Re: [Bug-gnubg] EPC (Trice) and other counts

[Philippe Michel]

On Sat, 13 Nov 2010, Adi Kadmon wrote:

> Hello all
>
> I'd like very much to have the following information:
>
> (1) What is GNUbg's formula for calculating "EPC"/"Trice count" of bearoff
> positions?
>
> (2) Is that formula, or any approximation thereof, practically usable by a
> human player during a real game?

There is no formula. It either comes from a bearoff database (calculated 
recursively) or from a rollout.

What a player can do is use gnubg to setup positions and get a general 
idea of how much wastage there is in common formations (a closed board 
with spares, a "racing" board with a more or less triangular checker 
formation, "usual" formations with fewer checkers, etc...)

There is an approximate formula for short bearoffs described here,
page 40 : 
http://www.dbgf.dk/gammonblad/Gammon_117_Online.pdf
(it's in danish, basic formula is easy to understand nevertheless, but 
adjustments less so).

> (3) Is the EPC/Trice count, if calculated precisely, actually more "correct"
> than, for instance, Ward count or Keith count, for cube decisions in bearoff
> positions?

Generally speaking, Ward and Keith work well for longer races and 
relatively smooth bearoffs (they are supposed to account for some wastage, 
but not for totally ragged positions). Trice is most useful for positions 
with many checkers stacked on the low points and shorter bearoffs.