Please send the gcg.  I would guess it is the same as the issue I
previously commented on - partial racks don't evaluate correctly.  I
can determine if that is the issue from the gcg fairly readily.

Matt

--- In [email protected], "zev_kesselman" <ke...@...> wrote:
>
> John, 
> 
> I think Chris has the right idea.  I have a similar issue, and I have 
> a .gcg for it.
> 
> With one tile in the bag, Expert Player lists 15 possible moves, the 
> first of which has a 50% chance of winning - depending on which of 
> the 8 tiles is in the bag (3 of them are E's, simplifying the 
> scenario).  Play number 5 has 0% chance.  Both evaluations seem 
> correct.
> 
> But when I do a sim of #1 against #5, #5 is reported as 100% win.  
> I've looked at the XML file, Quackle's moves don't add up to a win.  
> So Chris' concluding sentence seems appropriate.
> 
> Zev
> 
> 
> --- In [email protected], "John O'Laughlin" <olaughlin@> 
> wrote:
> >
> > Chris,
> > 
> > Do you have a gcg for this game?
> > 
> > John
> > 
> > On Tue, Jan 6, 2009 at 7:16 PM, tendenc <tendenc@> wrote:
> > > Quackle AAGINR 451 pts
> > > Me NITROSO 368 pts
> > >
> > > New Player 1: Turn 12
> > > A B C D E F G H I J K L M N O
> > > ------------------------------
> > > 1|B E E C h N U T P =|
> > > 2| - " R E " U - |
> > > 3| - G O X O N - |
> > > 4|' - I D U T '|
> > > 5| N O D R |
> > > 6| " I F " I " |
> > > 7| Q A T ' ' P E ' |
> > > 8|= f I B R I L S L ' =|
> > > 9| O ' A ' Y E ' |
> > > 10| V T H E M E D " |
> > > 11| E E - D |
> > > 12|Z A G - S ' - W '|
> > > 13|I L ' ' O Y H A J|
> > > 14|T O U C A N W A V E R |
> > > 15|S M ' A L K I E ' F E|
> > > ------------------------------
> > >
> > > The winning play is NITROSO at O5 for 74 points. This wins the 
> game.
> > >
> > > When I simulate the generated choices, NITROSO at M5 (62 points)
> > > gives a 100% win, when in actuality loses the game immediately.
> > >
> > > I don't pretend to understand all the intricacies of simulation 
> but
> > > it would seem that a math function to calculate the score after an
> > > outplay to validate the win% would seem appropriate.
> > >
> > > Cheers!
> > >
> > > Chris ten Den
> > >
> > >
> >
>


Reply via email to