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 Historical content remains viewable.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

bad pedagogy in backgammon.

1 view
Skip to first unread message

paulde...@att.net

unread,
Apr 8, 2008, 5:22:22 AM4/8/08
to
I've noticed an annoying habit among various analysts. Suppose there
are various candidate plays: A, B, C. The commentator tries to
explain why A is better than B or C. Various reasons are given and
one of the reasons might be that play A enables a take next turn
whereas after plays B or C, a pass will be necessary. However, this
mode of explanation is obviously a pure example of circular
reasoning. The reason that only play A allows a take is precisely
_because_ A is better than B or C. The take-next-turn comment
doesn't help at all in establishing which of the checker plays is
best.

Paul Epstein

gcmortal

unread,
Apr 8, 2008, 8:39:07 AM4/8/08
to

"Can I take-next-turn?" is a viable, practical approach.

If you can find the one single move that leads to a take, then that
is, by definition, the checker play with the highest equity. If
multiple moves are a take, at least you have pruned the tree. If every
variation is a pass, then perhaps then a try exists that may cause
your opponent to incorrectly hold the cube.

Peter Schneider

unread,
Apr 8, 2008, 9:12:44 AM4/8/08
to
Hi Paul,

wrote

Purely logically you are right. Seen more practiclly, the "take?" question
is a criteria, a rule of thumb for estimating the resulting equities which
is certainly a good start.

Imagine you have a chouette and people are not sure what to play. Somebody
brings up the "after A we'll have a take" argument. Everybody nods and
agrees -- fine, that's all what's needed. Only if somebody disagrees the
need for more explanations comes up. From a good player I would expect
that they can express some of their intuitive pattern recognition in
words, like "we won't leave so many blots" or "we anchor so a gammon is
less likely" or the like. I completely agree.

Peter aka the juggler


psim...@iwprint.com

unread,
Apr 15, 2008, 11:31:30 AM4/15/08
to
On Apr 8, 8:12 am, "Peter Schneider" <schneiderp_REMOVET...@gmx.net>
wrote:

At some scores, a given play may cause a double, and the same play may
not at another score. Play A could be right if the cube is turned,
but wrong because it is not. So the cube action may be the actual
"reason" why one play is better than another, and that is why that
logic makes sense.

Phil Simborg

Adam Tansley

unread,
Apr 27, 2008, 9:23:51 AM4/27/08
to

There is a similar rule of thumb - "Play A leads to a cube turn,
whereas play B doesn't; so I'll make play B".

In "Vision Laughs at Counting", Danny Kleinman shows a couple of
interesting exceptions to this rule:

GNU Backgammon Position ID: IgAAMAEAAAAAAA
Match ID : QYkQAAAAAAAA
+24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg (Cube: 2)
OOO | O O | | | 0 points
OOO | | | |
OOO | | | |
OO | | | |
OO | | | |
| |BAR| |v
XX | | | |
XX | | | |
XX | | | |
XXX | X | | | Rolled 14
XXX | X X | | | 0 points
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: tansley

The correct play - 4/off 2/1 - provokes a double. The vastly inferior
4/3 3/off staves off a double.

More complex is:

GNU Backgammon Position ID: LwAAAMAAAAAAAA
Match ID : QYkIAAAAAAAA
+24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg (Cube: 2)
OO | O O | | | 0 points
OO | O | | |
OO | O | | |
OO | O | | |
OO | | | |
| |BAR| |v
XX | | | |
XX | | | |
XXX | | | |
XXX | | | X | Rolled 12
XXX | | | X | 0 points
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: tansley

Once again the correct play - 9/8 9/7 - provokes a cube, and the
(slightly) inferior 9/6 staves off the cube.

tansley

bob

unread,
Apr 27, 2008, 11:02:27 AM4/27/08
to
GNU Backgammon Position ID: 294DAAJrsz0AAA
Match ID : cAkAAAAAAAAA
+24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
| O O O O O | | | 0 points
| O O O O O | | |
| O O | | |
| O O | | |
| | | |
| |BAR| |v (Cube: 1)
| | | |
| | | X |
| | | X |
| X X X X | | X X | On roll
| X X X X O X | | X X | 0 points
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: gnu


X to play 61 in a money game. What is the best move and what is the
cube action for O after X's two best plays?

Bob Koca

David C. Ullrich

unread,
Apr 28, 2008, 7:42:16 AM4/28/08
to

??? Oh, right - with the first play O has no reason not to
redouble since it's the last turn, while with the second
play O doesn't redouble because if he holds the cube
X can't reredouble if O misses.

Do you know any exceptions where the cube is in the middle?

>More complex is:
>
> GNU Backgammon Position ID: LwAAAMAAAAAAAA
> Match ID : QYkIAAAAAAAA
> +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg (Cube: 2)
> OO | O O | | | 0 points
> OO | O | | |
> OO | O | | |
> OO | O | | |
> OO | | | |
> | |BAR| |v
> XX | | | |
> XX | | | |
>XXX | | | |
>XXX | | | X | Rolled 12
>XXX | | | X | 0 points
> +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: tansley
>
>Once again the correct play - 9/8 9/7 - provokes a cube, and the
>(slightly) inferior 9/6 staves off the cube.
>
>tansley

David C. Ullrich

EthicalBackgammon

unread,
May 8, 2008, 2:29:40 PM5/8/08
to

I completely see your logic that this constitutes circular reasoning
and agree with you in the sense that is it synonymous to say that Play
B is worse than play A because if you make Play A it's no double, and
if you make Play B it is a double. The reason many people, including
myself, point out the doubling ramifications of a play is that this
helps you back in to the correct decision. When I am debating between
two plays, before I decide, I ask myself if one play is likely to get
me the cube and another play less likely, and I take that into account
in making the decision. Or I ask myself if I would take if I make one
play and drop if I make the other, and that also helps me decide
between the two plays. So after the fact, it is circular reasoning,
but in the decision process, looking at the cube possibilities is
quite valid.

0 new messages