|
Position 1.
# Ply Move Equity
1 3 bar/23* 3/2 0.343
2 3 bar/23* 6/5 0.282 (-0.061)
3 3 bar/23* 4/3 0.272 (-0.071)
4 2 bar/23* 8/7 0.135 (-0.208)
5 1 bar/23* 5/4 -0.290 (-0.633)
This is a bit of a cheeky one. I just wanted to see who was alert
enough to spot the duplication of twos, which I think is the idea
here.
In the game, I missed it and played 6/5with the one. I am happy to
roll this one out if anyone wants but I think Snowie gets these
ones right on 3 ply.
Bar 23*3/2: 5 points
All other plays: 1 point
Position 2
Cube action equity
Rollout Money equity: 0.396
0.9% 17.3% 65.1% 34.9% 8.3% 0.3%
95% confidence interval:
- money cubeless eq.: 0.396 ±0.051,
- live cube no double: 0.570 ±0.114,
- live cube double take: 0.547 ±0.147.
Rollout settings:
Full rollout,
108 games (equiv. 2888 games),
played 3-ply (fast), cube 3-ply,
seed 1, without race database.
Evaluations
1. No double 0.631
2. Double, take 0.605 (-0.027)
3. Double, pass 1.000 (+0.369)
Proper cube action: No double, take 7%
Live cube
1. No double 0.570
2. Double, take 0.547 (-0.023)
3. Double, pass 1.000 (+0.430)
Proper cube action: No double, take 5%
This looked like a pass to me, and I passed it in the game.
Far be it from me to provide an analysis, there a couple of
salient things here:
1. Green has a number of rolls which don’t do much or are
actually bad e.g 6-3, 6-2, 2-2, 1-1
2. The 21 point is very strong for white and he never gets wiped
out except in the variations where he runs off the anchor and
gets pointed on.
3. White actually has some firepower to attack green and can
afford to with the 21 point.
No double/take: 7 points;
Double/take: 3 points;
Double/pass: 0 points
Position 3
Cube action equity
Rollout Money equity: 0.615
0.9% 33.5% 68.3% 31.7% 9.2% 0.2%
95% confidence interval:
- money cubeless eq.: 0.615 ±0.038,
- live cube no double: 0.641 ±0.061,
- live cube double take: 0.968 ±0.108.
Rollout settings:
Full rollout,
108 games (equiv. 4660 games),
played 3-ply (fast), cube 3-ply,
settlement 0.550 at 4 pts,
seed 1, without race database.
Evaluations
1. Double, take 0.999
2. No double 0.690 (-0.309)
3. Double, pass 1.000 (+0.001)
Proper cube action: Double, take
Live cube
1. Double, take 0.968
2. No double 0.641 (-0.327)
3. Double, pass 1.000 (+0.032)
Proper cube action: Double, take
This is very close. Green gets gammoned lot but wins 31.7% of
games.
Over the board I’m not sure if there’s anyway to calculate this
without actually knowing the position or one very similar.
Double/take: 5 points
Double/pass: 4 points
No double/take: 0 point
Position 4
Cube action equity
Rollout Money equity: 0.492
1.6% 28.8% 68.2% 31.8% 16.4% 1.3%
95% confidence interval:
- money cubeless eq.: 0.492 ±0.048,
- live cube no double: 0.438 ±0.063,
- live cube double take: 0.670 ±0.140.
Rollout settings:
Full rollout,
72 games (equiv. 2887 games),
played 3-ply (fast), cube 3-ply,
settlement 0.550 at 4 pts,
seed 1, without race database.
Evaluations
1. Double, take 0.737
2. No double 0.490 (-0.248)
3. Double, pass 1.000 (+0.263)
Proper cube action: Double, take
Live cube
1. Double, take 0.670
2. No double 0.438 (-0.232)
3. Double, pass 1.000 (+0.330)
Proper cube action: Double, take
This position confounded me and I passed over the board,
although I wasn’t sure.
I thought that even if white fans green still has a disjointed
position, but if white does, green has quite a decent attack on
with most dice rolls working for him.
If white hits one man, green still has the four point and even a
two-four back game. I still don’t like the position but there
you go.
Double/ take = 7 points,
No double/take = 5 points,
Double/pass = 0 points
|