Yes, for those I checked at least.
I don't know what went wrong with that preseason Martinez example, but after spot checking a half dozen of those queries I illustrated earlier the results match the chart in the OP.
I don't know if those queries are an unintended back door or a teaser, but don't tell PFF about it just in case.
I suppose the most surprising grade would be Clinton-Dix earning "elite" status when measured against the eye test. Then again, safeties are the hardest to assess from a broadcast viewing because much of what they do is not on the screen. The same is true of corners to a large extent. Current perceptions may be colored by past season observations or "confirmation bias", my phrase of the day it seems.
PFF states they grade each play on a -2 to +2 scale along the lines of traditional coaches grading.
What's interesing about that scale is the following PFF QB example:
"At one end of the scale you have a catastrophic game-ending interception or pick-six from a quarterback, and at the other a perfect deep bomb into a tight window in a critical game situation, with the middle of that scale being 0-graded, or ‘expected’ plays that are neither positive nor negative."
https://www.profootballfocus.com/pff-player-grades
This bets a lot of questions.
Is an interception a +2 play? Maybe. Is a +2 play worth more than double a +1 play? That example seems to suggest so. Does a pick far outweigh a couple of missed tackles that don't result in big gainers? I'd presume so. If a CB or S has tight coverage such that the ball is not thrown his way is that +1 or a merely "expected" 0? I would think that's +1 since that's one kind of shut down corner. In Asomugha's 2008 - 2010 seasons (3 Pro Bowls, 2 first team All Pros) he was targetted a grand total of 98 times, about 2 times per game. Do you grade him 0 on those 94% of opponent passes, give or take, when opponents avoid throwing his way like the plauge while providing little opportunity to make the big +2 splash play? I don't think so.
I've not been able to find anything specific from PFF as to the fine points of assigning grades or the algorithm they plug them into. That's not surprising. It's surely deemed proprietary.
Getting back to the C-D grade, lets say those picks weigh heavily into the grade. He also has 3 passes defended in 4 games while his previous season high was 7 in 2016. As for the plays where Clinton-Dix is playing high safety and we don't see what he's doing, perhaps PFF has found his anticipation, decision making and coverage has been spot on in those instances. Under these conditions, it is plausible to think a grading system would make a modest charge against the grade for a few missed tackles that did not result in big extra yardage while still concluding "elite".
I don't know about anybody else, but I'm not going through 4 games of coaches tape, which are available on Game Pass, to focus on what Clinton-Dix has done on every snap. I doubt anybody here will either.
A PFF grade is not infallible. They are more credible now with 3 levels of review than they were in the early years. They say the 3rd. level is performed by ex-coaches deciding on any disagreement at the first two levels. But since it's more credible than overweighting a small handful of poor plays without being able to see the whole picture, I think it best to go with it until there is something says otherwise.
As for the rest of the grades, they seem about in line to me.
