I'd be interested to see what PFF bases their "News and Analysis" on for two rookies and a system vet that played on another team last year.
Those are what peeps in the analysis biz call unconstrained variables.
I agree that it'd be interesting to see. PFF's ratings are a total black box, and often seem to be based on nothing more than "these are the players we like." I like
FootballOutsiders, but I'd be a lot more comfortable with DVOA and other
FO stats if they were explained with formulas and stuff. I'll give them credit for telling us what goes into the recipe, but not the amounts and procedure for making the DVOA cake. So DVOA is also a black box, but it's a little less black-boxy than PFF's "pffffff!"-worthy ratings.
That said, one of the first important things Bill James discovered in baseball, way back when he was a night watchman at a warehouse, was that minor-league stats
could in fact tell you quite a bit about how well a player should be expected to perform in the majors,
as long as you properly adjusted them. The adjustment was key, because, for example, the PCL was a very offense-inflating league (the parks probably), so sometimes a big PCL hitting prospect would be hyped based on his monster triple-A numbers, and then fail to hit as expected upon hitting the majors, especially if it was with the Dodgers, whose stadium was an offense graveyard in the '80s and '90s.
Techniques for projecting performance have advanced a lot in recent years, and I have very good reasons to believe that rookie NFL performance can be predicted better than you might expect, using just a player's college (NFL-minor-league) stats and, when available, his Combine and/or Pro Day numbers.
So I generally think PFF ratings are kind of a joke, but it
is possible that they're basing their assessment of the rookies in the Seahawks' 2022 OL on their college performance and numbers from the Combine and/or Pro Days.
