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 Historical content remains viewable.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

opening rolls: split vs. slot according to computers

2 views
Skip to first unread message

MeMyselfAndI

unread,
Apr 1, 2001, 7:04:00 AM4/1/01
to
Hi all,

Jellyfish and Snowie prefer to split in the opening when they have to
play a 1 in the opening (4;1 5;1 2;1) . We know that bg programs
still have problems to handle backgames well. On the other side,
backgames very often occur if you slot the 5 point in the opening.
Is it possible that computers prefer to split because of their bad
backgame abilities. Will computers in some decades switch to a
slotting style?

Kind regards

slotalot

Daniel Murphy

unread,
Apr 1, 2001, 6:24:59 AM4/1/01
to
On Sun, 01 Apr 2001 11:04:00 GMT, inter...@yahoo.de (MeMyselfAndI)
wrote:

It's possible that the computers' overvaluation of racing equity and
undervaluation of backgame or many-men-back positions influence their
opening play preferences.

But I wouldn't characterize JF and Snowie as "bad" backgame players
for either side. Perhaps a larger problem, and one that seems to
impact JF more than Snowie, is that their equity estimations of
backgame-type positions may be way off, with detrimental affect to
cube decisions.

Although the bots lack world-class technique in rolling primes home,
catching and trapping a second checker, and recirculating checkers,
all in all they play normal-type backgames well.

Whether or not the bots are wrong, or by much, on opening 21, 41, and
51, current expert opinion seems to agree with the bots about 41 and
51, and disagree about 21.

From what I've observed, the (not unanimous) consensus on these plays
(at most match scores) is :

21: 13/11 6/5

51: 24/23 13/8

41: 24/23 13/9.

World class player Neil Kazaross commented on the 41 slot in a 1994
interview: "The most commonly misplayed position? In my opinion, it
would be an opening 4-1. Anybody who slots makes a mistake, probably
by a percent or two. There's more types of positions than that that
are misplayed, but I'm not talking about types of positions. This one
is common."

Complete interview at

ftp://ftp.cybercom.net/pub/users/damish/backgammon/kazaross.txt


Daniel Murphy
rac...@best.com, Raccoon on FIBS, GamesGrid

13th Nordic Open, 12-16 April 2001
50,000 DKK added http://www.nordicopen.dk

Vi ses, og tak for alle fiskene!

0 new messages