|
SEM wrote:
> > Elsewhere their advice is wrong:
> > "In a long-running game, double if you lead by over 10% of your
> > pipcount. If this lead is over 15%, your opponent should concede."
>
> What is the correct version of this advice?
Roughly, 8% lead to double from centre, 9% to redouble, at most 12% behind
to take. (Incidentally, I'm always amazed that there is a working formula
which is a constant percentage of the pipcount). Short races and positions
with odd bearoff distributions may need modification, which can be done by
the Thorpe count and other specialised counts.
Before anyone asks, the Thorpe count is pipcount, plus 2 for each chequer
still on the board, minus 1 for each inner board point with any chequers
on, plus 1 for each chequer on the ace point. If the leader's count is over
30, he must then add 10%. The leader should double if his count is at most
2 greater than the trailer's, redouble if only 1 greater, and the trailer
should drop if the leader is 2 ahead. Got that?
At different match scores, you need to know your percentage winning chance.
Does anyone know a SIMPLE formula for that? (That doesn't rely on me having
to multiply by numbers like 73 over the board!)
--
Stephen R. E. Turner
Stochastic Networks Group, Statistical Laboratory, University of
Cambridge e-mail: sret1@cam.ac.uk WWW:
http://www.statslab.cam.ac.uk/~sret1/home.html "You may notice that your
Customer Reference Number has changed" British Gas
|
![[GammOnLine forum]](pics/GammOnLine.gif)
![[Long message]](pics/Long.gif)
Long message
![[Recommended reading]](pics/Star2.gif)
Recommended reading
![[Recent addition]](pics/New.gif)
Recent addition