Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.

Dismiss

2 views

Skip to first unread message

Sep 9, 1991, 5:39:28 PM9/9/91

to

I saw a book by Phillip Martyn on a trip to England in 1978. I

didn't buy the book (to my later regret) but recall it mentioned a

technical paper by Edward (Beat the Dealer) Thorpe on running games.

Does anybody have a copy of this? I would also be interested in

mathematical analyses by other people.

Sep 10, 1991, 8:39:58 PM9/10/91

to

Edward Thorpe worked out a formula for calculating race doubles

(an elaborated pip-count) which is claimed to be 99% accurate. There

are certain positions (described in Bill Robertie's "Advanced Backgammon")

where it falls down a little. On the whole, it is extremely accurate

for medium and long races, and has the advantage of being easily (:-)

calculated over the table.

This is the formula:

. Compute the leader's doubling number as follows:

1) The leader's pip count,

2) plus two for each of the leader's remaining checkers,

3) plus one for each checker on the ace point,

4) minus one for each point covered in the home board,

5) plus 10% of the total so far, if that total is greater

than 30.

. Call the final result L, the leader's adjusted pip count.

. Compute the trailer's doubling number by following steps

1-4, but omitting step 5. Call this result T, the trailer's

adjusted pip count.

. If T > L-2, the leader should double.

. If T > L-1, the leader should redouble.

. If T > L+2, the trailer should pass.

Looks harder than it is. Anyone serious should be able to do a

normal pip count fairly quickly. Adding/subtracting the other bits

just takes a bit of practise.

If you want something a little easier, try this:

. Compute both player's pip count. Find the difference and

compare that difference to the leader's count.

. With a difference of 8% or more, the leader has an initial double.

. With a difference of 9% or more, the leader has a redouble.

. With a difference of 12% or less, the trailer has a take.

Note that these formulae apply to money-game races, and do not take

any account of match-play scores. Use at your own risk!

Cheers,

Simon si...@Aus.Sun.COM

COM

\

0 new messages

Search

Clear search

Close search

Google apps

Main menu