There are count systems for non-contact positions like Thorp count and Keith count. But they are only for money games. Is there also a count system for matches that takes the the match score into account?

Here’s my approach:

1) Know your takepoint at the score
2) Calculate your money pip-count-of-last-take (PLT) using your preferred method (e.g., for Keith count, if opponent’s adjusted pip count is X, then PLT = X + floor(X/7) - 2)
3) The money PLT corresponds to about 22% game winning chances and each pip near PLT is worth about 2% game winning chances, so adjust accordingly. If you want a formula, if e = (your Keith-adjusted pip count) - (your money PLT), then your game-winning chances are GWC = 22 - 2e
4) Compare GWC to your takepoint at the score to determine whether you can take or not.

It looks complicated in words but it’s not too bad. At most scores you’re only going to adjust your take/pass decisions by a pip or two in either direction.

Thank you. When I read about those count systems there is quite a lot of discussion which of these is "optimal" or "outdated". Is there any consensus what the best count system is at the moment?

Of the strong players I’ve talked to, I would say most use Keith, some use ISIGHT, and a few savant types use the Kleinman formula or EPC (but I wouldn’t recommend bothering with these until you’ve mastered one of the simpler ones which give very good results).

Following situation:



White should redouble here. But both count systems I use say "no redouble". Is there a count system that would say redouble here?

White’s Keith adjusted count is 54, PLT is 54 + 7 - 2 = 59, Black’s Keith adjusted count is 56 which is within 3 of PLT so Keith count says borderline redouble for White. However, although the Keith count penalizes the gap on the 5 pt, it only penalizes it by 1 pip while it really should be penalized more because Black has so many checkers on the 6 pt (Black has many opportunities to roll 5s which are very wasteful). That pushes this from a borderline redouble to solid redouble IMO (haven’t looked at the evaluation — I would guess it’s 0.900 D/T)

[QUOTE=sdfsgf;58056166]White’s Keith adjusted count is 54, PLT is 54 + 7 - 2 = 59, Black’s Keith adjusted count is 56 which is within 3 of PLT so Keith count says borderline redouble for White. However, although the Keith count penalizes the gap on the 5 pt, it only penalizes it by 1 pip while it really should be penalized more because Black has so many checkers on the 6 pt (Black has many opportunities to roll 5s which are very wasteful)./QUOTE]

Keith Count says redouble within 3 of raw count, within 1 of PLT; hence, solid no redouble using the count alone.
Your observation about the open 5 point is VERY important here.... you waste a number anytime you roll a 5 before you have cleared the 6pt. HUGE! If you open the 4 pt by moving its' lone checker any lower point, obviously the situation becomes even worse as you now waste both 4s and 5s..