Does anyone create a roster based directly on PFF advanced stats?

[SyncereBlackout]

Quote: Originally Posted by Aestis

Found his post on reddit and an explanation of his methodology. He said it's all from his own film watching using NFL Game Pass. I'm not a big fan of using one person's opinion even though I'm sure he knows his stuff and may be a PFF employee. The search continues...

Let me know what you end up with. Creating a fictional draft class and it’s imperative there’s continuity between the base roster and my class.


Sent from my iPhone using Tapatalk

[Steelers4190]

This football coach thinks PFF grades are wonky as all hell. Love the advance metrics and stats. Hate their player grades.

[briz1046]

Quote: Originally Posted by Steelers4190

This football coach thinks PFF grades are wonky as all hell. Love the advance metrics and stats. Hate their player grades.

I think a lot of the distrust of PFF comes from people not being able to attribute the proper weight to statistics in general , and using them inappropriately and with no context .
Like any other form of data the information they hold is only as good , and thus useful, as the individual's interpretation of them

[Aestis]

Quote: Originally Posted by SyncereBlackout

Let me know what you end up with. Creating a fictional draft class and it’s imperative there’s continuity between the base roster and my class. Sent from my iPhone using Tapatalk

I chatted a bit with the guy who does them mentioned above in the thread. I'm going to see what I think of his M19 stuff. #1 it may take a while and I probably won't want to wait, and #2 it's only very loosely based on PFF grades. I'm not sold on his rosters yet really.

[Aestis]

Quote: Originally Posted by Steelers4190

This football coach thinks PFF grades are wonky as all hell. Love the advance metrics and stats. Hate their player grades.

Quote: Originally Posted by briz1046

I think a lot of the distrust of PFF comes from people not being able to attribute the proper weight to statistics in general , and using them inappropriately and with no context . Like any other form of data the information they hold is only as good , and thus useful, as the individual's interpretation of them

As briz said, a lot of PFF distrust really does come from lack of understanding. They've invited naysayers (long-time NFL HC & positional coaches) behind the doors and to a man, every coach has come out blown away and retracting their criticism.

That said, the system isn't perfect. No single number is. Their seasonal grades are more effective for certain positions than for others, but overall they are pretty good. From the research I've done, I consider their coverage grades to be the weakest because their system doesn't do a good job (as far as I can tell) of controlling for difficulty of assignment. Their DL & OL grades I feel are probably a bit more accurate.

But overall it's a really strong "directional" grading source. It's not nearly as subjective as guys who haven't ever really read about the methodology seem to assume.

[chunt04]

I had done a test with PFF grades as ratings in Madden. I edited two full teams and played around with it for a bit. It actually worked surprisingly well, as you had to really gameplay around strengths and weaknesses. For instance, sliding your protections was a must, as you may have one offensive lineman rated in the 50s while another may be high 80s. You had to mask those weaknesses to be successful. The same goes for roster depth. Injuries could really cause issue, as the drop off between that back up and starter may be huge.

[SyncereBlackout]

Quote: Originally Posted by chunt04

I had done a test with PFF grades as ratings in Madden. I edited two full teams and played around with it for a bit. It actually worked surprisingly well, as you had to really gameplay around strengths and weaknesses. For instance, sliding your protections was a must, as you may have one offensive lineman rated in the 50s while another may be high 80s. You had to mask those weaknesses to be successful. The same goes for roster depth. Injuries could really cause issue, as the drop off between that back up and starter may be huge.

Thanks for testing this out. Always interested in peoples results.
How many games did you play?
What were the parameters for the games?
Did you play the same game without PFF rosters?Did you call the same plays in both sets of games?



Sent from my iPhone using Tapatalk

[Aestis]

This just got much tougher with extra ratings that PFF doesn't publish to my knowledge. Run block finesse vs strength, etc.