What is Josh Allen worth in wins and losses?

[Bferra13]

 

4 hours ago, wjag said:

Josh is worth money. 

[Big Turk]

Well according to EPA, Allen is worth about a touchdown per game.

[DCofNC]

This is interesting, I have thought about this without real data and kind of concluded it would look similar to this.  My thought was, the difference between Allen and Mahomes for wins would be likely zero, Allen to the other “elite” QBs would be 1-2 more wins, Allen to the “good” QBs would be 2-4 wins and then Allen to the bottom half guys could range from 5-10 wins depending on just how bad they really are.   For example, I think you could put Allen on the Broncos this year and add 4 wins, but stick him on the Raiders and he makes that team 10 games better.  Yes I genuinely believe the Raiders could be THAT much better with him.   Now the Giants.. woof.. they’d win more, but that roster is a dumpster fire that could hold anyone back.  Hate to say it, but they should fire the whole front office and probably send Daboll with them.

[MJS]

High QBR does not separate the QB impact from the rest of the team, the culture, player development, etc. Having good players to distribute the ball to, a good oline to protect him, a good scheme to operate in, good play calling, and a good culture of winning and growth are all going to positively impact QBR.

 

The QB is the most important position in sports. He has the highest impact on wins and losses for sure. But it is also true that he can't do it alone. I'm not sure there is a great way to separate those impacts since they all play together, especially the coach and culture.

 

But for the players, you could look at pro bowl and all pro players that they play with. That might be a way to show which QBs have to overcome more of a talent gap from the rest of the team and which benefit from dominant players at other positions. It still isn't great, but it is a start.

[Mikie2times]

1 minute ago, MJS said:

High QBR does not separate the QB impact from the rest of the team, the culture, player development, etc. Having good players to distribute the ball to, a good oline to protect him, a good scheme to operate in, good play calling, and a good culture of winning and growth are all going to positively impact QBR.

 

The QB is the most important position in sports. He has the highest impact on wins and losses for sure. But it is also true that he can't do it alone. I'm not sure there is a great way to separate those impacts since they all play together, especially the coach and culture.

 

But for the players, you could look at pro bowl and all pro players that they play with. That might be a way to show which QBs have to overcome more of a talent gap from the rest of the team and which benefit from dominant players at other positions. It still isn't great, but it is a start.

I planned on looking at pro bowlers players and all pros eventually. 
 

I don’t necessarily disagree with you, but in the majority of examples with top performing QB’s it would be my expectation that they would still be in that group on other teams. Brady to Tampa as an example. Manning to Denver, etc. Using Allen, my opinion would be he is not declining exponentially on another team. Now does Ryan Tannahill? Sure. That is a much more clear example of a QB who is highly dependent on roster, scheme, process, etc as far as how he is able to rank high. But Josh is going to give you a 65 rating regardless of process, culture, roster (within reason) and that 65 is going to get you to being on average, a fringe playoff team. 

[T master]

4 consecutive AFC East Championships ...

[Mr. WEO]

20 hours ago, djp14150 said:

im not saying the NFL passer rating is the best, it’s not because it penalizes mobile/ running QBs

 

at the same time QBR produces utterly nuts ratings.  It penalizes QBs for sacks but how much is the sack the QBs fault?

 

 

How many incomplete passes are the QB fault?  doesn't matter--it's applied equally for all QBs. 

 

[Since1981]

1 minute ago, T master said:

 

4 consecutive AFC East Championships ...

 

Many people have been made to look good after Beane took a tough, calculated risk on an inaccurate passer. 

[chris heff]

$258,034,000

[Beck Water]

23 hours ago, Mikie2times said:

I have seen a ton of posts talk about culture, it taking a village, and other concepts to explain why Buffalo has been so successful over the last few years. Maybe that stuff is true, maybe it's not, I really have no idea how to quantify these concepts. What we can do is look at high level QB play over a period of time and see what type of impact that had on the teams they played for. So that's exactly what I did. 

 

I used QBR to split the groupings up. https://en.wikipedia.org/wiki/Total_quarterback_rating not a perfect system by any stretch, but much better at accounting for total production (like running). 

 

• Group one is QB's that finished the season 70 or higher and played a minimum of 10 games
• Group two is QB's that finished the season between 65-69.99 and played a minimum of 10 game

 

In most seasons these groupings would yield between 8-10 total players, 2-4 in the first group and about 5-6 in the lower tier group. 

 

You can see the results below, 1 PLA and 2 PLA is just the total players in each group that year. 

 

Teams with a QB that finished the year with a 70 or better QBR are 582-208 since 2010. Good for a 73.67% winning percentage.

Teams with a QB that finished the year with a 65-69.99 QBR are 475-288 since 2010. Good for a 62.25% winning percentage.

 

 

Josh Allen has had the following QBR's

 

2020: 76.6 

2021:  66.3

2022: 73.4

2023: 69.6

2024: 76.7

 

So three years he has been in group 1 and three years he has been in group 2. His record in that time is 61-21,  1% above the average record of Group 1 players. 

 

I intend on adding some additional data to this such as playoff appearances and results, but for now this is what I have. 

 

  

 

I love a data nerd, so Kudos for putting this together, but I think "total QBR" has been pretty well debunked as a metric of QB play, much less a QB value to the team in terms of W and L, so I don't think your analysis really says too much 

 

Let's just give it a couple eyeball checks.  

 

In 2022, the Bengals played the Chiefs for the AFC Championship, the Eagles played the 49ers for the NFC championship, and the Giants and Seahawks snuck into the playoffs at 9-7-1

total QBR has Mahomes and Allen at the top of the list, but then places Tua, Daniel Jones, Geno Smith, and Jacoby Brissett ahead of Joe Burrow in total QBR.  Does that pass the eyeball test to you for a metric determining a QB's value to his team?  Really, Daniel Jones and Jacoby Brissett of more value to their team than Joe Burrow?

 

In 2023, Brock Purdy and Dak Prescott top the QBR list and Herbert, Stafford, and Cousins all come out ahead of Mahomes.  Again, common sense eyeball test?

 

I'd love to look under the hood and understand how QBR is weighting different things, but I can't; QBR is calculated by a proprietary algorithm using 10,000 lines of code.  ESPN has never released the complete formulas and procedures to calculate QBR

And of course, there's this famous example:

Quote

Further controversy erupted when the Total QBR system gave the Denver Broncos' Tim Tebow a higher rating than the Green Bay Packers' Aaron Rodgers in their respective Week 5 contests in 2011. Noting that Rodgers completed 26 of 39 passes for 396 yards and two touchdowns in a win over the Atlanta Falcons, while Tebow completed four of 10 passes for 79 yards and a touchdown, and six rushes for 38 yards and a touchdown, in a loss to the San Diego Chargers. In a more recent example, a game played on September 24, 2017, Alex Smith of Kansas City Chiefs received an inexplicable QBR of 7.8, half as much as the equally-bad QBR of 16.1 for his counterpart Philip Rivers of the Los Angeles Chargers, even though Smith had a higher completion rate (16/21 vs. 20/40), a better average per completion (7.8 yds vs. 5.9), a far superior TD/int ratio (2-0 vs. 0-3), and won the game handily 24-10. For comparison, the RTG, 128.1 for Smith and 37.2 for Rivers, was by far a better metric of success.

 

So I think you have a great idea of trying to use some form of objective metric, but I think there's reason to believe that total QBR may not be a very good choice.

[Mikie2times]

1 minute ago, Beck Water said:

 

I love a data nerd, so Kudos for putting this together, but I think "total QBR" has been pretty well debunked as a metric of QB play, much less a QB value to the team in terms of W and L, so I don't think your analysis really says too much 

 

Let's just give it a couple eyeball checks.  

 

In 2022, the Bengals played the Chiefs for the AFC Championship, the Eagles played the 49ers for the NFC championship, and the Giants and Seahawks snuck into the playoffs at 9-7-1

total QBR has Mahomes and Allen at the top of the list, but then places Tua, Daniel Jones, Geno Smith, and Jacoby Brissett ahead of Joe Burrow in total QBR.  Does that pass the eyeball test to you for a metric determining a QB's value to his team?  Really, Daniel Jones and Jacoby Brissett of more value to their team than Joe Burrow?

 

In 2023, Brock Purdy and Dak Prescott top the QBR list and Herbert, Stafford, and Cousins all come out ahead of Mahomes.  Again, common sense eyeball test?

 

I'd love to look under the hood and understand how QBR is weighting different things, but I can't; QBR is calculated by a proprietary algorithm using 10,000 lines of code.  ESPN has never released the complete formulas and procedures to calculate QBR

And of course, there's this famous example:

 

So I think you have a great idea of trying to use some form of objective metric, but I think there's reason to believe that total QBR may not be a very good choice.

It's not a deceleration for QBR, but of the readily available data points that is best. If not that, what would it be from your view (Again, readily available). Pure rating doesn't even include rushing outputs? This is pro football reference historical data. Ultimately the data is good enough to show if you have a good performing QB you will win a lot. Not rocket science, but at least some data to support that statement. 

[MJS]

1 hour ago, Mikie2times said:

I planned on looking at pro bowlers players and all pros eventually. 
 

I don’t necessarily disagree with you, but in the majority of examples with top performing QB’s it would be my expectation that they would still be in that group on other teams. Brady to Tampa as an example. Manning to Denver, etc. Using Allen, my opinion would be he is not declining exponentially on another team. Now does Ryan Tannahill? Sure. That is a much more clear example of a QB who is highly dependent on roster, scheme, process, etc as far as how he is able to rank high. But Josh is going to give you a 65 rating regardless of process, culture, roster (within reason) and that 65 is going to get you to being on average, a fringe playoff team. 

Tom Brady's last year in New England was pretty paltry as the offensive roster around him was pretty bad. I was going to use that as an example, actually. He went to Tampa Bay and won another Superbowl because they surrounded him with a ton of talent.

40 minutes ago, Mikie2times said:

It's not a deceleration for QBR, but of the readily available data points that is best. If not that, what would it be from your view (Again, readily available). Pure rating doesn't even include rushing outputs? This is pro football reference historical data. Ultimately the data is good enough to show if you have a good performing QB you will win a lot. Not rocket science, but at least some data to support that statement. 

I like PFR Aproximate Value. You could look at that as a second option.

Edited Wednesday at 04:56 PM by MJS

[Mikie2times]

24 minutes ago, MJS said:

Tom Brady's last year in New England was pretty paltry as the offensive roster around him was pretty bad. I was going to use that as an example, actually. He went to Tampa Bay and won another Superbowl because they surrounded him with a ton of talent.

I like PFR Aproximate Value. You could look at that as a second option.

Brady was probably not the best example, I will agree with that. I think AV has just as much likelihood, if not more, to pull in factors outside of what the individual player contributes. As they say, perfection is the enemy of progress. I think this information is helpful in progressing to the most accurate place we can get in understanding how QB play impacts the bottom line. Perfect, it is not. 

Edited Wednesday at 05:18 PM by Mikie2times

[Rochesterfan]

21 hours ago, Mikie2times said:

My takeaway, is Josh Allen = playoffs expectations.

 

Void of nearly everything else.

 

Is it possible to still not make the playoffs with a top 10 performing QBR? Sure. Rare, but it happens. Even more rare when that QB crosses 70 QBR. Even more so when you look at the declining number of players hitting these levels recently. Only 2 players even above 65 last year and 3 players the year prior.

 

I have long made the argument that this franchises success and stability are from Allen more than anything else and I would triple down on that thought process   

 

That is only if you believe that McD has not influenced that.  
 

McD as a coach talking with Josh and working with him and the OC to not only come up with plans, but what to work on and plans for growth have allowed Josh to develop.

 

Without the success and stability of a head coach, you get Aaron Rodgers of 2024 and NYJ Sam Darnold versus Minnesota Sam Darnold.  You get the Bears blowing multiple high draft pick QBs not because of talent, but the change and lack of stability cause the QB failure.

 

Josh is Josh because McD allowed him to grow and foster.  
 

My belief is that Josh’s success is a direct reflection of his relationship with Beane and McD and their trust and attitude in how to build a franchise allowed Josh to grow and become a star and you do not get one without the other.

 

 

[Fan in Chicago]

21 hours ago, CincyBillsFan said:

I have a lot of experience in data analysis and football is not a good place to apply it.  Sure, you can get some insight but for something like this it's a significant overreach.  After all how do you factor this into the calculations:

 

 

I whole heartedly agree. He is a prime example of stats not telling the full story. 

[Sammy Watkins' Rib]

6 minutes ago, Fan in Chicago said:

I whole heartedly agree. He is a prime example of stats not telling the full story. 

 

Yep. The best example this season of stats not telling the full story is probably the 4th and 2 game against the Chiefs to end that game. How many "points" does traditional passer rating assign to that play for Josh Allen? That would be zero. And however many points Total QBR assigned to it is probably still not even enough. Play of the year as Jim Nantz rightfully called it. 

[Rochesterfan]

16 hours ago, GunnerBill said:

 

Of course they are. Franchise QB always  matters more than anything else. That isn't revelationary. It's just fact. But that doesn't get you to where you think it does on coaching.

 

Totally agree - you have franchise QBs where the expectations are SB, but with poor coaching - you get Cincinnati of this year on the outside looking in.  
 

 

[Mikie2times]

16 minutes ago, Rochesterfan said:

 

That is only if you believe that McD has not influenced that.  
 

McD as a coach talking with Josh and working with him and the OC to not only come up with plans, but what to work on and plans for growth have allowed Josh to develop.

 

Without the success and stability of a head coach, you get Aaron Rodgers of 2024 and NYJ Sam Darnold versus Minnesota Sam Darnold.  You get the Bears blowing multiple high draft pick QBs not because of talent, but the change and lack of stability cause the QB failure.

 

Josh is Josh because McD allowed him to grow and foster.  
 

My belief is that Josh’s success is a direct reflection of his relationship with Beane and McD and their trust and attitude in how to build a franchise allowed Josh to grow and become a star and you do not get one without the other.

 

 

Without a hall of fame QB you likely don’t have success and stability at head coach. One thing I guarantee is it is a hell of a lot easier to draw the line to a QB as a reason a team is successful than a head coach. I don’t think one example exists where a great HC hasn’t had a great QB. Plenty exist the other way. 

3 minutes ago, Rochesterfan said:

 

Totally agree - you have franchise QBs where the expectations are SB, but with poor coaching - you get Cincinnati of this year on the outside looking in.  
 

 

Except Cincinnati this year is a massive anomaly in this data compared to the entire sample 

[HappyDays]

40 minutes ago, Sammy Watkins' Rib said:

The best example this season of stats not telling the full story is probably the 4th and 2 game against the Chiefs to end that game. How many "points" does traditional passer rating assign to that play for Josh Allen? That would be zero. And however many points Total QBR assigned to it is probably still not even enough.

 

Actually QBR likely graded that play extremely high because it adjusts for opponent quality, down/distance, and game situation (score differential, time left in the game, etc.), and it includes QB scrambles as part of the calculation. So I would bet that play approaches the ceiling of what a single play could calculate out to in QBR.

[FireChans]

38 minutes ago, Mikie2times said:

Without a hall of fame QB you likely don’t have success and stability at head coach. One thing I guarantee is it is a hell of a lot easier to draw the line to a QB as a reason a team is successful than a head coach. I don’t think one example exists where a great HC hasn’t had a great QB. Plenty exist the other way. 

Except Cincinnati this year is a massive anomaly in this data compared to the entire sample 

John Fox with Peyton Manning:

38-10

 

John Fox with everyone else:

 

95-113

 

He must’ve installed an INCREDIBLE culture those 3 years in Denver. Luckily, it coincided with Peyton Manning coming to town. 

Edited Wednesday at 06:23 PM by FireChans