BGonline.org Forums

Othello Quiz 2015, Problem 1 (Variants, etc.)

Posted By: Nack Ballard
Date: Sunday, 25 October 2015, at 10:24 p.m.

In Response To: Othello Quiz 2015 (othello)

		White is Player 2 score: 4 pip: 73 7 point match pip: 140 score: 0 Blue is Player 1
XGID=-a---BE-BAB----A--bbBbbcc-:1:1:1:00:0:4:0:7:10
Cube action of Original Position

Analyzed in Rollout	No redouble	Redouble/Take
Player Winning Chances:	51.51% (G:5.65% B:0.24%)	51.66% (G:6.16% B:0.28%)
Opponent Winning Chances:	48.49% (G:18.87% B:0.18%)	48.34% (G:22.02% B:0.13%)
Cubeless Equities	-0.036	+0.515
Cubeful Equities
No redouble:	+0.513 (-0.002)	กำ0.003 (+0.510..+0.515)
Redouble/Take:	+0.515	กำ0.003 (+0.512..+0.517)
Redouble/Pass:	+1.000 (+0.485)
Best Cube action: Redouble / Take
Rollout details
20736 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:		84.9%
Take Decision confidence:		100.0%
Duration: 11 hours 10 minutes

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

		White is Player 2 score: 4 pip: 73 7 point match pip: 140 score: 0 Blue is Player 1
XGID=-a---BE-BAB----A--bbBbbcc-:1:1:1:65:0:4:0:7:10
Original Position (Problem 1)

1.	XG Roller++	15/9 6/1*	eq: +0.529
	Player: Opponent: 51.00% (G:6.82% B:0.19%) 49.00% (G:20.22% B:0.10%)
2.	XG Roller++	10/4 9/4	eq: +0.374 (-0.155)
	Player: Opponent: 48.14% (G:4.66% B:0.12%) 51.86% (G:15.29% B:0.13%)
3.	XG Roller++	20/9	eq: +0.309 (-0.220)
	Player: Opponent: 47.18% (G:4.00% B:0.11%) 52.82% (G:17.59% B:0.09%)
4.	XG Roller++	20/15 20/14	eq: +0.230 (-0.300)
	Player: Opponent: 45.64% (G:2.50% B:0.03%) 54.36% (G:17.29% B:0.10%)
5.	XG Roller++	9/3 8/3	eq: +0.168 (-0.362)
	Player: Opponent: 42.34% (G:4.19% B:0.11%) 57.66% (G:18.98% B:0.14%)
6.	XG Roller++	15/9 10/5	eq: +0.156 (-0.373)
	Player: Opponent: 41.77% (G:3.95% B:0.11%) 58.23% (G:15.73% B:0.13%)

 

Kudos to Othello-san for another superb quiz.

In this post, I discuss Problem 1 from the recent 2015 quiz.

In the first (left-hand) position above, the 20k rollout suggests it is likely -- though by no means certain -- that Blue "should" have redoubled before rolling. I trust nobody will lose sleep over this, but some may find it interesting.

In the second (right-hand) position above, the XGR++ evaluation supports N (Near, 15/9 6/1*) as the best play, and it is a "wipeout" (the Ballard/Weaver error scale term for a play that is better than the second best play by at least .100). Othello's web-posted rollout confirms the same order of plays, with the error sizes being .177, .244, .326, .371 and .396. (I would have posted it if I had the file.)

In the first (left-hand) pre-variant below, I've altered White's buried-spare position by 1 pip. This time, it is almost certainly proper for Blue to roll without redoubling.

The second (right-hand) variant below is a rollout of the 65 with that minor spare alteration. The margins are noticeably tighter (compare .128, .192, .266, .333 and .353 to the green numbers above), but the N play can still easily boast the status of "wipeout."

My nitpicky point is that with a slight spare alteration (2pt to 3pt), it is possible to eliminate the possible cube controversy, albeit with some sacrifice in best-play clearance margin.

		White is Player 2 score: 4 pip: 74 7 point match pip: 140 score: 0 Blue is Player 1
XGID=-a---BE-BAB----A--bbBbcbc-:1:1:1:00:0:4:0:7:10
Cube action of Variant 1

Analyzed in Rollout	No redouble	Redouble/Take
Player Winning Chances:	50.53% (G:5.34% B:0.23%)	50.74% (G:5.67% B:0.25%)
Opponent Winning Chances:	49.47% (G:19.44% B:0.22%)	49.26% (G:22.62% B:0.15%)
Cubeless Equities	-0.063	+0.460
Cubeful Equities
No redouble:	+0.473	กำ0.008 (+0.466..+0.481)
Redouble/Take:	+0.460 (-0.014)	กำ0.008 (+0.452..+0.468)
Redouble/Pass:	+1.000 (+0.527)
Best Cube action: No redouble / Take
Percentage of wrong pass needed to make the double decision right: 2.5%
Rollout details
5184 Games rolled with Variance Reduction. Dice Seed: 31027603 Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:		99.2%
Take Decision confidence:		100.0%
Duration: 3 hours 57 minutes

		White is Player 2 score: 4 pip: 74 7 point match pip: 140 score: 0 Blue is Player 1
XGID=-a---BE-BAB----A--bbBbcbc-:1:1:1:65:0:4:0:7:10
Variant 1 (White 23pt spare to 22pt)

1.	Rollout1	15/9 6/1*	eq: +0.509
	Player: Opponent: 49.63% (G:7.44% B:0.34%) 50.37% (G:24.43% B:0.21%) Conf.: กำ 0.011 (+0.498...+0.520) - [100.0%] Duration: 31 minutes 50 seconds
2.	Rollout1	10/4 9/4	eq: +0.381 (-0.128)
	Player: Opponent: 47.62% (G:5.43% B:0.20%) 52.38% (G:18.71% B:0.25%) Conf.: กำ 0.009 (+0.372...+0.390) - [0.0%] Duration: 28 minutes 48 seconds
3.	Rollout1	20/9	eq: +0.316 (-0.192)
	Player: Opponent: 46.71% (G:4.59% B:0.16%) 53.29% (G:19.38% B:0.10%) Conf.: กำ 0.009 (+0.308...+0.325) - [0.0%] Duration: 30 minutes 30 seconds
4.	Rollout1	20/15 20/14	eq: +0.243 (-0.266)
	Player: Opponent: 45.54% (G:2.90% B:0.09%) 54.46% (G:18.53% B:0.16%) Conf.: กำ 0.008 (+0.235...+0.251) - [0.0%] Duration: 38 minutes 48 seconds
5.	Rollout1	9/3 8/3	eq: +0.176 (-0.333)
	Player: Opponent: 41.88% (G:4.65% B:0.15%) 58.12% (G:21.22% B:0.21%) Conf.: กำ 0.008 (+0.168...+0.184) - [0.0%] Duration: 31 minutes 54 seconds
6.	Rollout1	15/9 10/5	eq: +0.156 (-0.353)
	Player: Opponent: 41.04% (G:4.42% B:0.17%) 58.96% (G:18.50% B:0.20%) Conf.: กำ 0.010 (+0.146...+0.165) - [0.0%] Duration: 34 minutes 20 seconds
1 1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

 

Next, I respectfully submit Variant 2 below (with one of Blue's 6pt spares moved back to his 8pt). I find this problem even more difficult, because (a) Blue has the option of making an inside point without leaving a blot in his outer board, and (b) There is one fewer spare on the 6pt, making the ace-point hit less attractive.

I changed the score to gammon go for two reasons. One, there can be no criticism of cube action. Two, it increases the margin cleared by the best play from .055 (rollout at original match score situation, not shown) to the .070 you see below. [I noticed that of the ten 2015 problems, the smallest such margin is a blowout .068 -- Othello seems to hold these margins to a high standard.]

		White is Player 2 score: 6 pip: 73 7 point match Crawford pip: 142 score: 5 Blue is Player 1
XGID=-a---BD-CAB----A--bbBbbcc-:0:0:1:65:5:6:1:7:10
Variant 2 (Blue 6pt spare to 8pt)

1.	Rollout1	15/9 6/1*	eq: +0.290
	Player: Opponent: 54.99% (G:9.53% B:0.53%) 45.01% (G:25.77% B:0.11%) Conf.: กำ 0.007 (+0.283...+0.297) - [100.0%] Duration: 11 minutes 04 seconds
2.	Rollout1	9/3 8/3	eq: +0.220 (-0.070)
	Player: Opponent: 53.88% (G:7.13% B:0.39%) 46.12% (G:17.62% B:0.12%) Conf.: กำ 0.005 (+0.215...+0.225) - [0.0%] Duration: 9 minutes 38 seconds
3.	Rollout1	20/9	eq: +0.192 (-0.099)
	Player: Opponent: 53.39% (G:6.19% B:0.27%) 46.61% (G:19.26% B:0.13%) Conf.: กำ 0.006 (+0.186...+0.197) - [0.0%] Duration: 9 minutes 22 seconds
4.	Rollout1	10/4 9/4	eq: +0.177 (-0.113)
	Player: Opponent: 52.44% (G:6.41% B:0.31%) 47.56% (G:20.91% B:0.11%) Conf.: กำ 0.005 (+0.172...+0.182) - [0.0%] Duration: 10 minutes 03 seconds
5.	Rollout1	20/15 20/14	eq: +0.128 (-0.162)
	Player: Opponent: 52.46% (G:3.95% B:0.15%) 47.54% (G:18.69% B:0.11%) Conf.: กำ 0.005 (+0.123...+0.133) - [0.0%] Duration: 8 minutes 43 seconds
6.	Rollout1	15/9 10/5	eq: +0.063 (-0.228)
	Player: Opponent: 47.17% (G:5.96% B:0.34%) 52.83% (G:22.92% B:0.05%) Conf.: กำ 0.005 (+0.057...+0.068) - [0.0%] Duration: 9 minutes 41 seconds
1 1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

I do not claim my variants to be superior to the original position; I'm not sure how Othello balances the criteria. However, I am interested in his opinion. As it's his quiz, I naturally defer to his judgment.

Finally, below I offer a pair of variants that contrast each other. Variant 3 (left-hand) has the same match score, cube placement and board position as Othello's original problem, except Blue's 5pt is moved back to his 11pt. Variant 4 (right-hand) has that point nudged back slightly further, to the 12pt.

If there is sufficient interest, I'll try to find time to post variants of other Othello-quiz positions.

Nack

		White is Player 2 score: 4 pip: 73 7 point match pip: 152 score: 0 Blue is Player 1
XGID=-a----E-BABB---A--bbBbbcc-:1:1:1:65:0:4:0:7:10
Variant 3 (11pt)

1.	Rollout1	15/9 6/1*	eq: +0.511
	Player: Opponent: 49.98% (G:7.31% B:0.50%) 50.02% (G:29.14% B:0.45%) Conf.: กำ 0.023 (+0.488...+0.534) - [100.0%] Duration: 42 minutes 33 seconds
2.	Rollout1	20/9	eq: +0.361 (-0.150)
	Player: Opponent: 47.26% (G:5.82% B:0.27%) 52.74% (G:27.67% B:0.14%) Conf.: กำ 0.021 (+0.340...+0.382) - [0.0%] Duration: 35 minutes 00 second
3.	Rollout1	20/15 20/14	eq: +0.157 (-0.355)
	Player: Opponent: 43.51% (G:3.36% B:0.12%) 56.49% (G:29.71% B:0.30%) Conf.: กำ 0.014 (+0.143...+0.170) - [0.0%] Duration: 48 minutes 20 seconds
4.	Rollout1	10/4 9/4	eq: +0.105 (-0.406)
	Player: Opponent: 40.16% (G:4.93% B:0.23%) 59.84% (G:28.81% B:0.37%) Conf.: กำ 0.010 (+0.095...+0.115) - [0.0%] Duration: 45 minutes 30 seconds
5.	Rollout1	15/9 11/6	eq: +0.076 (-0.435)
	Player: Opponent: 39.26% (G:4.48% B:0.19%) 60.74% (G:24.76% B:0.39%) Conf.: กำ 0.011 (+0.065...+0.087) - [0.0%] Duration: 44 minutes 45 seconds
6.	Rollout1	11/5 10/5	eq: +0.041 (-0.470)
	Player: Opponent: 39.28% (G:4.85% B:0.22%) 60.72% (G:35.93% B:0.74%) Conf.: กำ 0.010 (+0.032...+0.051) - [0.0%] Duration: 43 minutes 43 seconds
7.	Rollout1	9/3 8/3	eq: -0.031 (-0.542)
	Player: Opponent: 36.15% (G:4.92% B:0.22%) 63.85% (G:31.04% B:0.39%) Conf.: กำ 0.010 (-0.041...-0.021) - [0.0%] Duration: 43 minutes 16 seconds
1 1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

		White is Player 2 score: 4 pip: 73 7 point match pip: 154 score: 0 Blue is Player 1
XGID=-a----E-BAB-B--A--bbBbbcc-:1:1:1:65:0:4:0:7:10
Variant 4 (12pt)

1.	Rollout1	20/9	eq: +0.236
	Player: Opponent: 44.14% (G:5.60% B:0.34%) 55.86% (G:30.96% B:0.11%) Conf.: กำ 0.014 (+0.223...+0.250) - [100.0%] Duration: 39 minutes 13 seconds
2.	Rollout1	15/9 6/1*	eq: +0.189 (-0.047)
	Player: Opponent: 42.04% (G:6.01% B:0.35%) 57.96% (G:32.02% B:0.28%) Conf.: กำ 0.016 (+0.173...+0.206) - [0.0%] Duration: 53 minutes 32 seconds
3.	Rollout1	20/15 20/14	eq: +0.039 (-0.197)
	Player: Opponent: 40.42% (G:3.28% B:0.09%) 59.58% (G:31.97% B:0.15%) Conf.: กำ 0.011 (+0.028...+0.050) - [0.0%] Duration: 52 minutes 17 seconds
4.	Rollout1	10/4 9/4	eq: -0.004 (-0.240)
	Player: Opponent: 37.50% (G:4.75% B:0.18%) 62.50% (G:31.13% B:0.56%) Conf.: กำ 0.010 (-0.014...+0.006) - [0.0%] Duration: 46 minutes 44 seconds
5.	Rollout1	15/9 12/7	eq: -0.057 (-0.293)
	Player: Opponent: 37.37% (G:5.01% B:0.21%) 62.63% (G:45.37% B:0.50%) Conf.: กำ 0.012 (-0.069...-0.045) - [0.0%] Duration: 41 minutes 40 seconds
6.	Rollout1	9/3 8/3	eq: -0.073 (-0.309)
	Player: Opponent: 35.22% (G:4.89% B:0.12%) 64.78% (G:33.53% B:0.40%) Conf.: กำ 0.011 (-0.083...-0.062) - [0.0%] Duration: 43 minutes 09 seconds
1 1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

 

 

Messages In This Thread

• Othello Quiz 2015
  othello -- Wednesday, 14 October 2015, at 6:24 p.m.
  • Othello Quiz 2015
    Stick -- Saturday, 17 October 2015, at 2:08 a.m.
    • Othello Quiz 2015
      Snua -- Saturday, 17 October 2015, at 11:35 a.m.
      • Othello Quiz 2015
        Stick -- Saturday, 17 October 2015, at 3:35 p.m.
        • Othello Quiz 2015
          Snua -- Sunday, 18 October 2015, at 8:45 a.m.
          • Othello Quiz 2015
            Stick -- Sunday, 18 October 2015, at 9:00 a.m.
  • Rollouts and result
    othello -- Sunday, 18 October 2015, at 9:45 a.m.
  • Othello Quiz 2015
    Timothy Chow -- Saturday, 24 October 2015, at 8:36 p.m.
    • After looking at the rollouts...
      Timothy Chow -- Saturday, 24 October 2015, at 9:01 p.m.
  • Othello Quiz 2015, Problem 1 (Variants, etc.)
    Nack Ballard -- Sunday, 25 October 2015, at 10:24 p.m.
    • Othello Quiz 2015, Problem 1 (Variants, etc.)
      othello -- Thursday, 29 October 2015, at 6:39 a.m.
  • Othello Quiz 2015 variants
    Timothy Chow -- Monday, 26 October 2015, at 12:59 a.m.