Rollout: The easiest contact bearoff play ever?

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

BGonline.org Forums

Rollout: The easiest contact bearoff play ever?

Posted By: Albert Steg
Date: Monday, 19 June 2017, at 9:43 p.m.

In Response To: Rollout: The easiest contact bearoff play ever? (David Rockwell)

I thought I'd post this position in order to resurrect this super-useful thread that I recently discovered while revisiting a chapter of Bill Robertie's '501 Essential Backgammon Problems' filtered through Timothy Chow's rollouts of that text:
http://timothychow.net/robertie/robertie322.html
I had this position in my "Substantial Errors" folder where I dump evaluations of positions I got wrong but haven't quite come to terms with, and it turns out David's explanation here covers it beautifully:

White is Player 1

score: 0
pip: 169

Unlimited Game
Jacoby Beaver

pip: 21
score: 0

Blue is Player 2

XGID=-bDCA--a-aaa-----bbcb-----:1:-1:1:62:0:0:3:0:10
Blue to play 62

1. Rollout¹ 4/Off 2/Off eq: +1.698
Player:
Opponent: 90.75% (G:78.57% B:13.84%)
9.25% (G:0.00% B:0.00%) Conf.: ą 0.007 (+1.691...+1.706) - [100.0%]
Duration: 1 minute 05 seconds
2. Rollout¹ 4/2 3/Off eq: +1.614 (-0.085)
Player:
Opponent: 89.45% (G:73.14% B:13.89%)
10.55% (G:0.00% B:0.00%) Conf.: ą 0.009 (+1.605...+1.623) - [0.0%]
Duration: 1 minute 12 seconds

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

eXtreme Gammon Version: 2.10

If I'm counting right, the "safe" play leaves only 11 shots while the "bold" play (ripping 2 checkers) leaves a whopping 20 -- but the latter promises to get at least 11 checkers off before getting hit when the bad sequence happens as opposed to 10. What a difference in equity!
Thanks Bill! Thanks Tim! Thanks Dave!