AgentDib":1nukpw2u said:
This isn't too surprising if you understand that PFF's sole product is their black box player grades. If that's the only input then it would be silly to be optimistic about our O-Line. Of course reality is a lot more complicated as we are moving players to new positions and continuing to develop them, but PFF is marketing their product here and looking at the big picture doesn't help them do that.
FO's stat based alternative is much better at gauging overall offenses, defenses, and team production. For something like this FO is useful while PFF provides no value. FO fails too, however, as you start to look at individual components. Sadly, the people who do get the balance between film and stats right work for NFL teams and don't publish all their information for free on the internet.
I've had a lot of the same thoughts.
I do think FO is actually pretty solid at individual stats, but their formulas only allow them to evaluate a fraction of the positions (namely, skill positions on offense), and unlike PFF, they grade ONLY by results and not by the eye test, which has obvious limitations.
You are right that PFF's micro analysis of each player doesn't work on a team level, because doing so ignores how teams are often less than or greater than the sum of it's parts. Additionally, it seems like of the 30 or so PFF reviewers, some are good at what they do and others aren't. Their grades are subjective and their accuracy will depend on the skill of the reviewer. That's one reason why some of their individual grades seem right on the money, while other grades are completely baffling.
In regards to how teams grade players,
these days evaluating a player is more about "fit" than ever. How good a free agent or draft pick plays or produces is secondary to how well he fits scheme. Or at least, it seems to be this way among the very best teams- Seattle, Green Bay, New England, etc. So I'm guessing that many if not most NFL teams have evaluation stats on players that are unique to their own organizations.