Automatic play - or not
Grunty
X to play [2] [1]
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| X O O | | O O O X |
| X O O | | O O X |
| O | | O O X |
| | | O |
| | | |
| |BAR| [2] [1] |
| | | |
| | | |
| | | X |
| X X X | | O X |
| X X X | | O X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
Pipcount X: 151 O: 126
CubeValue: 1
tc...@lsa.umich.edu
Grunty <grunti...@yahoo.com> wrote:
>X to play [2] [1]
>
>+24-23-22-21-20-19-------18-17-16-15-14-13-+
>| X O O | | O O O X |
>| X O O | | O O X |
>| O | | O O X |
>| | | O |
>| | | |
>| |BAR| [2] [1] |
>| | | |
>| | | |
>| | | X |
>| X X X | | O X |
>| X X X | | O X X |
>+-1--2--3--4--5--6--------7--8--9-10-11-12-+
I would play 10/8 24/23. O still has two checkers back, so my best chance
is to make an advanced anchor; hanging back on the 24-point is a losing
proposition. Thus I must split. My slightly stronger board protects me
against a loose hit. And since I can also clean up my blot, I do so.
The alternative of 13/10 looks considerably worse to me. It leaves me
with a stripped position with very little flexibility, and does nothing to
challenge O's plan to prime me.
If X were a lot further behind in the race then hanging back on the 24-point
might be right because the only chance of winning would be hitting a shot,
and the ace-point game gives better chances of that. In the given position,
though, X is not that desperate yet.
--
Tim Chow tchow-at-alum-dot-mit-dot-edu
The range of our projectiles---even ... the artillery---however great, will
never exceed four of those miles of which as many thousand separate us from
the center of the earth. ---Galileo, Dialogues Concerning Two New Sciences
Walt
Making the 10 point is the "obvious" choice. But it leads to a TMP
problem - Too Many Points. We don't need to make the 10 point. We need
to safety the blot on the 10 and get our back men moving. With a better
board we shouldn't be shy about splitting our back men and exchanging hits.
So, 10/8 24/23 for me. (Although I'd probably play the automatic move
over the board without thinking about it very much.)
//Walt
tc...@lsa.umich.edu
Walt <walt_...@SHOESyahoo.com> wrote:
>Making the 10 point is the "obvious" choice.
>So, 10/8 24/23 for me. (Although I'd probably play the automatic move
>over the board without thinking about it very much.)
That's interesting...for me, the splitting play is the automatic one. In the
opening, I look to split as soon as possible. This is probably due to my
being influenced by bots. The error I tend to make is to try to split even
when it's too late and I have to stay on the 24-point to avoid getting
blitzed. In this position, though, I don't think the blitz is that big
a threat, so I think splitting is right.
David C. Ullrich
>In article <mfeim.545326$4p1....@en-nntp-03.dc1.easynews.com>,
>Walt <walt_...@SHOESyahoo.com> wrote:
>>Making the 10 point is the "obvious" choice.
>[...]
>>So, 10/8 24/23 for me. (Although I'd probably play the automatic move
>>over the board without thinking about it very much.)
>
>That's interesting...for me, the splitting play is the automatic one.
Heh. I otoh couldn't figure out what the "automatic" play mentioned
in the title of the thread was supposed to be.
> In the
>opening, I look to split as soon as possible. This is probably due to my
>being influenced by bots.
Heh^2. Seems to me _my_ problem is not splitting when I should.
Has to do with learning the game a few decades ago when splitting
was out of fashion and slotting the five point was the thing to do.
(Then for a few years recently all my opponents were much weaker
than me and had no idea how to defend against a back game - that
didn't help me get over my tendency to slot instead of split.
Alas one of the guys has improved a great deal recently. Seems to
me when he splits he escapes or makes an anchor, when I split
I get blitzed...)
>The error I tend to make is to try to split even
>when it's too late and I have to stay on the 24-point to avoid getting
>blitzed. In this position, though, I don't think the blitz is that big
>a threat, so I think splitting is right.
Here the split seems clear even to me, for various reasons that
have already been mentioned.
David C. Ullrich
"Understanding Godel isn't about following his formal proof.
That would make a mockery of everything Godel was up to."
(John Jones, "My talk about Godel to the post-grads."
in sci.logic.)
Grunty
Snowie full rollout, played 3-ply, cube 2-ply
1. 24/22 10/9 -0,463
5.7% 34.2% 65.8% 19.9%
Cubeful result: -0,040
50 games (equiv. 1292 games),
2. 24/23 10/8 -0,552 (-0,089)
5.8% 30.4% 69.6% 21.4%
Cubeful result: -1,600
50 games (equiv. 1047 games),
3. 10/9 8/6 -0,577 (-0,114)
5.3% 27.7% 72.3% 17.1%
Cubeful result: -1,440
50 games (equiv. 1325 games),
4. 13/10 -0,579 (-0,116)
5.3% 28.5% 71.5% 19.2%
Cubeful result: -0,680
50 games (equiv. 1096 games),
Visually, "automatic play" would be 13/10. Strategically, it would be
24/23 10/8.
No matter what, both turned out to be erroneous, the one conceptually
and the other technically.
The plan for X here is anticipating O's likely prime, so mobilizing
his back men is mandatory. 24/23 doesn't accomplish much in this
sense. Ask yourself how would you feel when O makes his 5-point. It
wouldn't make much difference whether both your back men stand on the
24 or one on the 23. Would you take a double in either case?
24/22 10/9 is certainly a very bold play. You leave even another man
exposed, but maximize your chances against the coming prime. However
both split plays don't seem to differ much in gammon losses, so the
bigger split is preferable.
In the actual game where this position was taken from, Snowie was
playing X. I doubled, Snowie took and eventually ended up gammoned.
But it made no complaint at all, so I guess its play was correct.
tc...@lsa.umich.edu
Grunty <grunti...@yahoo.com> wrote:
>Visually, "automatic play" would be 13/10. Strategically, it would be
>24/23 10/8.
>No matter what, both turned out to be erroneous, the one conceptually
>and the other technically.
>
>The plan for X here is anticipating O's likely prime, so mobilizing
>his back men is mandatory. 24/23 doesn't accomplish much in this
>sense. Ask yourself how would you feel when O makes his 5-point. It
>wouldn't make much difference whether both your back men stand on the
>24 or one on the 23. Would you take a double in either case?
Very nice analysis! Thanks for posting this position. Once you point it
out, it is clear that 24/22 is mandatory.
Two reasons I missed this play are (1) I neglected to note that O will be
somewhat reluctant to break anchor to hit, and (2) I tend to underestimate
the defensive potential of having just one checker back. If X can't anchor
but has an opportunity to leap one checker out safely, then it is probably
right to do so in this position. Better to have one unanchored checker
back than two.
I did some rolling out with gnubg to confirm. The following are the results
of 1296 games each at 2-ply:
24/22 10/9 -0.854
24/22 24/23 -0.939
13/10 -0.953
24/23 10/8 -1.000
The cube is coming in any case. After my suggestion, 24/23 10/8, gnubg says
I can't take the double, so the play is tantamount to resignation!