Our Matchup Calls Went 1 for 7 in Week 4. The Usage Calls Went 4 for 4.

We published eleven lineup calls for Week 4. Seven of them were argued from defense versus position. Six of those seven were wrong.

Four were argued from workload and role. All four were right.

That is not a run of bad luck. It is a sample size problem we should have caught before publishing, and the split between the two groups is clean enough to act on.

The calls that came from matchup

Player	We said	Why	Result
Javonte Williams	Sit	Houston allowed fewest PPR to RBs	19 car, 62 yds, 3 TD, 5 rec. 31.3 PPR
Jonathan Taylor	Temper	Washington 2nd stingiest vs RB	20 car, 95 yds, 2 TD. 22.7 PPR
Josh Downs	Start	Washington worst in football vs WR	3 of 4 for 16 yds. 4.6 PPR
Travis Kelce	Start	Las Vegas 2nd most generous vs TE	2 of 4 for 15 yds. 3.5 PPR
Bucky Irving	Start	Green Bay most generous vs RB	16 car, 61 yds, 0 TD. 6.1 PPR
Tyler Higbee	Sit	Philadelphia 2nd stingiest vs TE	5 of 6 for 67 yds. 11.7 PPR
Tyler Warren	Start	Washington 3rd most generous vs TE	5 of 5 for 41 yds. 9.1 PPR

Javonte Williams is the one that should sting. We told people to bench him into the league's best run defense and he scored three times. Anyone who followed that advice lost a week they should have won.

Jonathan Taylor is the same mistake pointing the other way. We called him an uncomfortable temper and he put up 22.7.

The calls that came from usage

Player	We said	Result
Emeka Egbuka	Sit, nobody knows who is throwing	2 of 4 for 13 yds. 3.3 PPR
Jordan Addison	Start, but temper it	4 of 6 for 52 yds. 9.1 PPR
Dontayvion Wicks	Flex, not a difference maker	3 of 5 for 18 yds. 4.8 PPR
Braelon Allen	Speculative, not plug and play	14 car, 59 yds. 8.7 PPR

None of these were exciting. All of them were accurate. Addison in particular is the shape of a call we want to make more of: his targets were climbing, the matchup was poor, we said both things out loud and landed on a modest projection that is almost exactly what happened.

Why the matchup numbers lied

Three games is not a sample. It is three games, and what a defense has allowed through three games mostly describes who it has played.

Washington came into Week 4 having allowed more fantasy points to wide receivers than any defense in football. They had faced Philadelphia, Dallas and Seattle. Dallas alone has CeeDee Lamb, who has been one of the five best receivers in the league this year. That number was a readout of their schedule, not their coverage.

Houston looked like the stingiest run defense in football. They had faced Buffalo, Cincinnati and Indianapolis, and the game scripts in those matchups did not involve much grinding on the ground.

When we wrote that the middle of a defense versus position table is noise and the extremes are signal, we had it only half right. The extremes are where schedule distortion is largest, because it takes an unusual set of opponents to push a defense to the top or the bottom of a 32 team list after three weeks.

What changes

Workload is the primary input. A back with 19 or 20 carries and goal line work is a start regardless of the defense across from him, because the floor of that role survives a bad matchup. Javonte Williams had 19 carries. Jonathan Taylor had 20. We had both data points and overruled them.

Matchup becomes a tiebreaker, not an override. It is useful for choosing between two players with similar roles. It is not a reason to bench a workhorse or to start a player whose target share is unproven, which is exactly what we did with Josh Downs, who had four targets in a game we had called the softest spot on the board.

And the sample has to be big enough to mean something. We will not lead with defense versus position again until there are at least six games behind it, and when we use it we will say what the schedule behind the number looks like.

The full Week 4 box scores are at ESPN. Our current positional board is still on the matchup ratings page, now with the column labels fixed so rank one reads as the weakest defense rather than the best, and this week's adds are in the Week 4 waiver wire.