BGonline.org Forums

21 RO at dmp

Posted By: Nack Ballard
Date: Saturday, 5 February 2011, at 2:34 a.m.

In Response To: 21 RO at dmp (Ian Shaw)

2 * 36^3 is a lot of trials. But what if one of the 2nd roll responses is being played incorrectly? Wouldn't we gain more information by rolling out 21$ and 21D 23^2 times and combining the results.

Do you mean 21$ and 21S? ("21D" means 13/10.)

Why 23^2 = 529 trials?

For purposes of discussion, suppose 21$-11 is misplayed. Why would we gain more information by rolling out two single plays individually and then combining the data? (We'll ignore for the moment that the arithmetic would be a hassle.)

Nack

Messages In This Thread

• 21 RO at dmp
  Alexandre Le Siege -- Friday, 4 February 2011, at 9:16 p.m.
  • 21 RO at dmp
    Ian Shaw -- Friday, 4 February 2011, at 11:10 p.m.
    • 21 RO at dmp
      Nack Ballard -- Saturday, 5 February 2011, at 2:34 a.m.
      • 21 RO at dmp
        Ian Shaw -- Sunday, 6 February 2011, at 10:50 p.m.
        • 21 RO at dmp
          Nack Ballard -- Monday, 7 February 2011, at 7:27 a.m.
          • 21 RO at dmp
            Daniel Murphy -- Tuesday, 8 February 2011, at 9:39 p.m.
            • 21 RO at dmp
              Nack Ballard -- Wednesday, 9 February 2011, at 3:00 a.m.
              • 21 RO at dmp
                Daniel Murphy -- Wednesday, 9 February 2011, at 3:27 a.m.
                • Layered rollouts
                  Nack Ballard -- Wednesday, 9 February 2011, at 5:27 a.m.
  • 21 RO at dmp
    Nack Ballard -- Saturday, 5 February 2011, at 1:57 a.m.