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" <olaugh...@...> 
wrote:
>
> Chris,
> 
> Do you have a gcg for this game?
> 
> John
> 
> On Tue, Jan 6, 2009 at 7:16 PM, tendenc <tend...@...> 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