Against a back man

[Jean-Pierre]

Hpw to find the best play in this position

X to play 51 Money session.

+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O O O | | X X |
| O O O O O | | |
| O O O O | | | S
| | | | n
| | | | o
| |BAR| | w
| | | | i
| | | | e
| X | | |
| X X X X X | | |
| X X O X X X | | X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+

Pipcount X: 87 O: 64

CubeValue: 4, O owns Cube

[Nige M]

10/4
"Jean-Pierre" <jean-pier...@wanadoo.fr> wrote in message
news:44d66aab$0$876$ba4a...@news.orange.fr...

[Grunty]

I'd play 10/4, bringing to safety one blot, while retaining as much
outfield coverage as possible. X leaves 8 shots.
(11/5 is pointless, since 10-point gives less coverage and offers one
more shot than 11-point.)

16/10 is safest, but resigns outfield coverage at the same shot cost.
Its virtue is leaving less blots around, but the gammon threat is
significantly diminished by O's open 6-point. Should O had retained a
full board, this play would be correct.

[Lolek]

I would rather play 16/15 10/5 to get one more direct shot number from
the 5 point, and better coverage of the outfield. :-)

[David C. Ullrich]

On 6 Aug 2006 23:10:53 -0700, "Lolek" <pdol...@poczta.fm> wrote:

>I would rather play 16/15 10/5 to get one more direct shot number from
>the 5 point, and better coverage of the outfield. :-)

Why do you say that gives better outfield coverage than 10/4?

************************

David C. Ullrich

[mont...@lycos.com]

With 11-6, 13-12 you avoid a couple of killer doubles that might cost
you a gammon, and you also get some pick and pass numbers if he can't
try to make a run for it on his roll.

[Derek Ray]

paulde...@att.net wrote:

> Jean-Pierre wrote:
>> Hpw to find the best play in this position
>>
>> X to play 51 Money session.
>>
>> +24-23-22-21-20-19-------18-17-16-15-14-13-+
>> | O O O O O | | X X |
>> | O O O O O | | |
>> | O O O O | | | S
>> | | | | n
>> | | | | o
>> | |BAR| | w
>> | | | | i
>> | | | | e
>> | X | | |
>> | X X X X X | | |
>> | X X O X X X | | X X |
>> +-1--2--3--4--5--6--------7--8--9-10-11-12-+
>>

>> CubeValue: 4, O owns Cube
>

> I would hit O's blot.

X to play 5-1, not 6-1.

--
Derek

insert clever quotation here

[Grunty]

Derek Ray wrote:

> paulde...@att.net wrote:
> > Jean-Pierre wrote:
> >> Hpw to find the best play in this position
> >

> > I would hit O's blot.
>
> X to play 5-1, not 6-1.

OR... he meant playing some 10/5 or 11/6 AND hitting O's blot: 4/3* !

[Derek Ray]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

In which case, I'll buy as much of O's action as he's selling.

- --
Derek

insert clever quotation here
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFE133otQZlu3o7QpERAmA3AJ4/5yXhutR+SUIzpVzx8d2eAp3m2gCeMV1g
08EBXyrlk+5O6eR9ygv+GNY=
=R0+G
-----END PGP SIGNATURE-----

[Philippe Michel]

As X, you want to block O's back man or hit it when it escapes. So you
try for the best combination of outfield control and blocking points.

Points in indirect range don't work well (here making the 10pt would do
nothing and the 11pt would only block 44). Those in direct range do.

In the position above, it looks like 10/4 or 10/5 16/15 must be best. It
seems clear that the 5 has to be played from the 10pt. The blots on the 11
and 13pts shouldn't move (they're already a little too advanced for the
sake of outfield control). I have no idea whether the last one should be
on the 15 or 16pt.

If X had to play 41 on the other hand, it would be 13/9 10/9, blocking 61,
62, 63 and 66 at the cost of a moderate loss in outfield control.

[Derek Ray]

Philippe Michel wrote:
>> X to play 51 Money session.
>>
>>
>> +24-23-22-21-20-19-------18-17-16-15-14-13-+
>> | O O O O O | | X X |
>> | O O O O O | | |
>> | O O O O | | | S
>> | | | | n
>> | | | | o
>> | |BAR| | w
>> | | | | i
>> | | | | e
>> | X | | |
>> | X X X X X | | |
>> | X X O X X X | | X X |
>> +-1--2--3--4--5--6--------7--8--9-10-11-12-+
>>

>> CubeValue: 4, O owns Cube

(snip)

>
> If X had to play 41 on the other hand, it would be 13/9 10/9, blocking 61,
> 62, 63 and 66 at the cost of a moderate loss in outfield control.

(UNRELATED TO ORIGINAL ISSUE REMARK)

Actually, it's a near-zero loss in outfield control. The rolls that
really chafe your hide after 13/9 10/9 are 5-6 and 4-6 -- and those
rolls skipped past (or hit!) your checker on 13 if you'd left him there.

[mont...@lycos.com]

GNU 2 ply likes 10-4 (.032 better than 16-15, 10-5). A quick look
(after GNU analysis) suggests that leaving the fewest shots is best.
For example, after my play, 11-6, 13-12, getting hit with 5,5 means
about 50 less gammons than if you get hit with a 5,4 after the
supposedly best play. The two "nasty" doubles (4,4 and 5,5) after my
guess, are actually not nearly as bad as many would think. A good
position to keep as a reference.

[mont...@lycos.com]

GNU Backgammon Position ID: mZuDATA2d4LBAA
Match ID : cIkSAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: opponent
| X O X | | O O O O | 0 points
| X O X | | O O O |
| O | | O |
| | | |
| | | |
v| |BAR| | (Cube: 1)
| | | |
| | | |
| | | X X |
| O | | X X X X O | Rolled 54
| O X | | X X X X O | 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: me

Here's another interesting one: Setting the position upm as a money
game, GNU 2 ply has 18-13, 18-14 as .101 better than 13-8, 13-9 (my
play) but this was DMP, and my play wins 13 more games.

[mont...@lycos.com]

Correction: It's the other way around, that is, my play was better by
.101 at money, but at DMP, wins 13 fewer "plain" games.