Another reminder to disregard all PFF grades

TheMarathonContinues said:
He played solid against Ravens for mostly the whole game. Texans are a better defense and he looked abysmal I don’t know if it’s the gameplan or Dak or both but they play tough teams and get cute. They don’t have the defense to play as cute as they think they are.
Click to expand...
I actually think Dak made more difficult throws and spread the ball around much better against the Ravens. To me, this game - although it was the Texans D and they do deserve respect - was very simple because all he had to do was read coverage they were throwing Pickens and then find Lamb. They made a decision to frustrate Pickens by blanketing him and then seeing if they could single Lamb.

As for the defense, I am actually a bit higher on Christian Parker after 4 games than most here are. It's easy to blame the D when the O has been getting all the love for YEARS on this team. The defense has been a rudderless ship when all these different coordinator (and their staff's staff) and team leaders - Osi being the last one who was traded in the offseason. So what Parker has done in assembling talent, getting buy in from serious vets who want to win now and then calling games and seeing steady improvement has been impressive to me. Then you add all the young players and injuries. He's got his hands full and it's week 4.
 
OGSixshooter said:
Hold on, it's possible though. Dak nearly threw a few picks and Brisset could have had INTs that hit the receivers in the hands and bounced into the defeneder's. Unlike fans who feel good, PFF tries to take an analytical approach. And sometimes the result obscures the performance. I actually think Dak played better vs the Ravens as far as throwing the ball...so it's possible for a guy to have a not-so-great stat sheet but grade high on PFF.
Click to expand...
No
 
OGSixshooter said:
Hold on, it's possible though. Dak nearly threw a few picks and Brisset could have had INTs that hit the receivers in the hands and bounced into the defeneder's. Unlike fans who feel good, PFF tries to take an analytical approach. And sometimes the result obscures the performance. I actually think Dak played better vs the Ravens as far as throwing the ball...so it's possible for a guy to have a not-so-great stat sheet but grade high on PFF.
Click to expand...
Their QB grades aren't about stats they are about consistency. They did a whole segment on grading the QB when they first came out.
 
We can all agree that there is no perfect rating system for grading a player's contributions. If there were, the draft would be a more exact science.

That said, given their formula, it's entirely POSSIBLE for Brisset to throw 3 INTs with less yardage but grade higher - and that doesn't have to be ridiculous either. Dropped passes, end of half Hail Mary INTs, etc. count as real interceptions, but given PFF's formula, they don't count that against the QB as much. That's all I'm saying...it's possible.
 
Kwyn said:
PFFis a joke. Collinsworth sold the systems and engineers in March to a jock-bro tech company for millions and a stake in their company.

He fired almost everyone still with PFF, only leaving enough to keep the billing and website running and a guy to keep the AI content creation server cooled and watered. Oh, there’s one last human public face he kept on to tweet stuff out and give the illusion they still create content.

Now they use the sub money to license back the data they once owned and use the above mentioned ai to generate their grades.
Click to expand...
Thanks, now I hate even more PFF.

Layoffs are nasty.
 
OGSixshooter said:
We can all agree that there is no perfect rating system for grading a player's contributions. If there were, the draft would be a more exact science.

That said, given their formula, it's entirely POSSIBLE for Brisset to throw 3 INTs with less yardage but grade higher - and that doesn't have to be ridiculous either. Dropped passes, end of half Hail Mary INTs, etc. count as real interceptions, but given PFF's formula, they don't count that against the QB as much. That's all I'm saying...it's possible.
Click to expand...
That’s how PFF says it works

The problem is that the company Collinsworth sold the PFF b2b data systems to only brought over the entry level film reviewers. They didn’t convert any of the analysts or senior analysts whose job it was to review the work. Zero QA in their new structure, no oversight and no escalations.

Collinsworths company, PFF, then fired all of their analysts, they basically license the data back from Teamsworks and have AI create produce tweets.

In the past, a human would actually have reviewed this before it went out. If it was released and they could defend it, one of their analysts would create a post about it with an explanation
 
LouBrown said:
PFF used to be flawed because the grade was based on a sampling, not every snap. It was very imperfect, but was at least useful to contextualize player performance against their peers. And you could trust that it was honest, if not flawed.

Now, it's just an AI based on based on who knows what and has 0 credibility at all. Likely, it'll be a ratings shop for athletes, like 247 and Rivals.
Click to expand...
Ratings are not ai based
If it was it would be a lot better
 
goshann said:
Ratings are not ai based
If it was it would be a lot better
Click to expand...
Yes, that is what they do now. It's owned by Teamwork and they gutted PFF staff.

https://teamworks.com/blog/teamworks-series-f

It's an even dumber version than the one run by Chris Collinsworth, who had already dumbed down the site.
Former Cincinnati Bengals wideout and current Sunday Night Football commentator Cris Collinsworth purchased a stake in PFF way back in 2014 and assumed the title of CEO, which is how you and I ended up seeing the company’s rankings underneath every player during the SNF broadcast. Early PFF grades were somewhat tough to parse for customers, so the new head man mandated that the company switch over to a 1-100 scale for every player in the league, even if it meant potentially losing critical nuance in the process. Guess where Daddy got that idea from.

“My kids pushed me so hard to put it on a 1-100 grade,” Cris Collinsworth said in Coller’s book, Football Is a Numbers Game. “They know Madden. They know the Madden grades. They know what it means. I wasn’t very popular (internally) on that one.”
Click to expand...
https://defector.com/the-decline-of-pro-football-focus-is-bad-news-for-football-fans


Fundamentally, PFF is going to make money with enterprise (see NFL and college teams) data, not with a consumer-facing product. The PFF you and I see will have no merit, because they will not invest in it. Not only is it going to be an AI, it is going to purposely be a lesser installation than the proprietary version. They have no reason to use an effective rating system because they'll get more clicks being stupid.
 
The comparison is meaningless unless you watched every snap of both games.

Stats are context dependent.
 
You stopped the run, you get more opportunities on offense
 
OGSixshooter said:
We can all agree that there is no perfect rating system for grading a player's contributions. If there were, the draft would be a more exact science.

That said, given their formula, it's entirely POSSIBLE for Brisset to throw 3 INTs with less yardage but grade higher - and that doesn't have to be ridiculous either. Dropped passes, end of half Hail Mary INTs, etc. count as real interceptions, but given PFF's formula, they don't count that against the QB as much. That's all I'm saying...it's possible.
Click to expand...
which just shows you why no one should take PFF very seriously
 
You must log in or register to reply here.

Staff online

Members online

... and 174 more.
Total: 27,028 (members: 274, guests: 26,754)
Robots: 1,396

Latest posts

Forum statistics

Threads
476,216
Messages
14,673,204
Members
24,227
Latest member
Nathaniel
Back
Top