Gammonfever BG Quiz - Position #13

[Micke Nilsson]

X is doubling. Take or pass? And please, no bot solutions right away :o)

Money session. Score X-O: 0-0

X on roll, cube action
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O | | X O X |
| O O | | X O X |
| O O | | O | S
| O | | | n
| | | | o
| |BAR| | w
| | | | i
| X | | | e
| X | | O |
| X X X | | X O |
| O X X X | | X X O |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 126 O: 122 X-O: 0-0/Money (1)
CubeValue: 1

[danc...@yahoo.com]

This one looks a little tougher than some of the previous ones. The
position would lead me to guess to take. X doesn't have a lot of
market losers (53, 64, 55, maybe a couple of others). I could be wrong.

[paulde...@att.net]

It's a pass because of 0's weak stack on the 23 point. Without that
stack, the position resembles a classic double/take position along the
lines of 501 Essential problems by Robertie. But the weak stack
changes things a lot.

I think it's quite a big pass -- my guess is that taking loses over 0.1
points.

Paul Epstein

[Grunty]

18 rolls enlarge the prime.
Plus 22, 44 and 66 also very strong numbers.

O has a checker almost wasted on the 23 point.
His 5 point isn't made yet, in case he manages to step up to the edge
of the prime exposing himself to the attack.

I vote for a pass.

[paulde...@att.net]

Yes, after further thought, I'm now even more confident that it's a
pass. A great way to play backgammon is to continually read and reread
Robertie's 501 Essential .... and compare the position with a
near-equivalent from the book.

Look at position 215. This is Double/Take. However, the current
position is like 215 but strongly improved. The upcoming prime is
stronger, the number of rolls that make it is larger, and the 23 point
stack is horrendous. It's a clear pass, not a marginal pass.

Incidentally, the 23 stack looks like it was forced by a 66. That type
of stack occurs a lot and is worth thinking about in more general
situations.

Paul Epstein

[danc...@yahoo.com]

So what's the bot solution for this one?

[paulde...@att.net]

Well, Micke did say "no bot solutions _right away_." He didn't say "no
bot solutions at all". I think there's been a decent enough interval.
So, no one should mind if anyone posts one -- it doesn't need to be
Micke.

Paul Epstein

[paulde...@att.net]

A friend of mine researched this with gnu -- I was too lazy to check it
for myself. Apparently, it is a pass but taking only loses 0.076.

When it comes to this level of marginality, I think it is more or less
guesswork whether to take or pass and it was just chance that I got the
right answer and not dancl...@yahoo.com

I'm not even completely right because I saw it as a clear pass, not a
marginal one.

Any Snowie opinions?

Paul Epstein

[David C. Ullrich]

On 30 Sep 2005 14:06:13 -0700, paulde...@att.net wrote:

>A friend of mine researched this with gnu -- I was too lazy to check it
>for myself. Apparently, it is a pass but taking only loses 0.076.
>
>When it comes to this level of marginality, I think it is more or less
>guesswork whether to take or pass and it was just chance that I got the
>right answer and not dancl...@yahoo.com
>
>I'm not even completely right because I saw it as a clear pass, not a
>marginal one.

If it's any consolation it looks like a clear pass to me too.

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

David C. Ullrich

[paulde...@att.net]

I suppose I'm going to play the role of a bad student blaming his
teachers. As I said before, I used position 215 from 501 Essential...
as a reference, and judged that this position was a huge improvement on
215. However, Robertie never gave any indication of the marginality of
the cube action, and that makes it difficult to use such positions as
references.

Presumably, position 215 is very-marginal-double/ take (?)

Paul Epstein

[Micke Nilsson]

Hi,

I've been away for a few days, but now I'm back and I'll try to make the
rollout and post it tomorrow.

/Micke

<danc...@yahoo.com> skrev i meddelandet
news:1128086790....@g44g2000cwa.googlegroups.com...

[Philippe Michel]

Le 01-10-2005, paulde...@att.net <paulde...@att.net> a écrit :
> I suppose I'm going to play the role of a bad student blaming his
> teachers. As I said before, I used position 215 from 501 Essential...

> as a reference [...]

There is a huge difference between pos. #13 and pos. 215 from Robertie's
book (reproduced at the end of this post for those who don't have the
book). In the former, the backman is primed ; in the latter it is at the
edge of the prime.

This totally changes the dynamics of the position (in addition to the
equity :-). In pos. 215, X must hit the backman if he can. He will be hit
back quite often. If he cannot hit, the backman has significant chances to
escape. X could blow O away, but he could easily stumble as well. On the
other hand, in pos. #13, the most likely result of the next roll is that
nothing important happens : X breaks his midpoint and O moves 2 checkers.

Pos. #13 is more similar to what pos. 215 may become 2 rolls later : X
pointed on the backman, which then entered on the 24pt ; market lost.

> Presumably, position 215 is very-marginal-double/take (?)

It is, according to a gnubg rollout (and position 215 but with the backman
on the 24pt is a marginal take/drop ; *that* would have been a fine
reference position for the problem #13).

>> >A friend of mine researched this with gnu -- I was too lazy to check it
>> >for myself. Apparently, it is a pass but taking only loses 0.076.

1.076 is the 2ply evaluation. A rollout gives a much clearer pass.

Pos. 215 from _501 essential backgammon problems_ :

GNU Backgammon Position ID: WOeRAwSw25gHAA
Match ID : cAkAAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: White
| X O O | | O O O | 0 points
| X O | | O O |
| X O | | O |
| X O | | |
| | | |
v| |BAR| | (Cube: 1)
| | | |
| | | |
| O | | X |
| O X X X | | X X | On roll
| O X X X | | X X O | 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: Black
Pip counts: O 134, X 132

[paulde...@att.net]

I regard 215 as a reference for #13 in the sense that (as you correctly
say) #13 resembles a later development of 215 after X has lost its
market. Admittedly, I didn't phrase it in this way. However, I did
say that they were like each other _except that #13 is hugely better
for the favourite_. I then outlined a few ways in which #13 is better.

Nowhere did I suggest that 215 and #13 are very similar. Obviously, I
saw them as very different which is why I evaluated #13 as a clear pass
and 215 as only a marginal double.

I was puzzled by the 1.076 evaluation of position #13. The reason is
that the 1.076 evaluation represents a very efficient double. Assuming
what we both agree on, that #13 resembles a lost-market continuation of
pos. 215, it makes no sense to double early and prevent a lost market
if your typical market loss is only 0.076

All is explained if the 1.076 evaluation for #13 is inaccurate, and the
pass is actually much bigger.

Paul Epstein

[Micke Nilsson]

3-ply instant evaluation:

Money session. Score X-O: 0-0

X on roll, cube action
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O | | X O X |
| O O | | X O X |
| O O | | O | S
| O | | | n
| | | | o
| |BAR| | w
| | | | i
| X | | | e
| X | | O |
| X X X | | X O |
| O X X X | | X X O |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 126 O: 122 X-O: 0-0/Money (1)
CubeValue: 1

3-Ply Money equity: 0,664
0,6% 23,7% 73,7% 26,3% 5,1% 0,2%
1. Double, pass 1,000
2. No double 0,914 (-0,086)
3. Double, take 1,137 (+0,137)
Proper cube action: Double, pass

------------------------------ End ----------------------------------

3-ply full rollout:

Money session. Score X-O: 0-0

X on roll, cube action
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| O O O | | X O X |
| O O | | X O X |
| O O | | O | S
| O | | | n
| | | | o
| |BAR| | w
| | | | i
| X | | | e
| X | | O |
| X X X | | X O |
| O X X X | | X X O |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 126 O: 122 X-O: 0-0/Money (1)
CubeValue: 1

Rollout Money equity: 0,696
0,3% 22,1% 76,1% 23,9% 4,9% 0,2%
95% confidence interval:
- money cubeless eq.: 0,696 ą0,020,
- live cube no double: 0,932 ą0,024,
- live cube double take: 1,233 ą0,046.
Rollout settings:
Full rollout,
324 games (equiv. 11588 games),
played 3-ply (standard), cube 3-ply,
settlement 0,550 at 4 pts,
seed 1, with race database.
Evaluations
1. Double, pass 1,000
2. No double 0,937 (-0,063)
3. Double, take 1,218 (+0,218)
Proper cube action: Double, pass
Live cube
1. Double, pass 1,000
2. No double 0,932 (-0,068)
3. Double, take 1,233 (+0,233)
Proper cube action: Double, pass

------------------------------ End ----------------------------------

[danc...@yahoo.com]

Thanks Micke. Interesting position. I set it up and played it by hand
several times. Looks like my checker play may need a lot of work. I
was winning with the X checkers only about 70% of the time and getting
a closer decision than this.

Here's a question. If X rolls a 6-1 initially, should X hit loose or
make some other play?

Micke Nilsson wrote:
> 3-ply instant evaluation:

>
> Rollout Money equity: 0,696
> 0,3% 22,1% 76,1% 23,9% 4,9% 0,2%
> 95% confidence interval:

> - money cubeless eq.: 0,696 ±0,020,
> - live cube no double: 0,932 ±0,024,
> - live cube double take: 1,233 ±0,046.

[danc...@yahoo.com]

Thanks Micke. Interesting position. I set it up and played it by hand
several times. Looks like my checker play may need a lot of work. I
was winning with the X checkers only about 70% of the time and getting
a closer decision than this.

Here's a question. If X rolls a 6-1 initially, should X hit loose or
make some other play?

Micke Nilsson wrote:
> 3-ply instant evaluation:
>

> Rollout Money equity: 0,696
> 0,3% 22,1% 76,1% 23,9% 4,9% 0,2%
> 95% confidence interval:

> - money cubeless eq.: 0,696 ±0,020,
> - live cube no double: 0,932 ±0,024,
> - live cube double take: 1,233 ±0,046.

[paulde...@att.net]

I don't like 8-1 if X rolls 61 because it makes it harder for X to
make a 5 prime. My play for 61 is 13-7 6-5. (Assuming that O owns the
cube.)

Paul Epstein

[danc...@yahoo.com]

paulde...@att.net wrote:
> I don't like 8-1 if X rolls 61 because it makes it harder for X to
> make a 5 prime. My play for 61 is 13-7 6-5. (Assuming that O owns the
> cube.)
>
> Paul Epstein

I was playing this roll wrong and going for the loose hit. Breaking
the midpoint is not as dangerous as it looks. Thanks.