Whats the best strategy for this position?
MJR
Hi all,
I am hoping to generate some comments on what X's best strategy is in
this position:
Money game, cube in center:
-------------------------------
| O O O O O O | | - - - - - X |
| O O O O O O | | X |
| O O | | X |
| | | X |
| | | X |
| | | |
| | | |
| X X X X | | X |
| - O X X X X | | X - - - - - |
-------------------------------
How many blots should X be willing to leave in the outfield?
For example:
Is it OK to move 2 off midpoint on a 2-1 or 4-3, or is this too
dangerous.
Does X ever want to slot the 8 point, or hit loose?
How to play 3-2 or 4-1.
And how about rolls like 6-2 and 6-3..is it better to bring in a builder
for the 2 point, or better to leave a blot in the outfield?
Kit Woolsey
MJR (hac...@ptd.net) wrote:
: Hi all,
: I am hoping to generate some comments on what X's best strategy is in
: this position:
: Money game, cube in center:
: -------------------------------
: | O O O O O O | | - - - - - X |
: | O O O O O O | | X |
: | O O | | X |
: | | | X |
: | | | X |
: | | | |
: | | | |
: | X X X X | | X |
: | - O X X X X | | X - - - - - |
: -------------------------------
Plenty of interesting questions here, and I'm sure there would be some
disagreement even among experts.
: How many blots should X be willing to leave in the outfield?
: For example:
: Is it OK to move 2 off midpoint on a 2-1 or 4-3, or is this too
: dangerous.
The key is where those outfield men are moving. The idea is not to try
to build the eight point by getting a bunch of builders for it -- that
takes too much luck and runs obvious risks. However, putting blots in
the outfield for the purpose of hitting O if O escapes partially is
important. Thus, the farther back the blots are, the more valuable they
are. Also, X wants to bring in builders to make the two point or attack
O there. Thus with 4-3 I would play 13/6. However with 2-1 I would risk
13/11, 13/12.
: Does X ever want to slot the 8 point, or hit loose?
: How to play 3-2 or 4-1.
X never wants to slot the eight point. The one thing he can't risk is to
have O escape with a hit. While X has to take some positive action to
drive O's checker back since O is ahead in the race, slotting the eight
point is not the right idea. If X doesn't gut lucky and roll a joker to
build the eight point, he will plan to hit loose on the two point.
However, he cannot afford to break his five-prime to do this -- that
makes it too easy for O to escape if he hits back. So with 3-2 or 4-1 X
has no choice but to bring builders down.
: And how about rolls like 6-2 and 6-3..is it better to bring in a builder
: for the 2 point, or better to leave a blot in the outfield?
Here it matters where the outfield blot is going. As we have seen, a
blot on the nine point has virtually no containment value, while a blot
on the 12 point can be very valuable for containment. Thus with 6-4 and
6-3 I would move one checker. With 6-2 and 6-1 I'm not sure -- the
outfield checker is valuable, but so is the builder for the attack. With
6-1 I think I would play 13/12, 13/7 -- with 6/2 I think it is a close call.
Kit
Ron Karr
Kit Woolsey wrote:
>
> MJR (hac...@ptd.net) wrote:
>
> : Money game, cube in center:
>
> : -------------------------------
> : | O O O O O O | | - - - - - X |
> : | O O O O O O | | X |
> : | O O | | X |
> : | | | X |
> : | | | X |
> : | | | |
> : | | | |
> : | X X X X | | X |
> : | - O X X X X | | X - - - - - |
> : -------------------------------
>
> : Does X ever want to slot the 8 point, or hit loose?
> : How to play 3-2 or 4-1.
>
> X never wants to slot the eight point. The one thing he can't risk is to
> have O escape with a hit. While X has to take some positive action to
> drive O's checker back since O is ahead in the race, slotting the eight
> point is not the right idea. If X doesn't gut lucky and roll a joker to
> build the eight point, he will plan to hit loose on the two point.
> However, he cannot afford to break his five-prime to do this -- that
> makes it too easy for O to escape if he hits back. So with 3-2 or 4-1 X
> has no choice but to bring builders down.
>
Kit is correct that slotting the 8 point with a 3-2 or 4-1 is normally
wrong. For example, Jellyfish shows the equity from 13/8 as -.119 while
the equity from 13/9 13/12 is +.072.
However, I think slotting the 8 point is in fact correct under the
conditions stated (money game, cube in the middle). The reasons:
If X slots and O hits with a 6, O has now lost his market by a ton and
has significant gammon chances, so in a match, he would play on for
the gammon. However, in a money game (as this problem was stated),
the Jacoby rule is in effect. So after hitting O must settle for a
single
point by doubling. Therefore there's no "extra" cost to X to getting
hit.
And O is not strong enough to double BEFORE hitting in order to
activate gammons.
If X alternatively brings 2 builders down, O's sixes are pretty good
anyway. 4 numbers hit a blot and 1 number (66) runs to safety, thereby
winning with the cube. The other 6 sixes leave X roughly a double shot
to hit back. Let's say that of those, X wins 4 games and O wins 2. So
X
only wins an extra 4 games or about 11% by making the safer play.
What's the gain from slotting? When O doesn't hit, X now has an
extremely efficient cube turn, since he has many numbers that either
make the 8 point or hit on the 2. Jellyfish shows X's cubeless equity
as
around .52. With cube leverage, this should be a take for O, but it's
close to a pass. In other words, X can convert around a 75% chance of
winning to almost 100% by doubling.
Suppose X doesn't slot and O doesn't roll a 6. Now X is a favorite, but
not nearly enough to claim. He will still have to make the 8 point, or
hit on the 2 point and not get hit back, in order to win. Meanwhile O
continues to have escaping chances. Jellyfish says X's equity after
13/9
13/12 is .365.
So is the gain from being able to almost claim with the cube after a
slot
worth the extra 11% chance of being hit immediately? I think so. But
only because gammons won't count against X. If the Jacoby rule were
not in effect, or the cube had already been turned, the slotting play
would be wrong.
The concept here: it's worth taking a risk that increases the value of
the cube to you. If the risk succeeds here, X gets tremendous value
from doubling; if the risk fails, O gets no value from the cube.
Ron
Kit Woolsey
Excellent analysis, Ron. I was only looking at the plays from the point
of view of the best play, without taking cube considerations into
account. While there would be some number crunching to be done to get a
completely accurate result (keep in mind that if X brings two builders
down and O doesn't roll a 6 it may well be correct for X to double
anyway, even though now O has a trivial take), I would predict that your
conclusion will turn out to be correct.
Kit
Hugh B. McNeil
================
=oooooo[] x
=oooooo[] x
= oo[] x
= [] x
= [] x
= []
= xxxx[]x
= oxxxx[]x
==================
cube in the centre...
Kit suggests that x never wants to slot the 8 point... I suspect
that with o up ~ 37 pips that a slot by x, followed by a double
by x after a miss would be the best strategy...it might even be
better to slot the 8 with a 6-5 rather than hitting...
I'd be willing to prop this one for an hour or two...
is JF out there?
lee
With cube on bar (money game) I think I would slot the eight point with
a 5-any, and force O to roll a six or sweat out a double.
If O now doubles (in my regular chouette, at least) he would risk having
to face an Ealing of beavers.
In fact the more I look at this position the more a slot seems to be the
right move.
In article <572cf1$g...@freenet-news.carleton.ca>, "Hugh B. McNeil"
<ch...@FreeNet.Carleton.CA> writes
--
----------------------------------------------
l...@infoplus.demon.co.uk
Web site: http://www.ibmpcug.co.uk/~oak/info/
Email for links to/from this site
----------------------------------------------
Julian
In article <32920D...@best.com>, Ron Karr <ka...@best.com> writes
> However, in a money game (as this problem was stated),
>the Jacoby rule is in effect.
I wouldn't have assumed this - and I'm sure the majority wouldn't
have. Is Jacoby really such standard practice for money games that
it goes without saying?
---------------------------------------------------------------------------
Julian Hayward 'Booles' on FIBS jul...@ratbag.demon.co.uk
+44-1344-640656 http://www.ratbag.demon.co.uk/
---------------------------------------------------------------------------
"Any man can be 62, but it takes a bus to be 62A"
- Spike Milligan
---------------------------------------------------------------------------
Turnpike evaluation. For information, see http://www.turnpike.com/
Ivo Lima Brasil Jr.
Julian wrote:
>
> In article <32920D...@best.com>, Ron Karr <ka...@best.com> writes
> > However, in a money game (as this problem was stated),
> >the Jacoby rule is in effect.
>
> I wouldn't have assumed this - and I'm sure the majority wouldn't
> have. Is Jacoby really such standard practice for money games that
> it goes without saying?
>
using jacoby rule. To use it there must be a previous agreement.
Ivo BJ (IvoBJ on FIBS)
--
====================================================
Ivo Lima Brasil Jr. - Computer Science Engineer
PUC / RJ Research Team - BJ Computing Technical Director
phone : +55 (21) 511-2814
email : iv...@ele.puc-rio.br or iv...@bjcomputing.com
http://www.bjcomputing.com/ivobj
====================================================