BGonline.org Forums

31P-21S-65 — Rollouts — Same bot, same seed, same count, different results

Posted By: Taper_Mike
Date: Thursday, 2 February 2012, at 10:14 a.m.

In Response To: 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results (Stick)

I wondered whether a soft-error in RAM chip could have flipped a couple of bits in the value, so I checked the underlying bit patterns for the values 4.0 and 5.4. They are not close at all. (Note that 5.4 is a repeating binary, analogous to 1/3 in base 10.)

4.0 = 0100 0000 0001 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
5.4 = 0100 0000 0001 0101 1001 1001 1001 1001 1001 1001 1001 1001 1001 1001 1001 1010

Stick, you said you “might” have doctored the output. It’s a plausible idea only in that it would account for the evidence. But I find the notion unlikely. Do you have a positive recollection that you have done that to rollouts in the past?!! Why waste the time? I’m still prepared to chalk it up to some bizarre software issue as yet unexplained. And then forget it! The XG rollout I posted is close to the BGO value of 5.4. The 20k trial count may be insufficient, however, to wring out the variance of plays that are tied under the Ballard-Weaver error scale. Nack, didn’t you post somewhere that you and David are now rolling 61k for XG 3-ply rollouts where the results are within 0.01? Mike P.S. I have apologized elsewhere for noticing these nit-picking inconsistencies. What can I say? When I post the data to my rollout database, the rollouts are suddenly listed side-by-side, and the problem is hard not notice.

Messages In This Thread

• 31P.21S.65?
  Stick -- Tuesday, 30 October 2007, at 3:42 a.m.
  • 31P.21S.65?
    neilkaz -- Tuesday, 30 October 2007, at 4:18 p.m.
    • 31P.21S.65?
      Matt Cohn-Geier -- Wednesday, 31 October 2007, at 3:01 a.m.
      • 31P.21S.65?
        Stick -- Wednesday, 31 October 2007, at 3:24 a.m.
  • 31P.21S.65?
    Caryl Stephanson -- Tuesday, 30 October 2007, at 8:57 p.m.
  • Rollout
    Stick -- Thursday, 1 November 2007, at 8:17 a.m.
    • You missed a candidate
      Chuck Bower -- Thursday, 1 November 2007, at 7:42 p.m.
      • You missed a candidate
        Stick -- Thursday, 1 November 2007, at 8:11 p.m.
      • You missed a candidate
        Marty Storer -- Wednesday, 28 November 2007, at 6:42 p.m.
    • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
      Taper_Mike -- Wednesday, 1 February 2012, at 4:28 a.m.
      • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
        Nack Ballard -- Wednesday, 1 February 2012, at 6:26 a.m.
      • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
        Stick -- Wednesday, 1 February 2012, at 6:59 a.m.
        • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
          Nack Ballard -- Wednesday, 1 February 2012, at 2:17 p.m.
          • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
            Stick -- Thursday, 2 February 2012, at 8:12 a.m.
            • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
              Taper_Mike -- Thursday, 2 February 2012, at 10:14 a.m.
              • Degree of extension
                Nack Ballard -- Thursday, 2 February 2012, at 1:53 p.m.
            • 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
              Nack Ballard -- Thursday, 2 February 2012, at 1:14 p.m.
      • 31P-21S-65 — New XG Rollout
        Taper_Mike -- Thursday, 2 February 2012, at 1:52 a.m.