BuffaloBills

Start with the number that does not flatter us

Rush defense, z-scored across the 32 clubs with the sign flipped so higher is tougher: -1.73. That is 31 of 32. There is no version of this argument that begins anywhere else, because a supporter who leads with the good numbers in August is a supporter who has forgotten every January this club has had.

The offense is not the question and has not been for years. Pass efficiency 0.81, rush efficiency 0.87, quarterback input 0.93, a line at 10 of 32. Those four numbers are the reason the club has spent a decade being told it is close. Being close is not a plan, and this market has run out of patience with the word.

What a top-heavy roster actually is

Take the seven starting slots one at a time, because that is how the club figure is built and it is the only honest way to read it. Quarterback 100.0. First back 98.6. First receiver 84.9. Tight end 82.7. Then second receiver 64.3, flex 49.9, and a second back at 16.5. Four excellent players and three the model does not think much of.

The quarterback slot carries 28% of the club figure, which is the largest weight on the page and is the reason Buffalo rate where they do. Take him out and this is not a top-ten roster. That is not a complaint about the weighting, it is what the weighting is for: a quarterback is on the field for every offensive snap and a second back is not.

A HexScore compares a player to others at his own position, so Dalton Kincaid at 82.7 is an excellent tight end and not the equal of Josh Allen at 100.0. Worth saying plainly, because the numbers sit in the same column and the temptation to add them up is the whole reason the club figure weights them instead.

Where the model and I part company

My standing view is that a receiver room is judged by its second and third options rather than its first, because the first is never the one deciding a January game. On that test this room fails and it is not close: Khalil Shakir at 64.3 and Keon Coleman at 49.9, against a first receiver at 84.9. I have been making this argument about other clubs for years and I am not going to stop making it now that it lands on this one.

I also think a quarterback who runs raises the floor of everyone around him, and the model does not agree in the direction I would like. It rates the players around him where it rates them regardless, and a club with 2 players inside the league's top 50 is a club whose depth is an assumption rather than a measurement.

On age I am simply outvoted. The value-weighted age is 26.9, 21 of 32 counting youngest first, with only 12.4% of the club's rated talent held by players under twenty-five. That is not a rising roster. It is a roster that is as good as it is going to be, and it is very good.

What would have to be true

The schedule helps, and it is the one part of this that is not an opinion: strength of schedule ranks 5 of 32 counting easiest first, and 13 of 17 games are played outdoors, which in this division is a competitive fact rather than a hardship. The bye falls in week 7.

So the case is narrow and it is honest. A club that is 7 of 32 in roster strength and 8 of 32 on the blended team rating is a good football team and not the best one, and I am not going to tell you otherwise. Its bad number is a run defense at 31 of 32 and its second bad number is a starting seven that stops after four. The season turns on whether either moves, and one of them is easier to fix in a season than the other.