BGonline.org Forums

XBG 3ply RO: 53P21: Slot better as Split

Posted By: Maik Stiebler
Date: Friday, 2 October 2009, at 7:16 a.m.

In Response To: XBG 3ply RO: 53P21: Slot better as Split (W Womack)

Assuming normal distributions, you have to take the Pythagorean sum of the confidence intervals. If the sample means of the two plays are further apart than that, the p-value is higher than whatever XBG uses as the default value for p for computing confidence intervals.

Here, sqrt(0.010^2+0.011^2) < 0.446-0.429, so significance is high.

Messages In This Thread

• 53P21$@GS?
  Andreas -- Wednesday, 30 September 2009, at 5:40 p.m.
  • 53P21$@GS?
    Michael Depreli -- Wednesday, 30 September 2009, at 5:51 p.m.
    • 53P21$@GS?
      neilkaz -- Wednesday, 30 September 2009, at 6:27 p.m.
      • 53P21$@GS?
        Andreas -- Wednesday, 30 September 2009, at 6:51 p.m.
        • No - the problem isn't the settings
          David Rockwell -- Wednesday, 30 September 2009, at 7:45 p.m.
  • corrected settings: much closer...
    Andreas -- Thursday, 1 October 2009, at 5:38 a.m.
    • corrected settings: much closer...
      Stick -- Thursday, 1 October 2009, at 1:09 p.m.
      • XBG 3ply RO: 53P21: Slot better as Split
        Andreas -- Thursday, 1 October 2009, at 6:06 p.m.
        • XBG 3ply RO: 53P21: Slot better as Split
          W Womack -- Friday, 2 October 2009, at 1:17 a.m.
          • XBG 3ply RO: 53P21: Slot better as Split
            Maik Stiebler -- Friday, 2 October 2009, at 7:16 a.m.
          • XBG 3ply RO: 53P21: Slot better as Split
            Andreas -- Saturday, 3 October 2009, at 12:39 a.m.
        • XBG 3ply RO: 53P21: Slot better as Split
          neilkaz -- Friday, 2 October 2009, at 1:55 a.m.
          • XBG 3ply RO: 53P21: Slot better as Split
            Chuck Bower -- Friday, 2 October 2009, at 4:33 p.m.
    • corrected settings: much closer...
      David Rockwell -- Thursday, 1 October 2009, at 3:12 p.m.
  • Link to article please ! Can't find. *NM*
    neilkaz -- Thursday, 1 October 2009, at 4:51 p.m.
    • Link
      David Rockwell -- Thursday, 1 October 2009, at 5:46 p.m.
    • Link to article please ! Can't find.
      Andreas -- Thursday, 1 October 2009, at 5:48 p.m.