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

BGonline.org Forums

Short rollouts of variants (partial spoilers) and RO MAIN POS

Posted By: Snua
Date: Monday, 8 September 2014, at 6:11 p.m.

In Response To: Short rollouts of variants (partial spoilers) (Timothy Chow)

These are very interesting. I agree that the results are difficult to explain. At least I cannot explain them. But I'll try anyway.

In the main position, (MP), it's strange, because even though he's ahead in the race, we're the ones who’re breaking first. This is misleading and bizarre, because pip-wise he really needs to come of the 22 before we off the 21.

I noticed that some 4's play from the 22 for him. At least 45 will I think. And even though others might not (other than 44), this is thematically what he wants to do with them, so in an abstract sense, you don't mind giving him alternative good 4's. Considering other future rolls etc.

Variant 1: Seems to want to keep the 15 point when timing is worse. Could this be just because of his subsequent bad 33? Could it be that the explanation is this simple??

Variant 2: I think I sort of actually understand this one. Here's my theory anyway: our position is no way near as good as MP - see below. In MP we're in the region of 57% as opposed to about 46% here, so there may be some concept at work to do with the greater importance of preserving winning chances when those winning chances are greater. Could have something to do with being closer to cubing. Not sure. So when he's on the 21 instead of the 22, we actually are, contrary to intuition, happier not to play so flexibly and just play something down from the mid, in other words, it doesn't matter so much as far as actually winning the game goes.

Variant 3: This seems to chime with explanation in V2 above i.e. not so bad to break the 16 and sacrifice safety for flexibility where our winning chances are relatively higher and where we might actually win the game.

Variant 4: My higher winning chances theory would argue in favour of preserving 16 pt play since here the winning chances are much lower, then this would be further mitigated by the need not to get hit when the opponent has his 4, 5 and 6, instead of 3, 5 and 6. However running and coming down are about equal. Maybe what kicks in now is that because we're in so much worse shape, it now becomes relatively more important not to get hit. So the reverse effect is in operation.

Main position rollout:





White is Player 2

score: 0
pip: 133
Unlimited Game
Jacoby Beaver
pip: 146
score: 0

Blue is Player 1
XGID=---bBCB-C---bA--BbadbBb---:0:0:1:21:0:0:3:0:10
Blue to play 21

1.Rollout116/13eq: +0.277
Player:
Opponent:		57.00% (G:13.72% B:0.44%)
43.00% (G:10.29% B:0.34%)		Conf.: ± 0.013 (+0.264...+0.291) - [100.0%]
Duration: 22 minutes 05 seconds
2.Rollout113/10eq: +0.203 (-0.074)
Player:
Opponent:		54.55% (G:14.50% B:0.57%)
45.45% (G:10.40% B:0.35%)		Conf.: ± 0.012 (+0.192...+0.215) - [0.0%]
Duration: 22 minutes 07 seconds
3.3-ply13/11 5/4eq: +0.160 (-0.118)
Player:
Opponent:		54.69% (G:12.55% B:0.41%)
45.31% (G:9.92% B:0.32%)
4.3-ply8/6 5/4eq: +0.094 (-0.183)
Player:
Opponent:		52.89% (G:12.37% B:0.37%)
47.11% (G:11.20% B:0.37%)
5.3-ply13/11 8/7eq: +0.070 (-0.207)
Player:
Opponent:		53.04% (G:12.35% B:0.44%)
46.96% (G:13.07% B:0.49%)
1 1296 Games rolled with Variance Reduction.
Dice Seed: 91819330
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.10

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.