EDGE LABS
From the lab
Weekly columns from Ace, Stack and Buck, and the research notes behind our ratings.
Ace's Outright Dog
By Ace Wisconsin at UCLA: take the stable side
Wisconsin arrives with its quarterback order set, while UCLA is being asked to hold favorite status through a major change.
Column tally+1.46 units, through Oct 15
Buck's Primetime Props
By Buck Bills at Rams: the board is paying for the shootout
Every number on the Monday night sheet assumes a track meet, and I am taking the unders on the pieces that need one to cash.
Column tally+3.69 units, +3.05 in week 5, through Oct 15
CFB ratings movers, week 6: Buffalo's win at Toledo is the biggest swing, and Georgia retakes number 1
After week 6, Buffalo's 23 to 8 win at Toledo was the biggest swing in our college football rating, Georgia's win at Alabama flipped our top two, and Notre Dame lost rating points in a 51 point win. Each move below is traced to the game and the margin our number expected.
Buck's Primetime Props
By Buck Baltimore Ravens at Atlanta Falcons: the unders the board forgot
Buck walks the whole Sunday night board, takes five props at their prices and names five he would not touch.
Stack's Teaser Card
By Stack Raiders, Bengals and Browns, moved through the key numbers
Three legs, six points each, and a football reason under every one of them.
Column tally0-1, -1.00 units, through Oct 15
Ace's Outright Dog
By Ace Colts at Steelers: the price on the visitor
Indianapolis has flaws, but Pittsburgh has not given its offense a clean case at this price.
Buck's Primetime Props
By Buck Tampa Bay Buccaneers at Dallas Cowboys: I am under on the noise
A rookie quarterback, a home favorite laying 8 and a board shaded for the highlight reel. I take five unders and stay off the loud names.
Ace's Outright Dog
By Ace The short favorite has some explaining to do
USC brings the cleaner offense to Penn State, while Central Michigan meets an Ohio attack that stopped moving.
Buck's Primetime Props
By Buck Falcons at Saints: the big numbers are one night deep
Three unders on a Monday night board where the lines sit above what the recent games show, and one big-name prop I would leave alone.
CFB ratings movers, week 5: Missouri climbs 12 spots, and LSU drops after a 49 point win
After week 5, Missouri's 28 point win over Florida was the biggest swing in our college football rating, and LSU lost rating points despite winning by 49. Each move below is traced to the game and the margin our number expected.
Ace's Outright Dog
By Ace Jacksonville and Atlanta take aim on the road
Two road underdogs arrive with a reason to make their hosts answer.
What a surprise QB start really costs
We tested four feature layers against our NFL margin model, one at a time, across four development seasons. Two made it worse and one barely moved it. The one that survived says a team starting anyone but its real starter costs about 2.8 points.
What our model refuses to claim
Our NFL model beats its own predecessor on margin accuracy, loses to the closing spread in every held out season, and we print both numbers instead of picking the one that flatters us.
The preseason prior is worth eight games
Every team in our college football rating starts the season at its preseason number, then updates one game at a time on the live ratings board; expected wins on that board still come from the preseason walk. The weight on each new result shrinks all season, calibrated so the preseason prior carries the same influence as about eight games of evidence. Here is the math, and the two seasons where we checked it against what actually happened.
Why our model demoted Denver
Denver tied for the best expected win total our first NFL rating produced. In the model that replaced it, on its Aug 18 build, the Broncos ranked seventh, and the reason is the gap between winning games and winning them by enough.
What the transfer portal is worth in our CFB model
Our college football rating carries two portal inputs. The raw count of net transfers came out slightly negative and the average star rating of the incoming class came out positive. Volume is not the signal. Quality is, and here are both weights next to the teams that show the split.
The one totals angle that cleared our gate, and then failed its live test
We tested dozens of college football angles against eight seasons of closing lines and killed almost all of them. One survived the backtest: early season unders on high totals. Then the 2026 forward test graded it at 47.3 percent unders on 131 games, below the 52.38 percent gate. Here is the rule, the evidence, and the final result.
What a first-year head coach costs
A first-year head coach costs a team about 2.3 rating points in our college football preseason model, and every one of the six weights behind that number is public.
What returning production is really worth
Returning production is the input everyone quotes in August and almost nobody prices. In our college football model it carries a weight of 5.87 rating points per unit, third behind performance carryover and talent, and here is what that means with the real 2026 numbers behind it.
The whole CFB model in seven numbers
Our college football preseason rating is one line of arithmetic: a constant and six inputs. It beat an SP plus carryover baseline on average out of sample, across two held out seasons and without losing either, and here is every number, including the one that came out backwards.
How we built our NFL power rating
We published a simple model, got called on its blind spot, and rebuilt it on 6,223 games of opponent adjusted margins. The new version won all four test seasons. Here is the whole build, including the number most services would bury.
No columns from Stack yet.
Show everything