/4                     MWC:   0.000% (
> 0.000%)
>       0.0000 0.0000 0.0000 - 1.0000 1.0000 0.0000
>        2-ply cubeful prune [world class]
>
>
> ----
>
> To a human, it may seem natural to save gammon - to the bot it knew it
> was playing a lost cause so it decided not to even save itself from a
> gammon. The bot knows it has no way to win, and that all moves have 0
> equity. In this case the first move listed gets used which is 12/6 4/3
> which gets the bot gammoned. The human win 12-0. But  a human might say
> "hey wait a minute I should have only won by 11-0, what is the bot doing
> it is unnatural not to save a gammon - a human likely wouldn't do that).
>
> Same principle applies here. Bot doesn't care how much the human wins
> by. A win is a win no matter the number of points you go over by. Should
> there be an option that the bot must play to save gammons and
> backgammons in lost causes (absolutely 0% chance to win)? I'd say heck
> no, again pointless as the end result of a match loss remains unchanged.
>
> --
> Michael Petch
> CApp::Sysware Consulting Ltd.
> OpenPGP FingerPrint=D81C 6A0D 987E 7DA5 3219 6715 466A 2ACE 5CAE 3304
>
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/bug-gnubg