[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

Prime-rolling technique

Posted By: Nack Ballard
Date: Saturday, 23 February 2019, at 5:16 p.m.

In Response To: XG RO (Casper Van der Tak)

It depends on distribution. Starting with the ugly checker on the 4pt is much of the reason that hitting with the 3 is best in the original position. That is why I moved it back to the 6pt and changed the roll to 65 here, so that it is possible to roll the prime forward without committing a spare to the 4pt.

Below is that rollout again (but 4-ply, as I'm short on time today), this time with White's checkers all stacked on the 1pt:






White is Player 2

score: 0
pip: 31
7 point match
pip: 89
score: 0

Blue is Player 1
XGID=-aAABBCBBB--------------g-:0:0:1:65:0:0:0:7:10

No chance to pick up a second checker


1.Rollout18/3 8/2eq: +0.274
Player:
Opponent:
57.18% (G:0.00% B:0.00%)
42.82% (G:0.40% B:0.01%)
Conf.: ± 0.003 (+0.271...+0.277) - [70.9%]
Duration: 1 minute 09 seconds
2.Rollout19/3 6/1* eq: +0.273 (-0.001)
Player:
Opponent:
57.24% (G:0.00% B:0.00%)
42.76% (G:0.57% B:0.01%)
Conf.: ± 0.003 (+0.270...+0.276) - [29.1%]
Duration: 1 minute 33 seconds
1 5184 Games rolled with Variance Reduction.
Dice Seed: 77188237
Moves and cube decisions: 4-ply


[Naturally, 9/4 9/3 comes in third -- just as 9/6 9/3 would come in third in Highlander's original 63 position even if White's checkers are all on the 1pt -- because it puts/keeps a checker on the 4pt (too advanced.)]

However, even that situation is biased, because one or both of the semi-advanced 9pt checkers might be forced as low as the 4pt or even the 3pt.

To further isolate the prime-rolling decision, let's move one of the 9pt checkers into the back quadrant. Blue's position is more fluid, and hitting on the 1pt (before making the 2pt) becomes a noticeable error:






White is Player 2

score: 0
pip: 31
7 point match
pip: 103
score: 0

Blue is Player 1
XGID=-aAABBCBBA-------------Ag-:0:0:1:65:0:0:0:7:10

Fluid choice


1.Rollout18/3 8/2eq: +0.326
Player:
Opponent:
58.91% (G:0.00% B:0.00%)
41.09% (G:0.05% B:0.00%)
Conf.: ± 0.003 (+0.323...+0.329) - [100.0%]
Duration: 35.9 seconds
2.Rollout19/3 6/1* eq: +0.307 (-0.019)
Player:
Opponent:
58.19% (G:0.00% B:0.00%)
41.81% (G:0.32% B:0.01%)
Conf.: ± 0.003 (+0.304...+0.310) - [0.0%]
Duration: 56.9 seconds
1 5184 Games rolled with Variance Reduction.
Dice Seed: 36223632
Moves and cube decisions: 4-ply


In case that seems like sleight of hand, here's a rollout that more simply compares covering the 2pt vs hitting on the 1pt:






White is Player 2

score: 0
pip: 31
Unlimited Game
Jacoby Beaver
pip: 100
score: 0

Blue is Player 1
XGID=-aABBBBCB---------------gA:0:0:1:62:0:0:3:0:10

Simple 6 comparison


1.Rollout1Bar/23 8/2eq: +0.314
Player:
Opponent:
58.50% (G:0.00% B:0.00%)
41.50% (G:0.09% B:0.00%)
Conf.: ± 0.004 (+0.310...+0.317) - [100.0%]
Duration: 43.9 seconds
2.Rollout1Bar/23 7/1* eq: +0.301 (-0.012)
Player:
Opponent:
58.16% (G:0.00% B:0.00%)
41.84% (G:0.36% B:0.01%)
Conf.: ± 0.004 (+0.298...+0.305) - [0.0%]
Duration: 1 minute 06 seconds
1 5184 Games rolled with Variance Reduction.
Dice Seed: 77188237
Moves and cube decisions: 4-ply


Hope that helps.

Nack

Messages In This Thread

 

Post Response

Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:

 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.