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

BGonline.org Forums

6/5 4/1* ... 4-1

Posted By: Daniel Murphy
Date: Monday, 7 May 2007, at 6:57 a.m.

In Response To: Re: Re: Re: GNU-bg rollout (Chuck Bower)

The score (after 0 games) is: White 2, Blue 0 (match to 13 points)

Move number 6: White to play 41

White103


3O4O2X2X3X2X3X2X ' ' ' '

1O3O2O1O1O ' ' ' ' ' ' '

Blue186

Position ID: 3QoAwDtstxsAQA Match ID: EQGmASAAAAAA

• Gnubgh 2-ply moves bar/21*/20*

# Ply Move Equity
1 R bar/21* 7/6 +1.28468
0.827440.566650.14628-0.172560.000000.00004 +1.37900 +1.28468
0.005490.006820.00855-0.005490.002290.00002 0.01981 0.02378
Full cubeful rollout with var.redn.
648 games, Mersenne Twister dice gen. with seed 1058468222 and quasi-random dice
Play: world class 2-ply cubeful prune [world class]
keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
Skip pruning for 1-ply moves.
Cube: 2-ply cubeful prune [world class]
2 R bar/21*/20* +1.17868 ( -0.10600)
0.778790.569630.17439-0.221210.000080.00009 +1.31183 +1.17868
0.006740.008150.00943-0.006740.000920.00002 0.02447 0.02870
Full cubeful rollout with var.redn.
648 games, Mersenne Twister dice gen. with seed 1058468222 and quasi-random dice
Play: world class 2-ply cubeful prune [world class]
keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
Skip pruning for 1-ply moves.
Cube: 2-ply cubeful prune [world class]


Annotation
1. Cubeful 2-ply bar/21*/20* Eq.: +1.90858

0.94718 0.77958 0.24950 - 0.05282 0.00004 0.00000

2-ply cubeful prune [world class]

2. Cubeful 2-ply bar/21* 7/6 Eq.: +1.86792 ( -0.04067)

0.95432 0.77094 0.19842 - 0.04568 0.00015 0.00000

2-ply cubeful prune [world class]

Bar/21* 7/6 would have done better than if White refrained from hitting 21/20* with aces on the next roll.

Roll2-ply playRight playErrorError
6-4:bar/21*/15bar/150.1442*2/36 =0.008011
6-5:bar/20*/14bar/140.16106*2/36 =0.008947
6-1:bar/24*/18bar/19 7/60.22610*2/36 =0.012561
4-1:bar/21*/20*bar/21* 7/60.10600*2/36 =0.005888
0.035407

If the magnitude of these errors were to hold up in extended rollouts (a reasonable chance, I think), and also apply to the similar 24/23 6/5 4/3 2/1*, this play would now be in 4th place.

Next up: 4-3. Wrong: bar/21* 7/4. Right: bar/21*/18.

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.