When I consider adding more statistics to improve the model:
+ Week 6 Plays --
Maryland (-4) - Wake Forest
Maryland (-4) - Wake Forest
Louisiana-Monroe (-3) - Middle Tennessee St
Kentucky (+10.5) - Mississippi St
Hawai`i (+22.5) - San Diego St
Utah (+13.5) - Southern Cal
Western Michigan (-16) - Massachusetts
Stanford (-9.5) - Arizona
Eastern Michigan (+3.5) - Kent St
Connecticut (+9.5) - Rutgers
Syracuse (+3) - Pittsburgh
Texas (-7) - West Virginia
Miami OH (+21) - Cincinnati
Last Week: 9-10
Season: 47-30-1 (61.0%)
//
+ Gut Plays --
Kansas St (-24) – Kansas. 59-7 in 2010 and 59-21 in 2011.
Run on Monday afternoon. Posted Tuesday.
Kansas St (-24) – Kansas. 59-7 in 2010 and 59-21 in 2011.
Run on Monday afternoon. Posted Tuesday.
3 comments :
If you watch Stan against Wash and they look like a DIII program do you lay off when the model calls for it against Zona?
The model finds value through statistics that I've found contribute to success (through running countless regressions).
In this case Stanford lost to a ranked Washington program in Seattle, on national TV, and at night. I wouldn't penalize them for that too much. But your point is valid that the model doesn't account for emotion or the "eye test".
I'll also bet games that the model doesn't recommend, like Kansas State today.
Yea that was more of just an example, you hit the underlying question.
I should have used San Jose St game as an example instead. I was on SJ St. but Stan made it even easier than expected. Side note- Wash is ranked because of said Stanford game, they'll be unranked soon. I agree entirely on not putting a lot of weight on home team taking a thurs night game, but to your point it was more of an eye test.
Post a Comment