Containment: hit or not?

Discussion:

(too old to reply)

Tim Chow

2017-01-05 01:16:48 UTC

Permalink

Walt

2017-01-05 14:38:22 UTC

Permalink

X is way behind in the race, O's board is swiss cheese, and there's a
blot to shoot at. That says hit to me - knock O off the edge and give
16 dancing numbers. Then I'll play 18/16 to take some of the sting out
of the 11 joker.

18/16 6/2*

--
//Walt

BlueDice

2017-01-05 19:56:57 UTC

Permalink

Post by Walt

ditto
--
BD

Michael

2017-01-05 17:02:26 UTC

Permalink

She's at the edge of our prime it looks we have to hit because if she jumps out with a 6 we might not be able to contain her.
On the other hand there are other issues here, X's race deficit is already high, a hit followed by a hit back will increase it even more,while X has so many crossovers... Hitting definitely risks a gammon loss, but so be it.
6/2* 18/16

badgolferman

2017-01-06 00:47:27 UTC

Permalink

I don't know about hitting. O needs a 6 to get out. I think spreading
my outboard checkers around to contain him is wiser. Let's try 18/12.

Shawn K. Quinn

2017-01-06 01:35:33 UTC

Permalink

18/12

I'm willing to concede the joker 64 for a chance to hit and contain O. I
would also like to complete my prime at some point. This play happens to
also takes the one last blot O can hit out of reach. If getting hit
wasn't potential disaster I'd even consider something like 16/10.

--
Shawn K. Quinn <***@rushpost.com>
http://www.rantroulette.com
http://www.skqrecordquest.com

Tim Chow

2017-01-06 18:14:45 UTC

Permalink

XGID=--aBBBCB------A-BcAcabc-b-:1:-1:1:42:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O X | | O O O O O | +---+
| X O | | O O O O | | 2 |
| O | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| X | | X X X X |
| X | | X X X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 120 O: 89 X-O: 0-0
Cube: 2, O own cube
X to play 42

Quite a few players I know will automatically hit in this sort of position,
where by "this sort of position" I mean a containment position where the
opponent has a single straggler at the edge of our prime and we have the
stronger board and a big race deficit. But "always hit" is a poor rule of
thumb for this sort of position. O's board may not be perfect but it is
still a four-point board. Getting hit back is very bad because then O is back
on the edge of our prime and we're now on the bar. In contrast, if we don't
hit and O rolls a 6, it's bad but there's at least not much gammon danger,
and we have some outfield coverage. And if O doesn't roll a good number,
then putting O on the bar is only slightly better for us than having her on
the 2pt.

I didn't roll out every possible play but the rollout results below are enough
to illustrate the huge difference between hitting and not hitting.

1. Rollout¹ 18/14 6/4 eq:+0.375
Player: 71.57% (G:5.25% B:0.17%)
Opponent: 28.43% (G:4.65% B:0.06%)
Confidence: ±0.006 (+0.369..+0.381) - [47.3%]

2. Rollout¹ 18/12 eq:+0.374 (-0.001)
Player: 71.86% (G:6.16% B:0.19%)
Opponent: 28.14% (G:7.22% B:0.09%)
Confidence: ±0.006 (+0.368..+0.380) - [28.0%]

3. Rollout¹ 18/14 16/14 eq:+0.374 (-0.002)
Player: 71.60% (G:5.33% B:0.16%)
Opponent: 28.40% (G:4.72% B:0.05%)
Confidence: ±0.006 (+0.367..+0.380) - [24.8%]

4. Rollout¹ 18/16 6/2* eq:+0.254 (-0.121)
Player: 70.54% (G:8.41% B:0.28%)
Opponent: 29.46% (G:16.99% B:0.17%)
Confidence: ±0.008 (+0.246..+0.262) - [0.0%]

5. Rollout¹ 16/14 6/2* eq:+0.238 (-0.137)
Player: 70.01% (G:8.61% B:0.27%)
Opponent: 29.99% (G:17.84% B:0.36%)
Confidence: ±0.008 (+0.230..+0.247) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.208.pre-release

---
Tim Chow

BlueDice

2017-01-06 19:44:46 UTC

Permalink

Post by Tim Chow
XGID=--aBBBCB------A-BcAcabc-b-:1:-1:1:42:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O X | | O O O O O | +---+
| X O | | O O O O | | 2 |
| O | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| X | | X X X X |
| X | | X X X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 120 O: 89 X-O: 0-0
Cube: 2, O own cube
X to play 42
Quite a few players I know will automatically hit in this sort of position,
where by "this sort of position" I mean a containment position where the
opponent has a single straggler at the edge of our prime and we have the
stronger board and a big race deficit. But "always hit" is a poor rule of
thumb for this sort of position. O's board may not be perfect but it is
still a four-point board. Getting hit back is very bad because then O is back
on the edge of our prime and we're now on the bar. In contrast, if we don't
hit and O rolls a 6, it's bad but there's at least not much gammon danger,
and we have some outfield coverage. And if O doesn't roll a good number,
then putting O on the bar is only slightly better for us than having her on
the 2pt.
I didn't roll out every possible play but the rollout results below are enough
to illustrate the huge difference between hitting and not hitting.
1. Rollout¹ 18/14 6/4 eq:+0.375
Player: 71.57% (G:5.25% B:0.17%)
Opponent: 28.43% (G:4.65% B:0.06%)
Confidence: ±0.006 (+0.369..+0.381) - [47.3%]
2. Rollout¹ 18/12 eq:+0.374 (-0.001)
Player: 71.86% (G:6.16% B:0.19%)
Opponent: 28.14% (G:7.22% B:0.09%)
Confidence: ±0.006 (+0.368..+0.380) - [28.0%]
3. Rollout¹ 18/14 16/14 eq:+0.374 (-0.002)
Player: 71.60% (G:5.33% B:0.16%)
Opponent: 28.40% (G:4.72% B:0.05%)
Confidence: ±0.006 (+0.367..+0.380) - [24.8%]
4. Rollout¹ 18/16 6/2* eq:+0.254 (-0.121)
Player: 70.54% (G:8.41% B:0.28%)
Opponent: 29.46% (G:16.99% B:0.17%)
Confidence: ±0.008 (+0.246..+0.262) - [0.0%]
5. Rollout¹ 16/14 6/2* eq:+0.238 (-0.137)
Player: 70.01% (G:8.61% B:0.27%)
Opponent: 29.99% (G:17.84% B:0.36%)
Confidence: ±0.008 (+0.230..+0.247) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.208.pre-release
---
Tim Chow

At DMP hitting is still an error but now by just 0.020 according to XG++. So avoiding gammon losses appears to be the most important factor in the decision not to hit.
--
BD

Michael

2017-01-07 15:58:17 UTC

Permalink

One thing we missed is that even if she rolls a 6 and we miss, she still needs to roll more than 8 to escape

Peter Percival

2017-01-07 23:50:40 UTC

Permalink

16/14 18/14

--
Do, as a concession to my poor wits, Lord Darlington, just explain
to me what you really mean.
I think I had better not, Duchess. Nowadays to be intelligible is
to be found out. -- Oscar Wilde, Lady Windermere's Fan