FIBS Board backgammon forum
Backgammon => Fibsboard Forum Matches => FFM 7 => Topic started by: ah_clem on October 22, 2010, 01:10:50 AM
Zorba opted not to cube so he has a 5-4 to play
Position ID: hteMAyDYnoUBUA
Match ID: MAG2AAAACAAA
^^^^
Just making the golden anchor is tempting strategically, but tactically, the position asks for a bold play (I have more men back, I have a stronger board, opponent has a blot laying around, I'm not ahead in the race): B/21 6/1*. 25% dancing numbers are very juicy, and some other entering numbers don't play well. I may get a single or even double shot at sending a second man back. If I get hit myself, it's not too big a deal usually.
XG rollout + comments:
Spoiler
A big whopper by me. Apparently strategy overrides tactics here! Maybe it's the several double return-hits the Herd gets that make the loose hit this bad. The combination of not anchoring at the golden point but split instead, and hitting loose is too much here, even if several indicators point towards a bold play.
Just looking at the trade-off between single wins and gammon wins, the loose hit could be a good idea: it wins 3.5% less games but wins 8% more gammons. The catch is that the loose hit also loses a lot more gammons: 9%! That, combined with the cube still in the center pointing towards a less volatile play, makes the bold approach a big blunder.
XGID=-a--BBD-BA--cB-b-bad---bAA:0:0:1:54:0:1:0:5:10
1. Rollout¹ Bar/20 24/20 eq:+0,500
Player : 60,67% (G:18,05% B:0,39%)
Opponent: 39,33% (G:5,59% B:0,36%)
Confidence: +/- 0,009 (+0,491<E<+0,509)
Duration: 39 minutes 43 seconds
2. Rollout¹ Bar/21 6/1* eq:+0,349 (-0,151)
Player : 57,17% (G:26,05% B:1,20%)
Opponent: 42,83% (G:14,69% B:1,92%)
Confidence: +/- 0,010 (+0,339<E<+0,359)
Duration: 35 minutes 13 seconds
¹ 10368 Games rolled with Variance Reduction.
Moves: 1 ply, cube decisions: 3 ply
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
gnu rollout
Spoiler
Different bot, same story (c:
I think the right idea here is to make the golden point, then attack the single checker from a position of strength. It's likely not going anywhere anytime soon. Hitting now is premature.
1. Rollout bar/20 24/20 Eq.: +0.536
0.608 0.191 0.005 - 0.392 0.069 0.004 CL +0.401 CF +0.536
[0.002 0.002 0.001 - 0.002 0.002 0.001 CL 0.006 CF 0.015]
2. Rollout bar/21 6/1* Eq.: +0.316 ( -0.220)
0.567 0.267 0.012 - 0.433 0.159 0.018 CL +0.252 CF +0.316
[0.002 0.002 0.001 - 0.002 0.002 0.002 CL 0.006 CF 0.014]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 903967096 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]