BGonline.org Forums

probability for gammons in rollout?

Posted By: tansley
Date: Wednesday, 16 April 2008, at 4:43 p.m.

In Response To: probability for gammons in rollout? (Klaus Evers)

Hi Klaus

After D/T (and the auto redouble), the game is always played out to conclusion, and so the 30.15% gammon figure is based on fully rolled-out data.

In the case of no double (blue holds the 2 cube), many games will not be played to conclusion due to a future D/P. When this happens, the gammon % will be based on the bot's evaluation at that point - not on rolled-out games. Hence the difference.

There will be other reason's, as Daniel has pointed out, but in general the figures in a cube rollout in the D/T column will be more accurate than the figures in the ND column.

tansley

Messages In This Thread

• Cube action ...
  Mislav Radica -- Saturday, 12 April 2008, at 3:03 p.m.
  • Cube action ...
    Bill Riles -- Saturday, 12 April 2008, at 3:07 p.m.
  • Rollout convert gnubg evaluations by 0.156!
    Mislav Radica -- Sunday, 13 April 2008, at 7:31 a.m.
    • absolutely no surprise
      Klaus Evers -- Sunday, 13 April 2008, at 10:01 a.m.
      • Yes, you said that rollout for cube is must. *NM*
        Mislav Radica -- Sunday, 13 April 2008, at 11:25 a.m.
    • probability for gammons in rollout?
      Klaus Evers -- Sunday, 13 April 2008, at 10:12 a.m.
      • probability for gammons in rollout?
        Daniel Murphy -- Wednesday, 16 April 2008, at 1:50 p.m.
      • probability for gammons in rollout?
        tansley -- Wednesday, 16 April 2008, at 4:43 p.m.