BGonline.org Forums

Learning how to do proper RO with our bots.

Posted By: neilkaz
Date: Saturday, 25 September 2010, at 4:16 a.m.

In Response To: 63: Hit or not; efficiency - Rollout result (storm)

Here's a GNU RO with a setting that properly recubes after hit/dance but won't have it's head too far up its ass later on.

The score (after 0 games) is: gnubg 0, Neil 1 (match to 7 points)

Move number 9: Neil to play 36

gnubg	98
Neil	115
Position ID: a9sEAyC2c8DBAA Match ID: UYn5AAAACAAA

#		Ply	Move	Equity
	1	R	13/7 4/1*	+0.0465
			0.5025 0.1823 0.0129 - 0.4975 0.2832 0.0056 -0.1661 +0.0465 0.0015 0.0023 0.0022 - 0.0015 0.0023 0.0012 0.0046 0.0078
			Full cubeful rollout with var.redn.
			1296 games, Mersenne Twister dice gen. with seed 901728229 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 3-ply cubeful prune [grandmaster]
			Different evaluations after 5 plies:
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	2	R	13/4	-0.0244 ( -0.0709)
			0.4487 0.0530 0.0021 - 0.5513 0.1229 0.0018 -0.1855 -0.0244 0.0012 0.0010 0.0006 - 0.0012 0.0022 0.0002 0.0029 0.0043
			Full cubeful rollout with var.redn.
			1296 games, Mersenne Twister dice gen. with seed 901728229 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 3-ply cubeful prune [grandmaster]
			Different evaluations after 5 plies:
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	3	2	13/7 6/3	-0.1865 ( -0.2330)
			0.4281 0.0803 0.0011 - 0.5719 0.2264 0.0048
	4	0	18/9	-0.5987 ( -0.6452)
			0.3223 0.0884 0.0041 - 0.6777 0.3926 0.0096
	5	0	13/10 13/7	-0.7751 ( -0.8217)
			0.3131 0.0971 0.0017 - 0.6869 0.4927 0.0597

Messages In This Thread

• 63: Hit or not; efficiency.
  storm -- Thursday, 23 September 2010, at 2:13 p.m.
  • 63: Hit or not; efficiency.
    Timothy Chow -- Thursday, 23 September 2010, at 11:38 p.m.
  • 63: Hit or not; efficiency.
    Chuck Bower -- Thursday, 23 September 2010, at 11:49 p.m.
  • 63: Hit or not; efficiency.
    neilkaz -- Friday, 24 September 2010, at 4:47 p.m.
  • 63: Hit or not; efficiency - Rollout result
    storm -- Friday, 24 September 2010, at 5:07 p.m.
    • Corrupted rollout?
      Timothy Chow -- Friday, 24 September 2010, at 6:53 p.m.
      • Corrupted rollout?
        storm -- Friday, 24 September 2010, at 7:38 p.m.
      • Corrupted rollout?
        Stick -- Friday, 24 September 2010, at 8:15 p.m.
      • Corrupted rollout?
        Neil Robins -- Friday, 24 September 2010, at 8:33 p.m.
        • Corrupted rollout?
          Philippe Michel -- Friday, 24 September 2010, at 9:30 p.m.
    • Learning how to do proper RO with our bots.
      neilkaz -- Saturday, 25 September 2010, at 4:16 a.m.
      • Learning how to do proper RO with our bots.
        storm -- Saturday, 25 September 2010, at 6:56 a.m.
        • Learning how to do proper RO with our bots.
          neilkaz -- Saturday, 25 September 2010, at 2:40 p.m.
          • Learning how to do proper RO with our bots.
            Daniel Murphy -- Saturday, 25 September 2010, at 4:08 p.m.