I don't know what to make of them. It would be interesting to sit in on a session and see how they operate. I've worked in a stats office and there can be quite a bit of fluctuation from one guy to the next. Everybody has obvious strengths and weaknesses so it's sorted out along those lines, like players gravitating to their logical positions. I never like to comment on Pro Football Focus numbers since it's quite different than anything I've experienced, and we've never been treated to a sample, like a 5 minute clip of how they function. My best summary is they need some competition on the same level, to see how things match up. Heck, I can walk into Bird Bowl tomorrow and pretend a guy is awesome tour caliber until a legitimate star shows up two hours later.
I do know one thing, the percentages quoted on ESPN and elsewhere on likelihood of victory at given points of a game depending on situation are generic crap. If you want to hear guys howl, hang out with sports bettors when those percentages are asserted. It may sound swell if you want to claim it's a league average but once you try to pretend it applies to a given team in that particular game you're simply making a fool of yourself, even if unknowingly. For example, a 10 point favorite might be down by 3, early second quarter. The other team might have the ball first down at your 40. Now, there isn't a competent sports bettor alive who wouldn't realize that the 10 point favorite still enjoys a hefty likelihood to win the game, even if it's lower than initially. The favorite would be at least a 1/3 chalk. Yet ESPN clowns will use PFF numbers, or wherever they come from, and actually give the other team the theoretical edge. Unbelievable. There's something called "In Game" betting now that's increasingly popular. Instead of wagering stopping at kickoff, or second half wagers not taken once the second half kicks off, adjustments are made in the betting line throughout the game. Let me put it this way, every sharp bettor on the planet wishes those generic percentages were used during "In Game" wagering, and not the adjusted odds that account for team strength.