| |
BGonline.org Forums
31P-21S-65 — Rollouts — Same bot, same seed, same count, different results
Posted By: Nack Ballard In Response To: 31P-21S-65 — Rollouts — Same bot, same seed, same count, different results (Stick)
Date: Thursday, 2 February 2012, at 1:14 p.m.
Because I may not have thought anyone (you) was meticulously recording results and put the 15552 in myself not caring that the results might be off by a fraction of this once the rollout actually finished. (or not realizing 5 years later it would be referenced!)
I actually had considered the number of trials being manually overwritten as a possible explanation but dismissed it because I couldn't think of a good reason you would do that. If the file is, say, 13197 trials and you don't want to finish it or wait to finish it, why not just post it as the actual 13197 (rather than overwrite to 15552)?
Not that it matters much. The only important spoiler would be if someone is attempting to duplicate a result with the same seed, though at the time you manually overwrote (if you did in this case), presumably we hadn't yet discovered that some Gnu versions were corrupted. Hopefully, most if not all corruption-check tests were compared against hard file results.
(Another possible problem with overwriting, if done routinely, is that to the results-compiler variance will appear to be a bit higher than it really is. :)
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.