Week 3 Recap: Gibbs Goes for 41 and We Grade Our Own Start/Sit Calls

Fifteen of sixteen games are in the books, with Philadelphia at Chicago still to play. Five of them were decided by three points or fewer. One was not close at all.

The week in one screen

Jahmyr Gibbs was the best fantasy player in football. Twenty carries for 99 yards and two scores, plus seven catches for 65 and another, in a 31 to 24 win over the Jets. That is 41.4 PPR points and a reminder of why we kept him second in our rest-of-season rankings.

Jacksonville 35, New England 6, per the box scores. The Patriots managed six points and their leading receiver had 40 yards. Trevor Lawrence threw three touchdowns.

Five one-score finishes: Indianapolis over Houston by two, Washington over Seattle by two, Baltimore over Dallas by three, Pittsburgh over Cincinnati by three, and Cleveland over Carolina by three.

Jaxon Smith-Njigba put up 35.4 in a losing effort, ten catches for 128 and two scores. Bijan Robinson had 35.3 on Thursday. Brock Purdy threw four touchdowns without a pick.

Our Friday scorecard: three right, three wrong

We published six start/sit calls on Friday. Here is how they did, including the ones that did not work.

Right: start Garrett Wilson

We said Detroit was the softest matchup in the league for receivers at 42.9 points allowed per game, and that a Jets team favored to trail would throw. Wilson caught ten of thirteen targets for 107 yards and a score, 26.7 points. That is the call working exactly the way it was reasoned.

Right: temper Jonathan Taylor

Houston came in as the toughest run defense in football at 12.5 allowed. We said start him but expect closer to 12 than 25. He got 23 carries for 68 yards and no touchdowns, 9.2 points. The matchup won.

Right: sit Hunter Henry

Jacksonville had allowed 1.4 points per game to tight ends, the lowest figure at any position on the board. Henry caught one pass for five yards. If you started him because New England was a home underdog and tight end is thin, that is the week you had.

Wrong: start Dalton Schultz

This was our most confident call. Twenty two targets in two games, walking into the third softest tight end matchup. He saw three targets and finished with 30 yards.

Target share is the most predictive thing we have and it still misses. Houston lost by two in a game that stayed tight, which usually helps a possession tight end, and it did not. No excuse available.

Wrong: flex Omarion Hampton

We flagged the risk in the piece itself: Buffalo is soft against the run, but the Chargers were seven point underdogs and a team down two scores stops handing off. That is what happened. Fifteen carries, 56 yards, 5.6 points.

Naming the risk is not the same as getting it right.

Wrong: temper Trevor Lawrence

New England had allowed 2.3 points per game to quarterbacks, the lowest mark in the league. We said use a comparable option if you had one. Lawrence threw for 182 yards and three touchdowns in a 29 point win, 19.8 points.

Two weeks of defensive dominance was a smaller sample than it looked, which we wrote at the time and then did not weight heavily enough.

What it means going into Week 4

The matchup data that drove the three correct calls is the same data that drove the three wrong ones. Two weeks of points allowed is a real signal and a small one, and this week it was worth roughly a coin flip on the edges and quite a lot in the extremes. Detroit against receivers and Jacksonville against tight ends were both extremes, and both held.

Our matchup ratings now carry three weeks for most teams, which makes them meaningfully steadier. The waiver wire updates with the new week's adds.