[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Bug-gnubg] More cube analysis oddities
From: |
Albert Silver |
Subject: |
[Bug-gnubg] More cube analysis oddities |
Date: |
Fri, 21 Mar 2003 13:44:47 -0300 |
I downloaded the latest build of March 20th and redid the rollouts
exactly as the last time with 1296 trials, full rollout, 0-ply checker
play, and 1-ply cube decisions, Jacoby on). The results were not really
better, but Douglas Zare may have pointed out the problem. I can't post
the detailed results as none of the export functions will carry the Hint
window's contents, nor does the Copy button work. Basically it says:
Cubeless rollout equity: +0.875
Cubeful equities:
1. Double, take +0.944
57.8% 37.9% 10.2% - 42.2%
2. Double, pass +1.000 +0.56
3. No double +0.888 -0.55 (a rounding off problem I guess)
65.4% 46.7% 10.0% - 34.6%
Proper Cube action: Double, take
Obviously, it is still very strange to have a +0.875 cubeless equity be
considered a take by any margin. However, in a previous rollout with
Jacoby OFF, as a test, it yielded:
Rollout details
Win W g W bg Lose L g L bg Cubeless
Cubeful
Centered 1-cube 59.8% 40.4% 10.1% - 40.2% 0.0% 0.0% +0.701
+0.714
Standard error 0.3% 0.2% 0.2% - 0.3% 0.0% 0.0% 0.007 +0.014
Player White owns 2-cube 57.0% 37.8% 9.9% - 43.0% 0.0% 0.0% +1.236
+0.877
Standard error 0.3% 0.3% 0.2% - 0.3% 0.0% 0.0% 0.015 +0.022
This had been posted in the Gammonline BBS. Chuck Bower, a top american
player, commented:
-------------------------------------------------------
This looks odd to me. The cubeless equity is 0.70 yet it's still a take?
Am I having trouble with this because I'm used to Snowie and JF
cubeless-$ equities, and GNU reports something a bit different?
Janowski has a drop/takepoint of 0.64 (cubeless equity) here.
-------------------------------------------------------
To which Douglas Zare responded:
-------------------------------------------------------
The more meaningful cubeless number in gnu's report was the 1.236
cubeless equity on a 2-cube. That should be divided by 2 to say that the
position seems to have a cubeless equity of 0.618.
-------------------------------------------------------
- [Bug-gnubg] More cube analysis oddities,
Albert Silver <=