[Martinr]

Sportsbook Breakers at KillerSports/dotcom finds plays using SDQL. I think data bases are a great tool and I've used a similar db for horse race betting for years so this post isn't intended to denigrate their methods. Some of the stuff to be found using SDQL is pretty amazing.
In the latest KillerSports newsletter they publish a system that intends to capitalise on the Week 2 "overreaction" angle that I've heard ad nauseam this week. The system says to Fade teams in Week 2 that have a line 8 or more points lower than they had in Week 1. Seems logical. It looks good on paper too. 118 plays for a record of 48-66-4, which means fading these teams has had a 58% SR. Smallish sample size but if KillerSports thinks it sufficient then who am I to argue?

This year 5 teams qualified for the system. The Fade plays were/are -

Denver -6
New England -6.5
Carolina -13.5
San Diego -3
Chicago -3

If these teams were actually a 42% chance of covering then I make it a 3.1% chance that the system's poor showing this year is an expected norm. Then again, systems become obsolete (often, almost as soon as they are discovered), markets adjust, and coaches wise up to the dangers. Some of those 4 results this week buried the system 10' under.

Eagles @ Bears
Pass

[esulima]

Well they all covered!

[Martinr]

Well they all covered!

And the system had them as ATS Fades