Although my model’s spread projections haven’t done well the past couple of years (and continues to do poorly at 41.2% ATS), it still outperforms both Number Fire (38.9% ATS) and Team Rankings (33.3% ATS) – standards by which I measure my model’s relative reliability; moreover, its straight-up performance (68.7% SU) does show that, at the very least, it knows good teams from bad. What’s killing everyone’s spread projections (including Vegas line-setters) this year, is the impact of home field advantage or, rather, the lack thereof on actual margins of victory (MOV) – which has made road dogs Vegas heroes at 56.3% ATS. In these days of minimal crowds due to Covid 19, last minute schedule changes, and the relatively enormous number of injuries plaguing the NFL because of the lack of off-season training and no pre-season, past performance statistics are far less reliable indicators of prospective performance than they have ever been; and, trying disabuse algorithms of NFL normality is a lot easier said than done. Nevertheless, I’m confident that as the season progresses, the reliability of the numbers will improve, and having a statistical basis for wagering is always a good edge to have.
