Prescott's final season ranks (various metrics)

[percyhoward]

3rd in ESPN Total Adjusted EPA
4th in Total QBR (ESPN)
4th in Air Yards/Completion (PFR)
5th in DVOA
5th in ANY/A (Football Perspective)
5th in PFN OSM
6th in ANY/A (PFR)
7th in CPOE (Next Gen)
7th in EPA/play (Baldwin/NFLscrapR)
8th-lowest in Bad Throw %
9th in PFF offense grade
9th in On-Target %
10th in NFL passer rating
12th in CPOE (Baldwin/NFLscrapR)

NextGen's CPOE is based on target depth, receiver separation, and pressure on QB
Baldwin's CPOE is based on target depth only

FO hasn't updated DVOA yet, but I saw a twitter post (which I can't find now) with the final rankings that had Prescott 5th.

I generally left out things that are already included in other things (yards, completion percentage, yards per attempt, TD/INT ratio are all included in passer rating... adjusted completion percentage and average depth of target are already included in CPOE, etc.)

Depending on which source you want to believe, Prescott's receivers dropped 6.9% (PFF) of his catchable passes, which was the most in the NFL, or 5.7% (PFR) which was 5th-highest.

As for which stats are the best, it depends on how you measure that. Passer rating, ANY/A, and EPA all have strong win correlations, but don't concern themselves with how to divide credit/blame. Total QBR, PFF grade, and PFN's OSM all attempt to divide credit/blame.

The bottom line for objectivity and win correlation would be EPA/play, but there are other stats that predict future EPA better than EPA itself. Metrics with varying degrees of subjectivity (DVOA, PFF grade, and SIS Total Points) are even better at predicting future EPA because they include credit/blame considerations. (SIS doesn't have their 2019 total points in yet, but Dak has been Top 5-7 most of the season). Despite it's high correlation with wins, ESPN's Total QBR is not very good at predicting future EPA.

PFF's player grading system has evidently come a long way since they started. Based on my early experiences with them, my only involvement with PFF grades has been either to criticize them or tell people to ignore them. But PFF's grades' ability to predict future EPA better than EPA itself shows that there is an advantage to the subjectivity involved in assigning credit and blame, if you're doing it right. Based on the information below, I have to admit that the people at PFF know what they're doing.

Summary: You probably can't go wrong by taking Dak's average rank from the top 8 metrics you see below.

 

[WillieBeamen]

 

[Zman5]

3rd in ESPN Total Adjusted EPA
4th in Total QBR (ESPN)
4th in Air Yards/Completion (PFR)
5th in DVOA
5th in ANY/A (Football Perspective)
5th in PFN OSM
6th in ANY/A (PFR)
7th in CPOE (Next Gen)
7th in EPA/play (Baldwin/NFLscrapR)
8th-lowest in Bad Throw %
9th in PFF offense grade
9th in On-Target %
10th in NFL passer rating
12th in CPOE (Baldwin/NFLscrapR)

NextGen's CPOE is based on target depth, receiver separation, and pressure on QB
Baldwin's CPOE is based on target depth only

FO hasn't updated DVOA yet, but I saw a twitter post (which I can't find now) with the final rankings that had Prescott 5th.

I generally left out things that are already included in other things (yards, completion percentage, yards per attempt, TD/INT ratio are all included in passer rating... adjusted completion percentage and average depth of target are already included in CPOE, etc.)

Depending on which source you want to believe, Prescott's receivers dropped 6.9% (PFF) of his catchable passes, which was the most in the NFL, or 5.7% (PFR) which was 5th-highest.

As for which stats are the best, it depends on how you measure that. Passer rating, ANY/A, and EPA all have strong win correlations, but don't concern themselves with how to divide credit/blame. Total QBR, PFF grade, and PFN's OSM all attempt to divide credit/blame.

The bottom line for objectivity and win correlation would be EPA/play, but there are other stats that predict future EPA better than EPA itself. Metrics with varying degrees of subjectivity (DVOA, PFF grade, and SIS Total Points) are even better at predicting future EPA because they include credit/blame considerations. (SIS doesn't have their 2019 total points in yet, but Dak has been Top 5-7 most of the season). Despite it's high correlation with wins, ESPN's Total QBR is not very good at predicting future EPA.

PFF's player grading system has evidently come a long way since they started. Based on my early experiences with them, my only involvement with PFF grades has been either to criticize them or tell people to ignore them. But PFF's grades' ability to predict future EPA better than EPA itself shows that there is an advantage to the subjectivity involved in assigning credit and blame, if you're doing it right. Based on the information below, I have to admit that the people at PFF know what they're doing.

Summary: You probably can't go wrong by taking Dak's average rank from the top 8 metrics you see below.

Yeh but he doesn't pass my eye test. He sucks. Worst QB ever in the history of NFL.

 

[WillieBeamen]

 

[IheartRomo]

You do understand that all QBs over a large enough sample will play worse against the top 30% of the league than they do against the bottom 70%, right? This is not just football but any sport. The bad teams are bad for a reason.

All QBs in general play worse against good/great teams than they do against bad/terrible ones. This isn't rocket science.

 

[StevenOtero]

Not impressed

 

[percyhoward]

You have to understand that many of the above metrics (DVOA, Total QBR, PFF) are adjusted for opponent.

 

[BoysForLife]

Percy, please revise that thread title.
When I read "Dak's final season", you really got my hopes up for a second....

 

[CWR]

I agree, but I differ with you in that I dont think he sucks or cant be better than he already is. We need to franchise Dak. We cant saddle our new staff with a rookie QB. Dak is at least a part of the transition. However, we shouldn't make a long term decision on QB without their evaluation and input.

 

[Creeper]

Dak was actually really really good for most of the season, but let's face it. His 2019 season will and should be judged by how he played in the biggest game of the year, the division title game against the Eagles. Dak wasn't just bad, he was career bad. Replay that game and count the throws that were off target, some by a mile. I have no idea if he was injured or not but he was bad and it cost the Cowboys the game. I draw the comparison to Wentz who was throwing accurate passes to the right receivers the last month of the season. I am a Dak supporter but he needs to show up better in big important games. If that Eagles game is any indication of how Dak will play under pressure, the Cowboys have a big problem.

 

[shabazz]

3 field goals and no Touchdowns in the single most important game of the season against the mighty 7-7 Eagles. Losing the NFC East. Pretty stats though

 

[DFWJC]

For sure was a tale of two seasons.
I've said earlier that I've never seen a QB run up so many stats on poor teams and play so average or worse vs good teams.
I do NOT blame him for having good stats....certainly better than playing poorly vs ALL teams.

To me, the QB rating of 10 puts him about where he is in the league--which is still not terrible.

 

[big dog cowboy]

His 2019 season will and should be judged by how he played in the biggest game of the year, the division title game against the Eagles. Dak wasn't just bad, he was career bad.

It was also the most injured game of the year for him.

The problem was Garrett. He should have reduced Dak's passing attempts greatly and given the ball to Zeke. The fact Dak had more than 3 times the number of attempts compared to Zeke's rushing attempts is stupid crazy.

 

[HungryLion]

Dak is clearly in the 8-10 range of NFL QB’s. It’s clear. It’s obvious. Based on both film and watching the players, and the numbers themselves.

The only real argument is whether 8-10 is good enough to win a championship if he gets paid a lot of money.

I’ll add. Inconsistency is an issue. It would be better if he constantly played in the 8-10 range. Unfortunately he has streaks where he is top 5 and then streaks where he is like 20th. Which leads to overall 8-10.

 

[LandryFan]

3rd in ESPN Total Adjusted EPA
4th in Total QBR (ESPN)
4th in Air Yards/Completion (PFR)
5th in DVOA
5th in ANY/A (Football Perspective)
5th in PFN OSM
6th in ANY/A (PFR)
7th in CPOE (Next Gen)
7th in EPA/play (Baldwin/NFLscrapR)
8th-lowest in Bad Throw %
9th in PFF offense grade
9th in On-Target %
10th in NFL passer rating
12th in CPOE (Baldwin/NFLscrapR)

NextGen's CPOE is based on target depth, receiver separation, and pressure on QB
Baldwin's CPOE is based on target depth only

FO hasn't updated DVOA yet, but I saw a twitter post (which I can't find now) with the final rankings that had Prescott 5th.

I generally left out things that are already included in other things (yards, completion percentage, yards per attempt, TD/INT ratio are all included in passer rating... adjusted completion percentage and average depth of target are already included in CPOE, etc.)

Depending on which source you want to believe, Prescott's receivers dropped 6.9% (PFF) of his catchable passes, which was the most in the NFL, or 5.7% (PFR) which was 5th-highest.

As for which stats are the best, it depends on how you measure that. Passer rating, ANY/A, and EPA all have strong win correlations, but don't concern themselves with how to divide credit/blame. Total QBR, PFF grade, and PFN's OSM all attempt to divide credit/blame.

The bottom line for objectivity and win correlation would be EPA/play, but there are other stats that predict future EPA better than EPA itself. Metrics with varying degrees of subjectivity (DVOA, PFF grade, and SIS Total Points) are even better at predicting future EPA because they include credit/blame considerations. (SIS doesn't have their 2019 total points in yet, but Dak has been Top 5-7 most of the season). Despite it's high correlation with wins, ESPN's Total QBR is not very good at predicting future EPA.

PFF's player grading system has evidently come a long way since they started. Based on my early experiences with them, my only involvement with PFF grades has been either to criticize them or tell people to ignore them. But PFF's grades' ability to predict future EPA better than EPA itself shows that there is an advantage to the subjectivity involved in assigning credit and blame, if you're doing it right. Based on the information below, I have to admit that the people at PFF know what they're doing.

Summary: You probably can't go wrong by taking Dak's average rank from the top 8 metrics you see below.

I appreciate that you took time to put this together and post it. Thanks!

 

[percyhoward]

You do understand that all QBs over a large enough sample will play worse against the top 30% of the league than they do against the bottom 70%, right?

Yes, it's normal for a QB (especially a young QB) to struggle against great pass defenses. Aikman's 3rd-5th seasons:

First 6 games vs. Top 5 Pass Defenses (all 1991-92)
105 of 187 (56%) 1243 yd (6.6 ypa) 8 td 10 int 68.5

1991-93 vs. #6-28 Pass Defenses
755 of 1096 (69%) 8528 yd (7.8 ypa) 50 td 19 int 99.9

In his first six games against a defense that finished the season ranked in the Top 5 in passer rating, his passer rating was 31.4 points lower than it was in his other games. He didn't turn it around against elite defenses until the '93 playoffs.

 

[blueblood70]

3rd in ESPN Total Adjusted EPA
4th in Total QBR (ESPN)
4th in Air Yards/Completion (PFR)
5th in DVOA
5th in ANY/A (Football Perspective)
5th in PFN OSM
6th in ANY/A (PFR)
7th in CPOE (Next Gen)
7th in EPA/play (Baldwin/NFLscrapR)
8th-lowest in Bad Throw %
9th in PFF offense grade
9th in On-Target %
10th in NFL passer rating
12th in CPOE (Baldwin/NFLscrapR)

NextGen's CPOE is based on target depth, receiver separation, and pressure on QB
Baldwin's CPOE is based on target depth only

FO hasn't updated DVOA yet, but I saw a twitter post (which I can't find now) with the final rankings that had Prescott 5th.

I generally left out things that are already included in other things (yards, completion percentage, yards per attempt, TD/INT ratio are all included in passer rating... adjusted completion percentage and average depth of target are already included in CPOE, etc.)

Depending on which source you want to believe, Prescott's receivers dropped 6.9% (PFF) of his catchable passes, which was the most in the NFL, or 5.7% (PFR) which was 5th-highest.

As for which stats are the best, it depends on how you measure that. Passer rating, ANY/A, and EPA all have strong win correlations, but don't concern themselves with how to divide credit/blame. Total QBR, PFF grade, and PFN's OSM all attempt to divide credit/blame.

The bottom line for objectivity and win correlation would be EPA/play, but there are other stats that predict future EPA better than EPA itself. Metrics with varying degrees of subjectivity (DVOA, PFF grade, and SIS Total Points) are even better at predicting future EPA because they include credit/blame considerations. (SIS doesn't have their 2019 total points in yet, but Dak has been Top 5-7 most of the season). Despite it's high correlation with wins, ESPN's Total QBR is not very good at predicting future EPA.

PFF's player grading system has evidently come a long way since they started. Based on my early experiences with them, my only involvement with PFF grades has been either to criticize them or tell people to ignore them. But PFF's grades' ability to predict future EPA better than EPA itself shows that there is an advantage to the subjectivity involved in assigning credit and blame, if you're doing it right. Based on the information below, I have to admit that the people at PFF know what they're doing.

Summary: You probably can't go wrong by taking Dak's average rank from the top 8 metrics you see below.

now go remove all the sub .500 matchup games and say use NE-PHilly ONLY late int he year.. lest see how that data pays off.. I like dak but last 5 of 6 games when it was all onte line , he got worse each game and the 2nd half of the Philly game where it was ours to win with wide open receivers and time int he pocket, he was horrible..he lost us the division and a playoff game with that one half of football.. take the rest out of the equation...

hes done this in stretches in 2017 and to begin 2018, Ill throw those out he had some team excuses BUT NE-PHilly he could have won more of those and really he was the problem..

 

[zenmastersauce]

Dak was actually really really good for most of the season, but let's face it. His 2019 season will and should be judged by how he played in the biggest game of the year, the division title game against the Eagles. Dak wasn't just bad, he was career bad. Replay that game and count the throws that were off target, some by a mile. I have no idea if he was injured or not but he was bad and it cost the Cowboys the game. I draw the comparison to Wentz who was throwing accurate passes to the right receivers the last month of the season. I am a Dak supporter but he needs to show up better in big important games. If that Eagles game is any indication of how Dak will play under pressure, the Cowboys have a big problem.

I would say that his performance against the Eagles has been very good in the past. There were factors that made that game more important and thus carried more weight than many of the others. He needed to play better. He gets no excuse for injuries bc he was on the field. I do think it's not a pattern of failure in that regard yet. In pressure moments he has been pretty good.

 

[Gameover]

It was also the most injured game of the year for him.

The problem was Garrett. He should have reduced Dak's passing attempts greatly and given the ball to Zeke. The fact Dak had more than 3 times the number of attempts compared to Zeke's rushing attempts is stupid crazy.

Typical Garrett

Cowboys are 11-5 with Linehan and Dak finishes second in MVP voting

Garrett doesn’t know how to mesh the run and pass game. See 2011-2013.

 

[BoysForLife]

Typical Garrett

Cowboys are 11-5 with Linehan and Dak finishes second in MVP voting

Garrett doesn’t know how to mesh the run and pass game. See 2011-2013.

Here's the problem with this theory--
last year Dak's backers got Linehan run out of town on a rail because "SL's archaic gameplans and outdated offense" were hampering Dak's development production

Scott Linehan basically got the axe because he wasn't able to maximize Dak's "considerable talents". Poor Scott couldn't make an read option quarterback with horrific footwork, comparably bad throwing motion, and serious accuracy/timing issues look like Tom Brady. Therefore, he was shown the door.

Linehan was what we might refer to as one of the early Dak casualties.
Dez was another. Garrett is about to be another. Although a lot of JG's was self inflicted as well.