My bg-pet told me...
Grunty
| O O O O X | | X |
| O O O O | | X |
| O O O O | | |
| O | | |
| O | | |
| | X | [6][3] |
| | | |
| | | |
| | | |
| X | | |
| X X X | | X X |
| O X X X |[2]| X X |
+-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
Money game. X owns the cube.
He enters from the bar and has to play a 3.
What's your choice?
Håvard Raddum
I say hit with 6/3*. Not because we want to keep O on the bar or
because we want to build more inner-board points, but because we want to
keep O's last man away from the edge of our prime. If we don't hit and
O rolls a six, only 64 and 65 will force him into his own outer board
where we have very good chances of putting him on the bar again. Being
hit back is not very costly as the two open points in O's board will
never be filled. By hitting, O is twice as likely to enter on the 1 or
2 point instead of the 3, and this will give us some time to bring the
other checkers around to the prime.
paulde...@att.net
I agree with everything the first respondent said.
I would also add that, after 6/3*, X redoubles and it's a huge pass if
O misses X's blot.
Which adds another reason for 6/3*.
Paul Epstein
gammon_...@comcast.net
gammons must be considered. Hitting with the 3 allows O 14 hitting
numbers. O has the 4 point board, X has 7 checkers in the outer boards.
Who should fear the gammon more ?
If X plays Bar-19, 18-15 how bad is he ? Of O's 11 6s only 6-6 wins
outright, all other 6s leave multiple returns. While double and even
quadruple shots miss, I think I'd wait a roll and make the "passive"
play
Bob
paulde...@att.net
gammon_...@comcast.net wrote:
> This is not as obvious as it seems. It's a money game, cube in play, so
> gammons must be considered. Hitting with the 3 allows O 14 hitting
> numbers. O has the 4 point board, X has 7 checkers in the outer boards.
> Who should fear the gammon more ?
>
> If X plays Bar-19, 18-15 how bad is he ? Of O's 11 6s only 6-6 wins
> outright, all other 6s leave multiple returns. While double and even
> quadruple shots miss, I think I'd wait a roll and make the "passive"
> play
>
>
> Bob
The problem with your play is that it makes X less likely to cash the
game quickly.
The gammon parlay seems a bit remote. O needs to hit then O needs to
pop a 6 then X needs to miss etc.
Paul Epstein
Derek Ray
Hash: SHA1
Instinct says "hit".
Quiz factor says "play some fucked-up move".
Brief analysis says that:
a) 3-6 is O's only real joker.
b) If I'm hit and it isn't with 3-6, it's a quick race; I need to roll a
5 or 6 before O rolls a 6; I have plenty of outfield coverage.
c) If I don't hit, the position looks just like b) above, except I have
one less checker back... and that one checker is largely moot since I
already have men on the 19 and 20 points.
I hit and double if O doesn't hit back.
- --
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
iD8DBQFE6ln2tQZlu3o7QpERApEjAKCpRbw+5rPT7GK68IKdGO+2DXBtiwCcCRZp
iFU9wCbpgRb1AhHzjgPP3aA=
=+I/h
-----END PGP SIGNATURE-----
gammon_...@comcast.net
Now what ? X has no spares to cover the blot, Do you really think O is
dropping the rewhip ?
Bob
paulde...@att.net
gammon_...@comcast.net wrote:
> I also don't think the cash is all that obvious. X hits, O rolls 2-x.
> Now what ? X has no spares to cover the blot, Do you really think O is
> dropping the rewhip ?
>
The cash is obvious if you waste an enormous amount of time playing
backgammon. These positions occur all the time and they are standard
cashes.
Do I think O is dropping the rewhip? It depends how good O is. A
competent player would drop, yes. Taking is a clear error.
X has no spares to cover the blot but it takes O _ages_ to roll the 1
followed by the 6. Meanwhile O is crunching the home board. If O is
lucky enough to escape, X just hits him soon after the escape.
Paul Epstein
gammon_...@comcast.net
this game, expressed my opinion in a logical manner, didn't resort to
petty, sniping comments about your opinion.
If we met in person, I'd be willing to take a point to play the "huge
pass" as a prop.
Perhaps the original poster will share a Snowie or GNU analysis of this
position with us at this point.
Bob
Grunty
Needless to say, my "instinct" too would be automatically hitting.
However, my bg-pet ("X") made the waiting play 18/15. I got instantly
shocked it passed up one of its dream-shots!
Some rollout revealed that Bob got it right.
It seems the extra gammons "O" gets by rehitting make the hit not
worth, *compared with* the wins the waiting play gets for "X".
Also I find Bob's comments about "O" landing on 2-point and freezing
the prime quite accurate. I guess taking one of the other checkers to
act as a spare anywhere, even stacked on the 6-point, would shift the
preference back to the hit play. Worth a rollout too.
Most interestingly, I wondered about the other possible shot: 5-3, thus
anchoring on 20-point. Now, bg-pet preferred to hit! (didn't roll this
one out though). May be the additional safety provided by the anchor
*or/and* the somewhat poorer outfield coverage (less diversification)
make the waiting play less effective in this case, calling for a hit at
all cost.
I´ll post a Snowie rollout later, of course anybody feel free to post
one of their own.
paulde...@att.net
So Bob got the play right. Is Bob also correct when he says that if X
hits with 6/3*, and O fails to hit back, then X doesn't always have a
cash?
I was sure that these variations all lead to cashes, and it was a major
part of the reason I liked 6/3*.
Is this issue another one where Paul is wrong and Bob is right?
Yes, I know I could roll it out myself ...
Paul Epstein
Grunty
> Is Bob correct when he says that if X hits with 6/3*, and O fails to hit back,
> then X doesn't always have a cash?
I too think that O hitting and then jumping over the prime and then
avoiding double/triple shots, is a long parlay.
I´d only be sure of a pass if O lands on the ace-point, which turns
8-checkers into builders. But from the 2-point...
Problem is X´s reinforcements are far back, taking on average a couple
of turns to get into direct range. This means O will get a couple of
shots at the edge, while still retaining a respectable board. And no,
O´s board isn´t likely to break another point for the next couple of
turns.
My whole impression is O may squeeze a marginal take standing on the
2-point. Any little improvement to X´s position (say, a single spare
on the prime) or any little worsening to O´s position (say, one less
spare on the 21-point) would make a miss a pass.
gammon_...@comcast.net
EASY or CLEAR take. I think it's pretty close. I just wouldn't give it
up at that point. In a chouette I wouldn't fault anyone for taking a
rewhip, nor would I chastise someone for dropping it.
I really wish I were as certain, and as precise and correct about these
things as Paul is.
Bob
paulde...@att.net
gammon_...@comcast.net wrote:
> I certainly never claimed landing on the 2 point positioned O for an
> EASY or CLEAR take.
Actually, you did claim it was a clear take: Here's a copy-paste of
that claim of yours (enclosed in quotes).
"Do you really think O is dropping the rewhip ? "
Saying " Do you really think X is true? " is equivalent to saying
"I'm sure that X is not true ..."
So your quoted statement means "I'm sure O shouldn't drop the rewhip"
-- in other words, a clear take according to your previous posting.
But you never said it was an easy take. There's a big difference
between a clear take and an easy take. An easy take means dropping is
a blunder. A clear take means that taking is definitely preferable to
dropping.
...
In a chouette I wouldn't fault anyone for taking a
> rewhip...
I would fault a taker.
> I really wish I were as certain, and as precise and correct about these
> things as Paul is.
>
Well, you shouldn't wish you were as correct as me. My accuracy record
has been horrible.
Paul Epstein
bob_...@hotmail.com
paulde...@att.net wrote:
>
> Saying " Do you really think X is true? " is equivalent to saying
> "I'm sure that X is not true ..."
>
Do you really think that those two statements are equivalent?
Bob Koca
paulde...@att.net
Very witty. I think that "Do you really think X is true?" is a very
concise way of saying a number of things at once, of which "I'm sure
that X is not true" is only one. (Hence the ellipsis.)
My opinion is that "Do you really think X is true?" is a concise way
of making the following three statements.
"I'm sure that X is not true. I'm surprised that you appear to believe
X is true. You are invited to give confirmation and further support of
your position."
Paul Epstein
By the way, are you the same Bob posting earlier here under an ID like
gammoncast (I forget the exact ID)? It didn't occur to me that I might
be replying to Bob Koca.
bob_...@hotmail.com
No, I've always posted here recently as Bob Koca or ,Bob Koca.
In the old days possibly as bobk.
In context the statements may be equivalent or very close in meaning
but I strongly disagree that the two statements are equivalent in all
cases.
Another connotation the statement could give is an expression of
surprise.
For example suppose I am talking with someone whom I consider to be a
religious and political conservative and the person says "school
vouchers are a bad idea". I could very well agree with the statement
but yet want confirmation that the speaker believes that. Perhaps I
misheard or the person misspoke.
Another meaning is that the truth value of the issue is clouded and
the questioner is probing for more information. Suppose someone says
"there will be a terrorist attack this month". Asking if one really
believes that could get to whether someone is just thinking it is more
likely than not or if someone believes it is a strong likelihood (i.e.
they REALLY believe it will happen). This I think is closest to the
meaning of the poster who responded to you.
You posited that if someone does indeed mean that they strongly think
a pass/take declaration is incorrect then it implies they think it is
wrong by a lot. That may or may not be the actual case. Suppose I have
calculated a bearoff exactly via calculation and I know that it is a
-.998 take. Someone claims it is a pass. If I say "do you really
believe that?" I could be expressing my doubt that his statement is
incorrect while in absolutely no way saying that it is an easy take.
Bob Koca
paulde...@att.net
bob_...@hotmail.com wrote:
>
> You posited that if someone does indeed mean that they strongly think
> a pass/take declaration is incorrect then it implies they think it is
> wrong by a lot.
If I did say that, I wrote the exact opposite of what I intended. This
is my usage but I'm not completely sure it is standard.
Easy take: If doubled, a take is the correct response and a pass would
be a blunder.
Clear take: If doubled, a take is the correct response, and that has
clearly been shown.
This usage makes it very possible for a take to both clear and
marginal.
However, I think others (maybe most) use "clear take" and "easy take"
as synonyms. I think that's unfortunate, because creating distinct
meanings makes it easier to discuss a variety of concepts.
Paul Epstein
gammon_...@comcast.net
certain match scores . For instance, at 7-7 to 9, X with a 2 cube and
on roll, and gammons irrelevant, I'm pretty sure hitting is correct.
Bob
Grunty
> Grunty wrote:
> > +24-23-22-21-20-19-+---+18-17-16-15-14-13-+
> > | O O O O X | | X |
> > | O O O O | | X |
> > | O O O O | | |
> > | O | | |
> > | O | | |
> > | | X | [6][3] |
> > | | | |
> > | | | |
> > | | | |
> > | X | | |
> > | X X X | | X X |
> > | O X X X |[2]| X X |
> > +-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
> >
> > Money game. X owns the cube.
> > He enters from the bar and has to play a 3.
> > What's your choice?
>
> Needless to say, my "instinct" too would be automatically hitting.
> However, my bg-pet ("X") made the waiting play 18/15. I got instantly
> shocked it passed up one of its dream-shots!
>
> Some rollout revealed that Bob got it right.
> It seems the extra gammons "O" gets by rehitting make the hit not
> worth, *compared with* the wins the waiting play gets for "X".
Snowie full rollout:
played 3-ply (medium, 33%), cube 2-ply,
settlement 0,550 at 8 pts,
seed 100, without race database.
1. bar/19 18/15 0,456
Cubeful result: 0,890
100 games (equiv. 3730 games),
2. bar/19 6/3* 0,373 (-0,083)
Cubeful result: 0,690
100 games (equiv. 4457 games),
3. bar/19 20/17 0,369 (-0,087)
Cubeful result: 1,060
100 games (equiv. 2820 games),
> Also I find Bob's comments about "O" landing on 2-point and freezing
> the prime quite accurate. I guess taking one of the other checkers to
> act as a spare anywhere, even stacked on the 6-point, would shift the
> preference back to the hit play. Worth a rollout too.
>
> Most interestingly, I wondered about the other possible shot: 5-3, thus
> anchoring on 20-point. Now, bg-pet preferred to hit! (didn't roll this
> one out though). May be the additional safety provided by the anchor
> *or/and* the somewhat poorer outfield coverage (less diversification)
> make the waiting play less effective in this case, calling for a hit at
> all cost.
X to play (5 3)
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O O X | | X |
| O O O O | | X |
| O O O O | | | S
| O | | | n
| O | X | | o
| |BAR| | w
| | | | i
| | | | e
| 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: 147 O: 54 X-O: 12-13 (1)
CubeValue: 2, X owns Cube
Snowie full rollout:
played 3-ply (medium, 33%), cube 2-ply,
settlement 0,550 at 8 pts,
random seed, without race database.
1. bar/20 6/3* 0,392
Cubeful result: 0,960
100 games (equiv. 3792 games),
2. bar/20 18/15 0,390 (-0,002)
Cubeful result: 0,600
100 games (equiv. 4309 games),
Both plays are practically equivalent, trading single wins for gammons.
Grunty
> paulde...@att.net wrote:
> > Is Bob correct when he says that if X hits with 6/3*, and O fails to hit back,
> > then X doesn't always have a cash?
>
> I too think that O hitting and then jumping over the prime and then
> avoiding double/triple shots, is a long parlay.
>
> I´d only be sure of a pass if O lands on the ace-point, which turns
> 8-checkers into builders.
X on roll, cube action
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O O X X | | X |
| O O O O | | X |
| O O O O | | | S
| O | | | n
| O | | | o
| |BAR| | w
| | | | i
| | | | e
| | | |
| X X X | | X X |
| O X X X X | | X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 138 O: 56 X-O: 12-13 (1)
CubeValue: 2, X owns Cube
Rollout Money equity: 0,677
0,2% 6,2% 85,3% 14,7% 9,0% 0,3%
Full rollout,
50 games (equiv. 1723 games),
played 3-ply (medium, 50%), cube 3-ply,
settlement 0,550 at 8 pts,
seed 2121, without race database.
Cubeless
Double decision Good enough: Huge (Huge/0,50)
Too good to double: 3%
Take decision Current Doubled Borderline
Equities: 0,677 0,677 0,564
Proper cube action: Double, pass
Cubeful
Double decision Player owns cube: 0,760
Center cube: 0,780
Opponent owns cube: 1,040
Take decision Opponent takes: 1,040
Borderline take/pass: 1,000
Proper cube action: Redouble, pass
> But from the 2-point...
> Problem is X´s reinforcements are far back, taking on average a couple
> of turns to get into direct range. This means O will get a couple of
> shots at the edge, while still retaining a respectable board. And no,
> O´s board isn´t likely to break another point for the next couple of
> turns.
>
> My whole impression is O may squeeze a marginal take standing on the
> 2-point. Any little improvement to X´s position (say, a single spare
> on the prime) or any little worsening to O´s position (say, one less
> spare on the 21-point) would make a miss a pass.
X on roll, cube action
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O O X X | | X |
| O O O O | | X |
| O O O O | | | S
| O | | | n
| O | | | o
| |BAR| | w
| | | | i
| | | | e
| | | |
| X X X | | X X |
| O X X X X | | X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 138 O: 55 X-O: 12-13 (1)
CubeValue: 2, X owns Cube
Rollout Money equity: 0,553
0,2% 5,8% 80,7% 19,3% 11,8% 0,3%
Full rollout,
100 games (equiv. 3335 games),
played 3-ply (medium, 50%), cube 3-ply,
settlement 0,550 at 8 pts,
random seed, without race database.
Cubeless
Double decision Good enough: 205% (1,02/0,50)
Too good to double: 0%
Take decision Current Doubled Borderline
Equities: 0,553 0,553 0,563
Proper cube action: Double, take
Cubeful
Double decision Player owns cube: 0,970
Center cube: 1,010
Opponent owns cube: 1,260
Take decision Opponent takes: 1,260
Borderline take/pass: 1,000
Proper cube action: Redouble, pass
This seems to be a pass too. (More extensive/precise rollouts are
welcome.)