How the model works
Two independent projections, one edge against the market, and every qualifying game graded at the published number.
The process is built to be repeatable and honest rather than clever. Every Thursday at 8:00 AM ET it runs every FBS-versus-FBS game on the board and publishes the result, with no hand selection.
Two independent projections
Each game is projected two separate ways, from SP+ and from FPI, with a home-field adjustment. Each projection is compared to the market's consensus spread (the median across US books). The gap is that model's edge.
The dual edge
When both projections land on the same side of the number, the game has a dual edge: the smaller of the two edges, so it only counts what both models agree on. When they disagree, the dual edge is zero. Every game on the board is ranked by its dual edge.
The plays
A game is a play when its dual edge is at least 1.5 points. Below that, the gap is inside the models' own margin of error. There is no fixed number of plays: some weeks have many, some have few, and the record takes the side both models favor on every one.
The Chives Alpha Ratio
Each game also shows the Chives Alpha Ratio (the Chives Ratio, for short): the dual edge divided by the projected margin, with a floor so a near pick'em doesn't divide by almost nothing. It shows how much of an edge sits in a game projected to be close, as context for each play.
How the record is graded
Every play is graded at the consensus spread when the report published, not a later line, at -110 for 1 unit. A win pays 0.91 units, a loss costs 1, and a push is zero. When the median across books lands on a quarter point, no book offers it, so it is rounded to the half point against the play; the rule never flatters the record. Games that kick off before the report publishes are left out. Results post as games finish, on the track record.
The weekly download
Every report comes with the full board as an Excel file: every game with its ratings, live formulas for the projections, edges, ratios, and plays, this season's team numbers, and the record to date. Change a rating or the edge floor and the board recomputes.
What this is, and is not
This is analysis and commentary for informational and entertainment purposes. It is not betting advice, and it is not a guarantee. The edges are small and most bettors lose. Numbers reflect the lines at publication; odds and lines move.