[Vague Title] It continues... Josh Allen...

[BillsMafia13]

19 hours ago, Bangarang said:

 

Didn’t Allen miss more games in college due to injury than Rosen?

 

And why do people bring up his dad’s profession as some sort of negative against him? If he’s that well off that he can just walk away at any point tell you how bad he wants it since he’s actually putting in the work to try and be great?

 Again, im not saying any of this is a driving reason, just throwing it out there.  Maybe long term health is  more of a factor in the Rosen household than the farm Allen came from.  Playing in the NFL is a 1 and a million opportunity, Chris Borland quit after 1 season.  Money Manziels family has a ton of money, he doesnt need the NFL either. These little facts make me think Allen was the right pick here

[Hapless Bills Fan]

[This is an automated response]

 

As a courtesy to the other board members, please use more descriptive topic titles. A better title will help the community find information faster and make your topic more likely to be read. The topic starter can edit the topic title line to make it more appropriate.

 

Thank you.

[Bangarang]

16 minutes ago, BillsMafia13 said:

 Again, im not saying any of this is a driving reason, just throwing it out there.  Maybe long term health is  more of a factor in the Rosen household than the farm Allen came from.  Playing in the NFL is a 1 and a million opportunity, Chris Borland quit after 1 season.  Money Manziels family has a ton of money, he doesnt need the NFL either. These little facts make me think Allen was the right pick here

 

I think you’re grasping at straws here but you’re entitled to that opinion.

[HappyDays]

3 hours ago, Sky Diver said:

QBASE gave Wentz a 61.9% chance of being a bust? How well did that work out?

 

You would have to look at every QB they've given a 62% bust probability to, and see if around 62% of them busted. If 100% of those QBs busted, that would actually mean the model was broken. This is the biggest misconception about QBASE and other analytics tools. QBASE also gives Allen a 62% chance of busting. That really isn't that high. In Derek Jeter's career he struck out about 69% of the time. So according to QBASE Josh Allen has a better chance of being at least an adequate starter than Jeter had of getting a hit. Just to make a comparison.

 

And to be fair a lot of these misconceptions happen because the Football Outsider writers overstate their case. I don't know why they're laughing off the possibility that Allen could succeed. Their own model says he has a 38% chance of being a starting level QB. That isn't a particularly low probability.

[Sky Diver]

55 minutes ago, Bangarang said:

 

Stop trying to say it’s flawed because it may not be right 100% of the time. If stats were a guaranteed method of predicting an outcome then scouts would be taken off the road immediately. They are simply another tool for evaluation. 

 

Would you be singing the same tune if the numbers lined up with what you want to believe?

 

The model obviously has limitations as it gave a high probability that Wentz, Watson, and others would be busts, and obviously that wasn't the case.

 

Are statistics useful? Of course. Should they be relied upon exclusively? Of course not. Is Allen destined to be a bust because of QBASE's model. Of course, not.

52 minutes ago, BillsMafia13 said:

 Again, im not saying any of this is a driving reason, just throwing it out there.  Maybe long term health is  more of a factor in the Rosen household than the farm Allen came from.  Playing in the NFL is a 1 and a million opportunity, Chris Borland quit after 1 season.  Money Manziels family has a ton of money, he doesnt need the NFL either. These little facts make me think Allen was the right pick here

 

I doubt injury risk is in QBASE's model. A player is only valuable when he's on the field.

17 minutes ago, HappyDays said:

 

You would have to look at every QB they've given a 62% bust probability to, and see if around 62% of them busted. If 100% of those QBs busted, that would actually mean the model was broken. This is the biggest misconception about QBASE and other analytics tools. QBASE also gives Allen a 62% chance of busting. That really isn't that high. In Derek Jeter's career he struck out about 69% of the time. So according to QBASE Josh Allen has a better chance of being at least an adequate starter than Jeter had of getting a hit. Just to make a comparison.

 

And to be fair a lot of these misconceptions happen because the Football Outsider writers overstate their case. I don't know why they're laughing off the possibility that Allen could succeed. Their own model says he has a 38% chance of being a starting level QB. That isn't a particularly low probability.

 

The Jeter comparison doesn't make much sense, and he failed to get a hit 69% of the time, he didn't strike out 69% of the time.

[BillsMafia13]

1 hour ago, Bangarang said:

 

I think you’re grasping at straws here but you’re entitled to that opinion.

Just throwing those out there, theres a million other factors that have to be evaluated too. I also thought we'd suck last year but McBeane broke the curse.  Its nice to follow blindly and sometimes be rewarded 

[Wayne Arnold]

Allen could be a bust.

 

He could be great.

 

He could be somewhere in between.

 

/thread

[Mikey152]

My problem with Analytics, as used in football, can be summed up very concisely with this article...

 

"Old School" vs. "New School":

 

JT Barrett completed 64.7% of his passes for 35 TDs and 8 INTs in the Big 10 this year.  Those are better numbers than anybody but Mayfield w/ regards to the big 4, and that doesn't even include his rushing stats.  On paper, he should have been a top choice in the draft this year...and yet it was his physical tools and game tape that kept him from getting drafted anywhere near the top of the draft.  Don't even get me started on Guys like Mike White and Luke Falk.

 

This, in a nutshell, is the problem.  All these indicators, like comp%?  They only appear to be "predictive" once you have already evaluated a prospect by more traditional means, like size, arm strength, level of comp, scheme, etc...so for them to in turn be so cavalier about the value of traditional scouting is hilarious.  

 

Also, call me crazy but I don't think I need analytics to know that most guys that had bad numbers in college probably won't have good numbers in the pros....but to act like there can't be justfiable rationale for it to occur based on a flawed statistical sample?  That's bastardizing stats.

[oldmanfan]

2 hours ago, Bangarang said:

 

Stop trying to say it’s flawed because it may not be right 100% of the time. If stats were a guaranteed method of predicting an outcome then scouts would be taken off the road immediately. They are simply another tool for evaluation. 

 

Would you be singing the same tune if the numbers lined up with what you want to believe?

That door swings both ways

[Figster]

1 hour ago, Mikey152 said:

My problem with Analytics, as used in football, can be summed up very concisely with this article...

 

"Old School" vs. "New School":

 

JT Barrett completed 64.7% of his passes for 35 TDs and 8 INTs in the Big 10 this year.  Those are better numbers than anybody but Mayfield w/ regards to the big 4, and that doesn't even include his rushing stats.  On paper, he should have been a top choice in the draft this year...and yet it was his physical tools and game tape that kept him from getting drafted anywhere near the top of the draft.  Don't even get me started on Guys like Mike White and Luke Falk.

 

This, in a nutshell, is the problem.  All these indicators, like comp%?  They only appear to be "predictive" once you have already evaluated a prospect by more traditional means, like size, arm strength, level of comp, scheme, etc...so for them to in turn be so cavalier about the value of traditional scouting is hilarious.  

 

Also, call me crazy but I don't think I need analytics to know that most guys that had bad numbers in college probably won't have good numbers in the pros....but to act like there can't be justfiable rationale for it to occur based on a flawed statistical sample?  That's bastardizing stats.

Old school vs new school, good summary, I like it. 

 

The NFL is an ever evolving, ever changing league IMO. Cam Newton with his low 19 wonderlic score and running skills wasn't supposed to succeed in the NFL. Russell Wilson because of a height disadvantage faced the same kind of scrutinty. Our very own T Mobile was a long shot to ever becoming a starter in the NFL and just took Buffalo to the playoffs for the first time in 17 years.

 

The mold at QB position is made to be broken... 

 

 

Edited July 25, 2018 by Figster

[mjt328]

On 7/24/2018 at 1:39 PM, Cornette's Commentary said:

It was said that Pat Shurmur himself had Allen rated as the #1 QB of this class.

 

Lots of people had Josh Allen ranked #1, or right around that spot. 

Just look in the Top 10:

 

Cleveland allegedly was debating for months between Allen and Baker Mayfield.

Like you mentioned, the Giants/Pat Shurmur have said Allen would have been their top choice if they decided to go QB.

Many have reported the Cardinals being "heartbroken" on settling for Josh Rosen instead.

We know what the Bills thought.

 

[Zerovoltz]

4 hours ago, Sky Diver said:

 

I don't know what their model is or what factors they might be missing.

 

As far as weather, it's certainly a factor. It's easier to throw a football in Hawaii than Alaska from Sept to December, for example.

 

Wentz played in ND. Pretty nasty there in winter. Maybe they aren't weighting weather enough or it's not even in their model?

 

...um...North Dakota State plays in an indoor stadium.  

 

The argument about Allen is this:  Given his physical tools and mental makeup, either A) he really does suck because we have a BIG SAMPLE size that says he does. or B) He has sucked so far because in his youth and college years, he never was around good coaching or other players that would push his development along and he won't suck anymore after he gets good coaching and has teammates and opponents who are better quality.

 

That's it.  I love metrics...I think metrics are really useful.....but it doesn't take advanced metrics to tell any reasonable person that Josh Allen's actual performance in football has been poor.  That includes the Bills staff.  

 

The Bills staff believes the answer is B....because you don't take him if you think it's A....and you also don't take him even if you believe it's B, but you aren't sure you can fix him.  They think he's undercoaced and underdeveloped.  

 

I don't think they can do it.  I would only say, in this case, where you may actually have a guy with a ton of talent, who just hasn't been developed properly for years...that could be something the analytics would miss because the analytics are analyzing high level football players, and makes no assumptions where they came from or how much football they played where, or against who.  This is where you'd not rely on a number spit out by a formula...but a human judgement.  Again..I don't think it will end well, but I can see a case here about why the analytic numbers may not apply.

[Figster]

Just now, mead107 said:

Let’s move forward.   It’s training camp time. 

 

Everything has has been hashed and rehashed 

about Allen.  

 

Time to forget about this thread.  

He is our pick and draft is over. 

 

Just my opinion.  

true that, 

 

Josh Allen is a project in my humble opinion.  A J McCarron on the other hand is not.

 

Despite the question mark when it comes to a franchise QB I really like what the Buffalo Bills have done in an effort to upgrade the position. 

[The Red King]

Good lord.  He might be a bust, he might be a star.  We won't know until the bullets fly.  I want a good QB.  Not just a QB with good stats.  A good QB, one that wins games.  McBeane saw the same stats we did, the same scouting reports, the same metrics, and they drafted Allen anyway.  They obviously think they can tweak what needs tweaked, and that he can hit his huge potential.  Time and again last season, the Bills made moves that had us screaming and questioning, but what did it get us?  A playoff birth, and a buttload of draft picks.  In my eyes that earns them the benefit of the doubt.  There has been a plan here, and calculated moves with benefits that weren't immediately obvious.  That, and I'm a damn Bills' fan.  All together, that means I'm going to support the kid and see what they can do with him.  Am I saying he can't fail?  Not at all.  I just see no reason to automatically assume McBeane has no idea what they're doing and condemn the young QB to the trash heap before he takes a single snap.

Edited July 25, 2018 by The Red King

[Shaw66]

27 minutes ago, Zerovotlz said:

 

...um...North Dakota State plays in an indoor stadium.  

 

The argument about Allen is this:  Given his physical tools and mental makeup, either A) he really does suck because we have a BIG SAMPLE size that says he does. or B) He has sucked so far because in his youth and college years, he never was around good coaching or other players that would push his development along and he won't suck anymore after he gets good coaching and has teammates and opponents who are better quality.

 

That's it.  I love metrics...I think metrics are really useful.....but it doesn't take advanced metrics to tell any reasonable person that Josh Allen's actual performance in football has been poor.  That includes the Bills staff.  

 

The Bills staff believes the answer is B....because you don't take him if you think it's A....and you also don't take him even if you believe it's B, but you aren't sure you can fix him.  They think he's undercoaced and underdeveloped.  

 

I don't think they can do it.  I would only say, in this case, where you may actually have a guy with a ton of talent, who just hasn't been developed properly for years...that could be something the analytics would miss because the analytics are analyzing high level football players, and makes no assumptions where they came from or how much football they played where, or against who.  This is where you'd not rely on a number spit out by a formula...but a human judgement.  Again..I don't think it will end well, but I can see a case here about why the analytic numbers may not apply.

Thanks.  This a really good nutshell summary.

 

I tend to think the Bills CAN do it, can teachAllen what he needs to learn.   Why do I think so?   Because the way this front office and coach operate, they've evaluated the guy's work ethic, his understanding of the game, his intelligence.  They've done that with testing and through interviews with the player and his coaches.  They analyzed kind of a learner he is.  

 

Remember, McBeane saw Cam Newton come into the league and learn the things he had to learn.   He played in a pretty simple college offense, and he had a lot to learn.   He's been learning for years.   So McBeane knows what it takes, they know what kind of learner Cam was and they know what kind of learner Allen is.   

 

I think when they took Allen, they had a high level of confidence that they can teach him what he needs to know.   And that confidence was based on actual investigation into the guy.  

[Sky Diver]

1 hour ago, Zerovotlz said:

 

...um...North Dakota State plays in an indoor stadium.  

 

The argument about Allen is this:  Given his physical tools and mental makeup, either A) he really does suck because we have a BIG SAMPLE size that says he does. or B) He has sucked so far because in his youth and college years, he never was around good coaching or other players that would push his development along and he won't suck anymore after he gets good coaching and has teammates and opponents who are better quality.

 

That's it.  I love metrics...I think metrics are really useful.....but it doesn't take advanced metrics to tell any reasonable person that Josh Allen's actual performance in football has been poor.  That includes the Bills staff.  

 

The Bills staff believes the answer is B....because you don't take him if you think it's A....and you also don't take him even if you believe it's B, but you aren't sure you can fix him.  They think he's undercoaced and underdeveloped.  

 

I don't think they can do it.  I would only say, in this case, where you may actually have a guy with a ton of talent, who just hasn't been developed properly for years...that could be something the analytics would miss because the analytics are analyzing high level football players, and makes no assumptions where they came from or how much football they played where, or against who.  This is where you'd not rely on a number spit out by a formula...but a human judgement.  Again..I don't think it will end well, but I can see a case here about why the analytic numbers may not apply.

 

Um, and how many of the other 50% of their games are indoors?

 

Or C, he doesn’t suck and is a tremendous talent worthy of a top pick. Check his QB rating.

Edited July 25, 2018 by Sky Diver

[Hapless Bills Fan]

4 hours ago, HappyDays said:

And to be fair a lot of these misconceptions happen because the Football Outsider writers overstate their case. I don't know why they're laughing off the possibility that Allen could succeed. Their own model says he has a 38% chance of being a starting level QB. That isn't a particularly low probability.

 

To go further, here's the FO 2018 QBase data.  Fundamentally, their algorithm loves Mayfield and gives him almost equal probability of being elite as being a bust.

2nd favorite is Rosen. Darnold and Allen are not that dissimilar in their analysis-they give both >50% chance of being a bust and 20-30% chance of being an adequate starter.

 

They say themselves " It's important not to overestimate the importance of a small difference in the QBASE projections. Most of this year's top quarterbacks are grouped together in the middle of the scale. It's a bit of a surprise that Sam Darnold came out with a lower QBASE rating than some of the other top prospects -- Josh Rosen, for example. But if I were the analytics director for an NFL franchise, I would feel no need to disagree with a scouting director who placed Darnold ahead of Rosen."

 

There's a similar difference between Rosen and Darnold as between Darnold and Allen, but for some reason FO is all up in dissing on Allen ("horrifying") while exculpating Darnold's performance in their model.  Of more significance to my mind is their commentary that "QBASE is meant to only be used on players chosen in the top 100 picks; after that, the judgment of scouts becomes even more important, and statistics become even less predictive."  Hold the mustard here - what they probably mean by that is that their model is more valid for prospects from major programs facing a similar level of competition, and doesn't do so well with players outside that model.

 

I saw you palm that card, Football Outsiders.

 

Here's their data:

 

Darnold

Allen

Rosen

Mayfield

And for comparison a couple from 2016:

Mahomes

Trubisky

Watson

 

 

[Sky Diver]

17 minutes ago, Hapless Bills Fan said:

 

To go further, here's the FO 2018 QBase data.  Fundamentally, their algorithm loves Mayfield and gives him almost equal probability of being elite as being a bust.

2nd favorite is Rosen. Darnold and Allen are not that dissimilar in their analysis-they give both >50% chance of being a bust and 20-30% chance of being an adequate starter.

 

They say themselves " It's important not to overestimate the importance of a small difference in the QBASE projections. Most of this year's top quarterbacks are grouped together in the middle of the scale. It's a bit of a surprise that Sam Darnold came out with a lower QBASE rating than some of the other top prospects -- Josh Rosen, for example. But if I were the analytics director for an NFL franchise, I would feel no need to disagree with a scouting director who placed Darnold ahead of Rosen."

 

There's a similar difference between Rosen and Darnold as between Darnold and Allen, but for some reason FO is all up in dissing on Allen ("horrifying") while exculpating Darnold's performance in their model.  Of more significance to my mind is their commentary that "QBASE is meant to only be used on players chosen in the top 100 picks; after that, the judgment of scouts becomes even more important, and statistics become even less predictive."  Hold the mustard here - what they probably mean by that is that their model is more valid for prospects from major programs facing a similar level of competition, and doesn't do so well with players outside that model.

 

I saw you palm that card, Football Outsiders.

 

Here's their data:

 

Darnold

Allen

Rosen

Mayfield

And for comparison a couple from 2016:

Mahomes

Trubisky

Watson

 

 

 

After the top 100, players are too inconsistent?

 

For Darnold, Rosen and Allen it’s basically a coin flip if they will be busts or not according to QBASE.

 

So what exactly are we learning here?

Edited July 25, 2018 by Sky Diver

[Hapless Bills Fan]

3 hours ago, Sky Diver said:

 

After the top 100, players are too inconsistent?

 

For Darnold, Rosen and Allen it’s basically a coin flip if they will be busts or not according to QBASE.

 

So what exactly are we learning here?

 

That's not the way the folks running Qbase present it, but that seems to be their bottom line when they ran the model. 

If it's no-big-difference between Rosen and Darnold, the other side of the error bar puts Darnold and Allen kinda close - Allen has a "bust potential" that is higher.

 

I think what we're learning is that some people who run analytics models are still using their Mach 1 Eyeball and interpreting their model accordingly.

 

 

[Sky Diver]

6 hours ago, Hapless Bills Fan said:

 

That's not the way the folks running Qbase present it, but that seems to be their bottom line when they ran the model. 

If it's no-big-difference between Rosen and Darnold, the other side of the error bar puts Darnold and Allen kinda close - Allen has a "bust potential" that is higher.

 

I think what we're learning is that some people who run analytics models are still using their Mach 1 Eyeball and interpreting their model accordingly.

 

 

 

QBASE looks like a lot of handwaving to me.