Josh Allen: Analytics

[cle23]

Josh Allen is Deshone Kizer on steroids.  Everything you could want in a QB but none of the production.

3 hours ago, cba fan said:

That shows how far the play went. NOT how far the QB threw the ball in the air.

 

My point is Allen throws the ball much much further in the air.

 

Also the following graph taken off same site your comp % was on shows Rosen and Allen are about the same. Rosen does better on 4 yrd passes and that is about it.

He's literally last in all but 1 spot, and in that he's 4th out of 5.

[Kemp]

16 hours ago, BillsFan4 said:

 

https://www.12up.com/posts/4792594-carson-wentz-is-only-an-eagle-because-of-analytics-and-the-browns

 

 

http://www.theherdnow.com/radio/the-browns-missed-on-carson-wentz-because-analytics-cant-measure-franchise-qb-traits/

 

 

Also, I think its worth mentioning that Allen and Wentz had pretty similar college stats -

 

https://www.sports-reference.com/cfb/players/carson-wentz-1.html

 

https://www.sports-reference.com/cfb/players/josh-allen-7.html

 

 

 

Thanks for the links. Interesting stuff.

 

I wonder if it's possible that Wentz isn't as good as currently believed. His team buzzed through the playoffs and won the Super Bowl without him and Foles put up better stats than Wentz under tougher circumstances. Since Foles had never shown this level of ability before, it at least makes me wonder if Wentz will have the career most assume he will have or the Eagles were just that good.

 

Just a thought.

[oldmanfan]

There are so many things that go into the relative success or failure of an NFL QB.  And there are so many things that go into how you perform in college as a college QB.  Quality of your receivers, the type of offense you run, the quality of the opposition, etc, etc, etc.  You would have to use pretty advanced multivariate statistical analysis to try and compare players because of that, so I always question the relative value of some of the articles I've seen, not only about Allen but about any college QB entering the league.

 

Allen to me is like any of the other guys drafted, in that he has work to do to make it as an NFL starter.  He may have more work to do in some areas than others, and vice versa.  Ultimately you just have to let him learn and develop and see if he makes it.  The Browns will do the same with Mayfield, the Jets with Darnold, and so on.  There is no other way to approach it.

[Wayne Arnold]

10 hours ago, cle23 said:

Josh Allen is Deshone Kizer on steroids.  Everything you could want in a QB but none of the production.

He's literally last in all but 1 spot, and in that he's 4th out of 5.

 

Didn't realize Josh Allen played for Brian Kelly at a national powerhouse surrounded by 4-5 star talents. Interesting. I can now see why so many are against him.

 

1 hour ago, Kemp said:

 

Thanks for the links. Interesting stuff.

 

I wonder if it's possible that Wentz isn't as good as currently believed. His team buzzed through the playoffs and won the Super Bowl without him and Foles put up better stats than Wentz under tougher circumstances. Since Foles had never shown this level of ability before, it at least makes me wonder if Wentz will have the career most assume he will have or the Eagles were just that good.

 

Just a thought.

 

Goes to show the importance of coaches.

[cle23]

2 hours ago, Wayne Arnold said:

 

Didn't realize Josh Allen played for Brian Kelly at a national powerhouse surrounded by 4-5 star talents. Interesting. I can now see why so many are against him.

 

 

Goes to show the importance of coaches.

Since when is ND a national powerhouse? They have a good year every 5 or 10 years and that's it. I actually like Allen, if you give him a LOT of time to sit. But accuracy isn't something that usually gets a ton better once in the NFL.

[Wayne Arnold]

33 minutes ago, cle23 said:

Since when is ND a national powerhouse? They have a good year every 5 or 10 years and that's it. I actually like Allen, if you give him a LOT of time to sit. But accuracy isn't something that usually gets a ton better once in the NFL.

 

Of course Notre Dame is a powerhouse. They have top recruiting classes every year.

 

Throwing accuracy isn't an issue for Allen.

[Hapless Bills Fan]

3 hours ago, Kemp said:

 

Thanks for the links. Interesting stuff.

 

I wonder if it's possible that Wentz isn't as good as currently believed. His team buzzed through the playoffs and won the Super Bowl without him and Foles put up better stats than Wentz under tougher circumstances. Since Foles had never shown this level of ability before, it at least makes me wonder if Wentz will have the career most assume he will have or the Eagles were just that good.

 

Just a thought.

 

That's always a fair question.  I have the impression that maybe both are true:

-Wentz is a capable, possibly good or excellent NFL QB

-DiFilippo as QB coach, Reich as OC, and Pederson as HC may just have been something special.  

 

It will be interesting to see how Wentz and Foles work next year with Reich and DiFilippo both moving on.

[Hapless Bills Fan]

3 hours ago, oldmanfan said:

There are so many things that go into the relative success or failure of an NFL QB.  And there are so many things that go into how you perform in college as a college QB.  Quality of your receivers, the type of offense you run, the quality of the opposition, etc, etc, etc.  You would have to use pretty advanced multivariate statistical analysis to try and compare players because of that, so I always question the relative value of some of the articles I've seen, not only about Allen but about any college QB entering the league.

 

Yes.  This is exactly the kind of thing I was trying to get at earlier in the thread.  If you're building a predictive model for NFL performance that's based largely on college QB who were recruited directly from HS to power-5 teams with decent to excellent talent overall on the team, playing for the most part spread offenses with simplified reads or protections coming in from the sideline, that's great - and it should work well when applied to the college QB who come from the same background.

 

But if you take a QB with a year in JUCO, running a pro-style offense on a team without NFL-draftable talent around him, does that predictive model which says something like "<60% completions, fewer than 30 starts = bust" apply?  Maybe it does.  Maybe it doesn't, because there just aren't enough QB who have that college background built into the model for it to be valid.  That's why every couple years we get a Russ Wilson or a Dak Prescott who was drafted in later rounds (because they didn't really fit the common model) and set everyone's hair on fire.

 

You would probably need to do a deep dive into college QB performance in different situational contexts vs their NFL performance and build a much more sophisticated model - deeper than Solak goes with Contextualized QBing because it doesn't take WR performance and OL performance and all into sufficient account. 

 

Now maybe the Bills and their nascent analytics department did just that, and that's why they said "yep, Allen's the guy" and traded up decisively for him.

 

Or maybe they just watched him throw in a private workout and fell in love with a guy with a good heart and a big arm.

 

Either way, there's no way now to tell just what he'll be until we see him play.  I'm not sure if I'm more irked with the hearts-and-flowers crowd who want to totally discount the problems a number of analysts (and posters here) see on film, or the model-worshippers who pen absurdities like that one article claiming "every number in the history of the world" says Allen will fail (or something like that).

 

Stats aren't for losers any more than compass planes or reciprocating saws are for losers.  Like compass planes and reciprocating saws, stats are a tool and need to be understood and used and interpreted correctly.

[ColoradoBills]

4 hours ago, Kemp said:

 

Thanks for the links. Interesting stuff.

 

I wonder if it's possible that Wentz isn't as good as currently believed. His team buzzed through the playoffs and won the Super Bowl without him and Foles put up better stats than Wentz under tougher circumstances. Since Foles had never shown this level of ability before, it at least makes me wonder if Wentz will have the career most assume he will have or the Eagles were just that good.

 

Just a thought.

 

The only reasonable conclusion that one can come to is that a QB's success is a combination of things.  Analytics is a good tool to use but as the input data

gets more variables the outputs become cloudier.  Oldmanfan touched on some good points below.  Eagles have a strong system (scheme and personnel)

currently, how long they can keep it working is the question.  It's why Analytics can be more successful evaluating a baseball players batting v a NFL QB.

 

3 hours ago, oldmanfan said:

There are so many things that go into the relative success or failure of an NFL QB.  And there are so many things that go into how you perform in college as a college QB.  Quality of your receivers, the type of offense you run, the quality of the opposition, etc, etc, etc.  You would have to use pretty advanced multivariate statistical analysis to try and compare players because of that, so I always question the relative value of some of the articles I've seen, not only about Allen but about any college QB entering the league.

 

Allen to me is like any of the other guys drafted, in that he has work to do to make it as an NFL starter.  He may have more work to do in some areas than others, and vice versa.  Ultimately you just have to let him learn and develop and see if he makes it.  The Browns will do the same with Mayfield, the Jets with Darnold, and so on.  There is no other way to approach it.

 

Great post oldmanfan.  The thing I can see about Allen is this, he has a very limited exposure to the variances in QB type play.  He may be given details

that can change his game completely.  This is what I see as his biggest RAW condition.  It's one of the reasons I agree with the evaluators who want to

bring him around slowly.

[Luxy312]

Gotta love all of these useless comparisons between players.  To even say that one QB's completion percentage was better than another's is pointless.  It implies on the surface that each QBs situation is the same in terms of making the same throws to the same caliber of talent of receiver playing against the same caliber of talent of defense.  Without a control group to make a true comparison, we're left simply with generalization.  When all is said and done, the "analytics" over the last 30 years have been about equally wrong as right in terms of quarterbacks taken in the top ten in the NFL draft.  For every Cam Newton, we get a Jake Locker.  That's even saying all of the mediocre QB's are part of the 50% "right" with the likes of Ryan Tannehill, Alex Smith, Kerry Collins and Trent Dilfer.  Not terrible QB's by any stretch, but certainly not HOF bound.  Until the analytics get better in terms of predictive value, I'll remain cautious in terms of how they're used.

[Mat68]

11 minutes ago, ColoradoBills said:

 

The only reasonable conclusion that one can come to is that a QB's success is a combination of things.  Analytics is a good tool to use but as the input data

gets more variables the outputs become cloudier.  Oldmanfan touched on some good points below.  Eagles have a strong system (scheme and personnel)

currently, how long they can keep it working is the question.  It's why Analytics can be more successful evaluating a baseball players batting v a NFL QB.

 

 

Great post oldmanfan.  The thing I can see about Allen is this, he has a very limited exposure to the variances in QB type play.  He may be given details

that can change his game completely.  This is what I see as his biggest RAW condition.  It's one of the reasons I agree with the evaluators who want to

bring him around slowly.

Raw is a vauge term now.  Allen played in a pro system closer than any player.  What Allen experianced at Wyoming is closer to what he will find in the NFL vs what Mayfield did at Oklahoma.  

[Luka]

It's a team sport. Tons of QBs have busted, but is it always all on the player? If you're picked high in the 1st round, you're usually going to a bad team with a new staff. Or if you're unfortunate enough to be picked by the Browns, you're going to a team that's won a single game in 2 years and has decided that their coaching staff is good enough. There are entirely too many factors to sit down and say this stat, this stat and this stat is what matters. Especially since none of these statistics are obtained in vacuum. No two situations are alike.

[HappyDays]

25 minutes ago, Luxy312 said:

Gotta love all of these useless comparisons between players.  To even say that one QB's completion percentage was better than another's is pointless

 

Where was completion percentage mentioned?

[ColoradoBills]

3 minutes ago, Mat68 said:

Raw is a vauge term now.  Allen played in a pro system closer than any player.  What Allen experianced at Wyoming is closer to what he will find in the NFL vs what Mayfield did at Oklahoma.  

 

Raw is a vague term.  Allen's history of High School, JUCO and let's say his first injured year at Wyoming IMO put him way behind the "normal" curve

of QB development.  Having him play in a more Pro style of offense is a help but most who evaluated that offense also admit it wasn't very successful.

The OL, WR's and RB were lacking in talent and execution.  Allen's QB play was affected, the issue is how much was on him and how much was on

the rest of the team.  There is a lot of games stat's showing Allen throwing a very low number of passes.

 

My point is, that is all he was exposed to for only a short period of time (20 some games).  Some evaluators may have "pigeonedholed" him to soon.

[Luxy312]

16 minutes ago, HappyDays said:

 

Where was completion percentage mentioned?

 

Did you even try to read the thread.  It's on this page.  LOL.  Terrible laziness on your part.

[Kemp]

4 hours ago, Wayne Arnold said:

 

 

 

2 hours ago, cle23 said:

Since when is ND a national powerhouse? They have a good year every 5 or 10 years and that's it. I actually like Allen, if you give him a LOT of time to sit. But accuracy isn't something that usually gets a ton better once in the NFL.

 

When has accuracy gotten a ton better from college to the NFL?i

[Wayne Arnold]

1 minute ago, Kemp said:

When has accuracy gotten a ton better from college to the NFL?i

 

Define "a ton".

[jrober38]

Everything I've read over the past couple weeks tells me there's no difference between Josh Allen and Christian Hackenberg, who just announced he's about to completely rebuild his mechanics this offseason as he goes into his THIRD year in the NFL. 

 

Both are big, strong, mobile QBs with great character, who were plagued by ineffectiveness and accuracy issues in college. 

 

Both worked with Jordan Palmer leading up to the draft. If you read up on what Palmer said about Hackenberg's mechanics, it's pretty much verbatim what he said about Josh Allen for the past two months. 

 

Hopefully the result is different for us than it was for the Jets. 

[Wayne Arnold]

2 minutes ago, jrober38 said:

Everything I've read over the past couple weeks tells me there's no difference between Josh Allen and Christian Hackenberg, who just announced he's about to completely rebuild his mechanics this offseason as he goes into his THIRD year in the NFL. 

 

Both are big, strong, mobile QBs with great character, who were plagued by ineffectiveness and accuracy issues in college. 

 

Both worked with Jordan Palmer leading up to the draft. If you read up on what Palmer said about Hackenberg's mechanics, it's pretty much verbatim what he said about Josh Allen for the past two months. 

 

Hopefully the result is different for us than it was for the Jets. 

 

Absurd.

[jrober38]

5 minutes ago, Wayne Arnold said:

 

Absurd.

 

In what way?