Analysis

Jesse McConnell, Oct 6, 2026. NFL, 5 min read

Offense and defense EPA per play for all 32 teams, free, on the team ratings page and every team file. Since 2015, 8 of 11 champions were above average on both sides of the ball, and none was below average on both.

Jesse McConnell, Oct 5, 2026. NFL, 10 min read

The Margin QB Grade is the points a quarterback adds per play, 50 is average. Against eleven other ratings on 2016 to 2025, it tracked a game's scoring best (0.818, vs 0.712 for QBR and 0.688 for QBERT), tied with a four-stat blend. On stability and prediction, QBR, QBERT and others match or edge it. Hand-charted throw data didn't help.

Jesse McConnell, Sep 30, 2026. NFL, 6 min read

A new page: for any NFL team at any point in its season, the 12 most similar teams since 1999 and how they finished. We tested it before launch, and it predicts worse than a simple formula. What comps are good for (range, style, context) and what they aren't (forecasts, bets, playoff odds).

Jesse McConnell, Sep 29, 2026. NFL, 6 min read

A new tool: every QB since 1999, game by game, with an honest range around each line and a 'right now' view. Brock Purdy leads today, but his range overlaps everyone in the top nine. Late bloomers are rare: 2 of 56 QBs who were average or worse after 32 games became good ones.

Jesse McConnell, Sep 29, 2026. NFL, 9 min read

Rebuilt from 5,814 plays and tested against mid-season QB changes back to 2015: yards after the catch and QB hits follow the quarterback, so they're out. With them gone, the 49ers are the only cast clearly above average, and the Browns, Steelers and Chargers are clearly below. The other 28 teams can't be ranked with any confidence yet.

Jesse McConnell, Sep 28, 2026. NFL, 10 min read

Draft slot predicts which QBs make it better than any college stat, mostly because high picks get the snaps. Among QBs who got a real chance, ESPN's pre-draft board predicted nothing about how well they played. A team taking a QB earlier than the board did, and college rushing, both did.

Jesse McConnell, Sep 27, 2026. NFL, 6 min read

We measured five seasons of NFL injuries six different ways. About a fifth of a team's injury burden carries from one year to the next. Arizona and Detroit top the list, Kansas City and Jacksonville anchor the healthy end, and the betting market already knows all of it.

Jesse McConnell, Sep 25, 2026. NFL, 8 min read

Love ranks 6th of 32 in EPA per dropback and 2nd of 17 in the playoffs. Herbert ranks 19th, and the efficiency stats never rated him. Clutch doesn't repeat. What separates Mahomes, Allen and Jackson is what happens when the play breaks. And since 2022, nobody has been more efficient than Purdy.

Jesse McConnell, Sep 17, 2026. NFL, 6 min read

Denver's opener was historic, at the 99.8th percentile of every team-game since 2018. It is also close to worthless as a predictor, and the widely shared 29.8 percent figure does not reconcile with the charting it claims to come from.

Jesse McConnell, Sep 15, 2026. Model log, 6 min read

NFL week 1 graded (the calls sheet went 2-3, -1.76 units at a flat paper unit per call) and the review that followed rebuilt the NFL sheet from four slots to five, replaced the kill bar with one...

Jesse McConnell, Aug 12, 2026. Calls, 5 min read

Most picks content sells you the loudest opinion on the board. We publish the quietest one, because that is where our model has been measurably least wrong. Every call carries the slot's actual track record, including the slots where that record is a coin flip.

Jesse McConnell, Jun 29, 2026. CFB, 8 min read

There's no public EPA in college past 2021, the schedules are a free-for-all, and the biggest award is decided by story as much as by snaps. So we built an opponent-adjusted rating that survives all three, then learned that predicting the Heisman means measuring the team, not just the quarterback, and that even then a stubborn slice of the trophy is pure narrative.

Jesse McConnell, Jun 29, 2026. NFL, 9 min read

Success isn't wins. It's points per drive, and the QB controls twice as much of that as he does winning. Accuracy is the engine, 'clutch' is noise, QBR is the least stable number in football, and the rating that beats it credits the thing the old stats ignore: running.

Jesse McConnell, Jun 14, 2026. CBB, 4 min read

Georgia leads the model's board at 41.8 percent, North Carolina sits second, and Troy is the 0.4 percent story nobody bracketed. The full title board, plus where to watch the model disagree with the book.

Jesse McConnell, Jun 13, 2026. NBA, 7 min read

Totals were the real edge. Killing a broken calibration table was the best decision we made. And the model favored the Spurs in a Finals the Knicks won. The full receipts, wins first.

Jesse McConnell, May 27, 2026. CBB, 3 min read

Every CWS champion since 2009 except 2022 Ole Miss passed all seven Selection-Day thresholds. Four 2026 teams matched the profile. The full write-up is on the methodology page.

Jesse McConnell, May 26, 2026. CBB, 5 min read

Where the model disagrees with DK on the 2026 NCAA baseball field, plus how the 50K Monte Carlo gets built and what not to do with it.