What is this based on? Just who he thinks?
Pretty much. PFF likes to create these proprietary rating scales and metrics to try and explain everything. I used to pay for their premium service when it was $30/year. This is back around 2012 or so when they would explain their formula for rating a specific metric. Their site was pretty rudimentary but it wasn't hard to navigate or anything. Lot's of measurements were displayed in a manner that was basically a glorified excel sheet. It was actually kind of nice.
They had various measurements that never really made sense to me. They had QB ratings "under pressure" and at the top of the list you'd basically see the elite QBs intermingled with guys like Vick. Of course a mobile QB is going to be "under pressure". They hold the ball relying on their mobility so they are always going to be scrambling away from defenders and that's where their production comes from. They still sucked, but somehow they were at the top of this list. Manning, Brady, Rodgers don't get into pressure situations nearly as often because they read the defense and get rid of the ball before anyone can move them off their spot.
PRP (Pass Rush Productivity) was one that I always found to be odd because it treated a QB Hurry as the equivalent of 1/2 sack. Not joking, a sack was worth 1 point, and a hurry was worth 0.5 points. Obviously, that's completely stupid because a hurry has no defined outcome (could be a throw away or could be a TD) and a sack is a guaranteed loss of yardage and down. That was when I started to wonder if I was wasting my $30/year. It was one of their premium stats and I just couldn't get on board with the idea that you can treat a sack and hurry that way. Not to mention their coverage stats for DBs was based on "closest to the ball". Suppose a CB gets tossed at the line and 15 yards downfield the closest DB is the safety trying to make up for it. That reception goes against the safety simply because he is now the closest defender? It just didn't make sense.
To their credit, I do remember looking at their PRP one year and Justin Houston was ranked really high even though his production wasn't through the roof. I think it was after his rookie season and I was wondering how a guy with only 5-6 sacks rated so highly. Well, he didn't really get a ton of playing time so based on the time he did get he was doing very well for the number of snaps he had. Turns out, dude went on to have some very good seasons once given the proper workload. That said, you don't need a convoluted algorithm to extrapolate. If a guy is getting 15-20 snaps/week and produces X amount...how about we just double it and see if he produces Y amount.
Watch any SNF or MNF game and you'll see weird stats. Amazon, PFF, or whoever. Odds of completion on some difficult play is a really weird one. The one where they're like, "23% chance of catching the ball". Says who and based on what? Even if the estimate is 100% accurate it's based on leaguewide data so we're gonna pretend like the odds are the same for all QB/WR combos? Guarantee you the odds of Aaron Rodgers completing a pass to any WR is a hell of a lot higher than Zach Wilson completing a pass to Randy Moss or Jerry Rice in their prime. Pair Rodgers with prime Moss or Rice, what are those odds gonna look like?
I appreciate numbers and whatnot, but that site is complete trash. Watt was 66th in the NFL in sacks, and he's 4th best? I don't care if it's injury or not, the best "ability" is and always will be availability. To me, it's okay if a guy falls out of the Top 10 based on playing time. There is no need to rate them higher based on past production if you are doing a season-by-season analysis. Looks like the 4th, 5th, and 6th leading sack players are all omitted. How is that possible? I don't care what you
could do for me, I care about what
did do for me.