 SBR Top-Rated Sportsbooks |
Best Sportsbooks List
|
|
| |||||
| #1 FanDuel | SBR rating 4.8/5 | Review | #6 BetRivers | SBR rating 4.1/5 | Review | #2 Caesars | SBR rating 4.7/5 | Review | #7 Fanatics | SBR rating 4.1/5 | Review |
| #3 DraftKings | SBR rating 4.7/5 | Review | #8 Betway | SBR rating 3.8/5 | Review | ||
| #4 BetMGM | SBR rating 4.6/5 | Review | #9 Borgata | SBR rating 3.5/5 | Review | ||
| #5 bet365 | SBR rating 4.6/5 | Review | #10 ClutchBet | SBR rating 2.9/5 | Review | ||
non linear statistical modeling in NCAAF
In most models I've worked with, there were certain linear properties of team strengths.
If A and B would be "Pick'em" on a neutral field, then you would expect the spread of a team against A or B to be identical. I'm trying a new approach where that no longer holds true.
Has anyone here experimented with more complex statistical power rankings where you see different spreads against quasi-equal teams?
Team C could be better suited to play in park A than in park B.
If you are more comfortable with linear models (as most are), then you may want to consider a functional transformation of the data.Originally Posted by Justin7
I myself am not too familiar with non-linear model.
Good luck Justin....
My models are linear regression-based, but I include interaction effects where appropriate. (i.e. a team with a high K rate vs. a high K pitcher). I think the baseline though is power rankings for specific factors, rather than an overall power ranking.
yep, linear model can pretty much do everything after you transform the data...why would you want to use a nonlinear model at all?
You may simply want to add a factor called 'over enthusiasm'.
In football you have teams that will play better against stronger teams and weaker against weak opponents (negative OE factor) or the other way round teams that like to crush weak opponents and give up against strong opponents (positive OE factor).
Good Luck, Justin7
Reply With Quote