Saturday, June 11, 2011

checker play #062

is Player 2

score: 1
pip: 225
9 point match
pip: 107
score: 6

is Player 1
XGID=-dbBBaDBC---bA---aAda-----:0:0:1:22:6:1:0:9:10
to play 22

Svaki potez koji barem ne udari na 5.polje je blunder. Čak je 13/5* koji propušta priliku zatvoriti polje drugi potez po snazi koji donosi i više gammona i običnih pobjeda od svih kandidata osim onog najboljeg. Očito bijeli treba biti gramzljiv i pokušati iscijediti iz pozicije što više može tj. ići po još protivnikovih checkera, bilo da pričeka na 18.polju jedinicu ili da sa bara, ukoliko ga crni udari, pokuša pokupiti checker sa 20.polja. Nakon 13/11 bijeli treba početi  razmišljati o tome kako očistiti 6.polje, a 6/4, kao bonus, popravlja i distribuciju. 

Every move which misses opportunity to make or at least hit loose on 5pt is blunder. White needs that point most. 13/5* wins more gammons and more wins than any other play except winning one. Reason for that is greed, which is desirable here when white should try to squeeze the most equity from position. He has to stay on black's barpt in order to hit blot there or after 13/5*, if black rehit, white gets chance from the bar to pick up black's homeboard blot. After 13/11 he starts to think how to clear 6pt and improve spare distribution.

1.Rollout113/11 7/5*(2) 6/4eq: +0,704
Player:
Opponent:		61,60% (G:41,49% B:22,04%)
38,40% (G:0,69% B:0,03%)		Conf: ± 0,017 (+0,687...+0,721)
Duration: 5 hours 37 minutes
2.Rollout113/5* eq: +0,669 (-0,036)
Player:
Opponent:		62,64% (G:41,95% B:19,71%)
37,36% (G:1,57% B:0,07%)		Conf: ± 0,017 (+0,652...+0,686)
Duration: 5 hours 14 minutes
3.Rollout113/9 7/5*(2) eq: +0,664 (-0,041)
Player:
Opponent:		60,82% (G:40,84% B:21,16%)
39,18% (G:0,60% B:0,02%)		Conf: ± 0,017 (+0,647...+0,681)
Duration: 4 hours 25 minutes
4.Rollout28/4 7/5*(2) eq: +0,661 (-0,044)
Player:
Opponent:		60,93% (G:40,85% B:20,89%)
39,07% (G:0,71% B:0,03%)		Conf: ± 0,017 (+0,644...+0,678)
Duration: 4 hours 02 minutes
5.Rollout118/14 7/5*(2) eq: +0,657 (-0,047)
Player:
Opponent:		61,24% (G:40,76% B:19,77%)
38,76% (G:0,65% B:0,04%)		Conf: ± 0,016 (+0,641...+0,673)
Duration: 4 hours 49 minutes
6.Rollout218/16 13/11 7/5*(2) eq: +0,634 (-0,071)
Player:
Opponent:		60,60% (G:40,16% B:19,78%)
39,40% (G:0,60% B:0,04%)		Conf: ± 0,016 (+0,618...+0,650)
Duration: 4 hours 11 minutes
1 9932 Games rolled with Variance Reduction.
Moves and cube decisions: 3 ply

2 9931 Games rolled with Variance Reduction.
Moves and cube decisions: 3 ply
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross
Labels:

2 comments:

  1. DavidJune 11, 2011 at 4:47 PM

    it strikes me as unlikely that 6/4 is .071 better than 18/16 after 7/5(2) 13/11. my guess is that the relative bot errors favor 6/4.

    ReplyDelete
  2. Dubrovnik Backgammon MagazineJune 11, 2011 at 11:54 PM

    18/16 13/11 7/5*(2) was my OtB choice. This is not statistical significant rollout regarding absolute equity. There is still +/-0.017 interval, so with more trials I guess difference between these plays will be smaller.

    Here I tried to find some reasons behind rollout winner. Another reason to stay on 18pt could be timing. White maybe likes to be hit in order to force black to bring those checkers from midpoint closer so result will be black's army divided into two pieces.

    ReplyDelete

Note: Only a member of this blog may post a comment.