Here's how the thousand Kenpom simulations did against the thousand Hanner simulations in Round One:
# Right | Kenpom | Hanner | ||
27 | 1 | 0 | ||
26 | 8 | 2 | ||
25 | 23 | 7 | ||
24 | 39 | 21 | ||
23 | 81 | 55 | ||
22 | 130 | 97 | ||
21 | 149 | 133 | ||
20 | 182 | 178 | ||
19 | 148 | 162 | ||
18 | 121 | 139 | ||
17 | 69 | 95 | ||
Average: | 20.034 | 19.367 |
I've consulted my wife, an epidemiologist, and the preliminary opinion is that these results are not "statistically significant" (whatever that means). They are, however, Torvistically significant—which in some (very small) circles is considered much more important.
No comments:
Post a Comment