Thorp Count questions
Ian Shaw
anyone could give me some pointers.
1) If the roller's count comes to 30 exactly, should you add 10% to make it
33?
2) When adding 10%, should you round normally, always round up or always
round down?
a) 71, is the final count 78 or 79?
b) 75, is the final count 82 or 83?
c) 77, is the final count 84 or 85?
3) If the non-roller is exactly 2 behind, is it a take or a pass? I've seen
both printed.
4) I've read that a Thorp count is better in short races, but that the
racing lead in % is better for long races. Does anyone know roughly where
the breakpoint comes?
--
Regards
Ian Shaw (ian on FIBS)
Chuck Bower
Ian Shaw <ian....@riverauto.co.uk> wrote:
>I've got a few questions regarding the Thorp Count. I'd be grateful if
>anyone could give me some pointers.
>
>1) If the roller's count comes to 30 exactly, should you add 10% to make it
>33?
How about adding 1.5. This 'discontinuity' is probably the Thorp Count's
weakest feature. Thorp is a mathematics professor and recognized that
simplicity should count for something, otherwise he would have made the
adjustment more complicated.
>2) When adding 10%, should you round normally, always round up or always
>round down?
> a) 71, is the final count 78 or 79?
> b) 75, is the final count 82 or 83?
> c) 77, is the final count 84 or 85?
If you don't round at all, you will often avoid #3 below. That is
my suggestion.
>3) If the non-roller is exactly 2 behind, is it a take or a pass? I've seen
>both printed.
I'm not sure what Thorp's original (and apparently difficult to find)
work said. I use 'take', but it probably doesn't really matter.
The Thorp Count is a guide but it is far from 100% accurate. It's better
than guessing and for bearoff's it is usually better than the popular
percentage methods.
>4) I've read that a Thorp count is better in short races, but that the
>racing lead in % is better for long races. Does anyone know roughly where
>the breakpoint comes?
Maybe you read something I wrote and thus this may not be an independent
answer. Personally I think Thorp loses much of its reliability when there
are checkers outside the home board. In addition, when there are only a
few checkers left (~5) for each side, it also suffers bigger uncertainties
(in general, IMO).
Thorp's method was made popular by Bill Robertie in his two volume
ADVANCED BACKGAMMON. He also covers other methods there which compliment
Thorp. I recommend careful (re)reading of those books to help solve some
of these dilemmas.
Chuck
bo...@bigbang.astro.indiana.edu
c_ray on FIBS