| |
BGonline.org Forums
42P-43: more data
Posted By: Timothy Chow In Response To: 42P-43: more data (Nack Ballard)
Date: Sunday, 14 February 2010, at 5:49 p.m.
Nack wrote:
it happens often enough that I (and Paul, and Stick) stopped believing the confidence intervals that the bots report long ago.
I might have said this before, but based on general principles, I would expect this problem (reported confidence intervals being smaller than the truth) to be particularly bad if the cube has a chance (even a relatively small chance) of getting jacked up high. That would tend to increase the error involved in approximating by a normal distribution. I would expect less and less of a problem as the match length gets shorter.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.