BGonline.org Forums

side note: xG MET

Posted By: Robert-Jan Veldhuizen (Zorba)
Date: Sunday, 28 June 2009, at 12:00 p.m.

In Response To: eXtreme Gammon or GnuBG - cube? (Robert-Jan Veldhuizen (Zorba))

I guess one of the reasons GnuBG and xG show a relatively large disagreement on these two cube actions I posted, is the MET used.

I don't know which MET xG used to make its decisions, but it's probably the default xG MET. A quick peek at interesting entries:

away score: xG MET% , g11 MET%, Dunstan MET%

-2,-1PC: 48.82 48.81 48.89

-2,-1 C: 31.12 31.85 32.23

-3,-2: 40.05 40.04 40.09
-4,-2: 32.46 32.82 33.12
-5,-2: 25.43 25.48 25.62

-4,-3: 42.24 42.63 42.82
-5,-3: 34.84 34.97 35.12

-5,-4: 42.27 42.25 42.18

The pattern is pretty clear, the METs are pretty similar except at gammon-go(-like) scores: there the xG MET shows clearly lower winning chances than g11 (0-ply play, 2-ply cube) and even more so compared to the Dunstan MET (2-ply play and cube).

I don't know how the xG MET was generated, would be interesting. For now I'm a bit suspicious about it, since the pattern of the past has always been that stronger play leads to higher gammon rates (the difference between g11 and Dunstan also shows this nicely).

Theoretically, it's of course possible that better (or equal but different) play, would lead to a better defense against getting gammoned.

Fortunately, xG allows you to load different METs and has several included that also come with GnuBG.

Messages In This Thread

• eXtreme Gammon or GnuBG - cube?
  Robert-Jan Veldhuizen (Zorba) -- Saturday, 27 June 2009, at 12:07 p.m.
  • eXtreme Gammon or GnuBG - cube?
    John O'Hagan -- Saturday, 27 June 2009, at 2:01 p.m.
  • eXtreme Gammon or GnuBG - cube?
    Neil Robins -- Saturday, 27 June 2009, at 2:21 p.m.
  • side note: xG MET
    Robert-Jan Veldhuizen (Zorba) -- Sunday, 28 June 2009, at 12:00 p.m.