Allen is NOT inaccurate unless Baker, Lamar, Darnold, Rosen, 2017 Watson & 2016 Wentz are, too

[Nextmanup]

3 hours ago, transplantbillsfan said:

This is the kinda stuff that made me start this exercise.

 

Schatz has been one of those really highly respected guys saying Allen is somehow so much more inaccurate than other rookies.

 

He's not wrong that Allen misses easy throws, but he's very wrong that Allen is unique in that sense among good rookie QBs.

 

This kinda statement is the equivalent of, "before you eat that lemon, watch out! That one is sour!"

He was on with Schopp & Bulldog yesterday.

 

He maintained that he has always been, and remains, a non-believer in Josh Allen.

 

Schopp asked why, and he responded that he is simply too inaccurate and that is not going to be improved by learning.

 

He also pointed out that his running was fantastic, those yards count, and QB runs can be a big addition to the offense.  

 

Schatz also mentioned that at the halfway mark of last season, the Bills were set to have the worst offense in something like 30 years of data as compiled/studied by Footall Outsiders.  But then we hit the easy part of the schedule and Allen was able to improve the stats to the point of us not even being in last place for the season.

 

Most of Schatz's segment discussed things other than Josh Allen.

 

 

 

[transplantbillsfan]

2 hours ago, Nextmanup said:

He was on with Schopp & Bulldog yesterday.

 

He maintained that he has always been, and remains, a non-believer in Josh Allen.

 

Schopp asked why, and he responded that he is simply too inaccurate and that is not going to be improved by learning.

 

He also pointed out that his running was fantastic, those yards count, and QB runs can be a big addition to the offense.  

 

Schatz also mentioned that at the halfway mark of last season, the Bills were set to have the worst offense in something like 30 years of data as compiled/studied by Footall Outsiders.  But then we hit the easy part of the schedule and Allen was able to improve the stats to the point of us not even being in last place for the season.

 

Most of Schatz's segment discussed things other than Josh Allen.

 

Yes, clearly Schatz doesn't like Allen

 

Doesn't matter.  Perception is reality.  Allen looks like he could be an excellent pocket passer if his OL actually gave him time to throw it.  He demonstrated that last year when he was given time and often when he wasn't given time and was smashed while making a great throw, like the TD to Foster in the Jags game.

 

I actually kinda do wish that others would try this exercise.  I look at something like PFF and see a series of tweets like this:

 

and I just can't take them seriously having watched all of these guys play.

[transplantbillsfan]

 

[Foxx]

53 minutes ago, transplantbillsfan said:

 

... I actually kinda do wish that others would try this exercise.  I look at something like PFF and see a series of tweets like this:

 

and I just can't take them seriously having watched all of these guys play.

so your saying... you don't agree with guys that are considered professionals for their analyses?

 

i have not been following along, can you tell me why? is it because their observations do not align with your confirmation bias or is it something else entirely? 

Edited January 29, 2019 by Foxx

[C.Biscuit97]

Sal C on WGR said Allen isn’t the most accurate qb. Must be biased. 

[Augie]

If we can “debunk” one theory, can we “bunk” another? 

[oldmanfan]

Looking at Palazzolo's graphs he seems to have a very limited definition of accuracy.  Accuracy is how close you are to a target.  If you want to say someone is inaccurate then you have to decide how far from the target should be constitute being inaccurate.  In statistical terms it's standard deviation; a measure of distance from the mean.  If you decide you have to be for example within 0.1 SD of the mean you have little if any margin for error before calling something inaccurate (for scientific data it's usually 2SD that is taken as a norm).  

 

The maps shown indicate the numbers on the uniform are the target.  And that any throw not hitting the uniform is inaccurate.  That's a pretty narrow definition, and would not take into account exactly where the QB was trying to throw it, whether the receiver got to the spot he was throwing to and such.  

 

In contrast, transplantfan says an accurate pass is within the catch radius.  This to me seems more reasonable; it takes more into account things like the receiver being in the right spot.  

 

I don't have access to all their data, but to understand it you'd need to see how they define their terms, how they do measurements, the error in their measurements and such.  They may get paid for this; it doesn't mean their methods are appropriate.  It is always reasonable to be skeptical of statistical analysis until one sees the methods.  I review dozens of scientific papers a year and reject most because their methods and/or analysis is flawed.

[transplantbillsfan]

21 minutes ago, Foxx said:

so your saying... you don't agree with guys that are considered professionals for their analyses?

 

i have not been following along, can you tell me why? is it because their observations do not align with your confirmation bias or is it something else entirely? 

 

It's because I actually watched every single pass of all 5 rookies along with Watson--so far---and as I watched every single pass I identified each pass simply as catchable or uncatchable and what I saw was that Allen's accuracy was at least similar to all the other rookies this year and Watson last year.

 

That's why.

 

It's not that their observations don't align with my confirmation bias.  It's that their system is clearly overly intricate and impossible to accurately  chart with NINE different categories each throw would fall into.

 

And how is each of those categories defined?

 

For example, are passes 20 yards down the field to a WR who has to stop in the middle of a crossing route to catch the football in his chest still one of those throws that they categorize in that "Accurate+" circle in the middle of the body?

 

Seems like my "confirmation bias" is no stronger than theirs, honestly.

[C.Biscuit97]

3 minutes ago, oldmanfan said:

Looking at Palazzolo's graphs he seems to have a very limited definition of accuracy.  Accuracy is how close you are to a target.  If you want to say someone is inaccurate then you have to decide how far from the target should be constitute being inaccurate.  In statistical terms it's standard deviation; a measure of distance from the mean.  If you decide you have to be for example within 0.1 SD of the mean you have little if any margin for error before calling something inaccurate (for scientific data it's usually 2SD that is taken as a norm).  

 

The maps shown indicate the numbers on the uniform are the target.  And that any throw not hitting the uniform is inaccurate.  That's a pretty narrow definition, and would not take into account exactly where the QB was trying to throw it, whether the receiver got to the spot he was throwing to and such.  

 

In contrast, transplantfan says an accurate pass is within the catch radius.  This to me seems more reasonable; it takes more into account things like the receiver being in the right spot.  

 

I don't have access to all their data, but to understand it you'd need to see how they define their terms, how they do measurements, the error in their measurements and such.  They may get paid for this; it doesn't mean their methods are appropriate.  It is always reasonable to be skeptical of statistical analysis until one sees the methods.  I review dozens of scientific papers a year and reject most because their methods and/or analysis is flawed.

I respect your tenacity.  But you are working really hard to defend a 52% nfl passer, a 56% Mountain West Passer, and a 49% JUCO.  If Allen was more accurate, he’d be the number 1 pick in the draft and won’t have had to go to Wyoming.  He needs to get better.

[oldmanfan]

8 minutes ago, C.Biscuit97 said:

I respect your tenacity.  But you are working really hard to defend a 52% nfl passer, a 56% Mountain West Passer, and a 49% JUCO.  If Allen was more accurate, he’d be the number 1 pick in the draft and won’t have had to go to Wyoming.  He needs to get better.

Here we go again.  I want to lay this out and ask you to respond.  I and many others have shown you many times now that accuracy is not completion percentage.  We have on many occasions pointed out the factual reasons why, including the extreme number of variables that affect whether a pass is complete or not ( such as if a pass is dropped or receiver runs a bad route).  I have pointed out more times than I can count that if one throws 30 passes in a game that if two become complete vs incomplete you go from 52% to 60%.  

 

 

Accuracy is how close you come to a given target, completion percentage is completely different.  It is fact.  Yet you continue to cling to completion percentage as a measure of accuracy.  As I have said before, it's as if you say the sky is purple and despite folks laying out all the scientific and other reasons why it's blue you ignore it and say it's purple.

 

Why do I challenge this stuff?  For several reasons.  One, I have training in statistics and have seen over a 40 year career how statistics are routinely misused to support a preconceived bias.  Two, I charted some games and also looked at accuracy for Allen and my analysis is closer to the OP.  Three, I hope my relative expertise in areas of statistics adds something to the discussion.

 

Does Allen need to improve?  Absolutely and I have said so.  I have pointed out he is accurate but needs to be more precise in his ball placement.  He needs more touch on his short throws.  Like all young QBs he needs to work on his pre-snap reads and make quicker decisions.

 

So I ask you again:  why do you cling to completion percentage as some Holy Grail of accuracy, when so many have show why you are in error?

[PrimeTime101]

1 hour ago, C.Biscuit97 said:

Sal C on WGR said Allen isn’t the most accurate qb. Must be biased. 

He is right. he is not.,. but you don't have to be the most accurate QB to be a franchise QB now do you? I think your take on looking at numbers like completion percentage and other numbers like this is LAZY! If you look at all the teams the rookie QB's went to, our Offensive situation was by far the WORST! yet you continue to ignore this fact.

 

Want to go look at numbers?

Look at the % he threw 15 yards or longer

Look at the % he had to run for his life

Look at the WR rankings with creating gap

Look at the WR drop rate, especially when Allen got better in the last 7 games

Look at the O Line starters from last year that likely would not start for any other team?

Look at the run game and how little it produced outside JA running

 

Then get a clue to why JA had the numbers you were referring to. I mean did you even watch one game? or is it because your 20 years old and have not ever seen a real QB on this team? witch is it?

 

Get a CLUE!

[LSHMEAB]

1 hour ago, C.Biscuit97 said:

Sal C on WGR said Allen isn’t the most accurate qb. Must be biased. 

HATERRRRRR!

[Doc]

No he is not the most accurate QB.  That would be Brees.

[CincyBillsFan]

1 hour ago, oldmanfan said:

Looking at Palazzolo's graphs he seems to have a very limited definition of accuracy.  Accuracy is how close you are to a target.  If you want to say someone is inaccurate then you have to decide how far from the target should be constitute being inaccurate.  In statistical terms it's standard deviation; a measure of distance from the mean.  If you decide you have to be for example within 0.1 SD of the mean you have little if any margin for error before calling something inaccurate (for scientific data it's usually 2SD that is taken as a norm).  

 

The maps shown indicate the numbers on the uniform are the target.  And that any throw not hitting the uniform is inaccurate.  That's a pretty narrow definition, and would not take into account exactly where the QB was trying to throw it, whether the receiver got to the spot he was throwing to and such.  

 

In contrast, transplantfan says an accurate pass is within the catch radius.  This to me seems more reasonable; it takes more into account things like the receiver being in the right spot.  

 

I don't have access to all their data, but to understand it you'd need to see how they define their terms, how they do measurements, the error in their measurements and such.  They may get paid for this; it doesn't mean their methods are appropriate.  It is always reasonable to be skeptical of statistical analysis until one sees the methods.  I review dozens of scientific papers a year and reject most because their methods and/or analysis is flawed.

 

On the face of it aiming only at the numbers on the uniform is dumb given the many different types of passes thrown by NFL caliber QB's.  Just off the top of my head it would be profoundly inaccurate (or stupid) for a QB to aim for the numbers of his 6 ft 6 inch WR being guarded by a 5 ft 10 in inch DB in the end zone.  I would think throwing the ball up high where his receivers height advantage could be used most effectively is the most ACCURATE pass in that situation.

 

Think about that spectacular pass Allen threw to Foster to convert a 3rd & 13 out of his end zone against Miami in their first meeting. The Throw was a laser beam deliberately put about a foot out of bounds so only Foster could catch it with his HANDS.  Aiming a pass for the number on Fosters Jersey would probably have resulted in an INT instead of a first down.

 

I would submit that half or more of an NFL's QB's throws are not intended for the numbers on the jersey. 

 

 

[oldmanfan]

1 minute ago, CincyBillsFan said:

 

On the face of it aiming only at the numbers on the uniform is dumb given the many different types of passes thrown by NFL caliber QB's.  Just off the top of my head it would be profoundly inaccurate (or stupid) for a QB to aim for the numbers of his 6 ft 6 inch WR being guarded by a 5 ft 10 in inch DB in the end zone.  I would think throwing the ball up high where his receivers height advantage could be used most effectively is the most ACCURATE pass in that situation.

 

Think about that spectacular pass Allen threw to Foster to convert a 3rd & 13 out of his end zone against Miami in their first meeting. The Throw was a laser beam deliberately put about a foot out of bounds so only Foster could catch it with his HANDS.  Aiming a pass for the number on Fosters Jersey would probably have resulted in an INT instead of a first down.

 

I would submit that half or more of an NFL's QB's throws are not intended for the numbers on the jersey. 

 

 

Correct.  The problem with all these measurements is that only the QB knows where he's aiming.

[formerlyofCtown]

Did you add Manning.

On 1/24/2019 at 6:39 PM, transplantbillsfan said:

 

I don't think ball velocity was much of a problem.  The complaint that Allen has no touch and puts too much into every pass is overblown, especially by the end of the season.

Judging by the numbers it possibly keeps his passes from being batted.

[YattaOkasan]

On 1/18/2019 at 11:48 AM, Shaw66 said:

You know, old man, I like words and I like precision, and I like precision in the use of words.   You and some others get off on this accuracy versus precision distinction every once in a while, and I sometimes pay attention to it.  

 

However, in my mind it's a distinction without any real significance.   I get that there's a difference - I think Brees is both accurate and precise, some guys are accurate and not precise and some may be precise but not accurate (although I doubt there are many of those).   

 

The point is that I seriously doubt that some coach in Buffalo is saying "we need to work on Allen's accuracy" and some other coach is saying "no, we need to work on his precision."   I think both coaches are looking at the same film of the same play and agreeing that the ball wasn't in as good a spot as they would prefer and then looking at the film to see if they can determine what it is that Allen needs to work on to get the ball in the right spot as often as possible.   It's not like they say "well, if he has an accuracy problem he needs to fix this and if he has a precision problem he needs to fix that."   For them, it's just semantics that doesn't add to the conversation.   They're just asking themselves what needs to be done to make Allen better.  

Really liked your contributions to this thread but wanted to reply to this topic. 

 

I think coahes could get to the accuracy/precision distinction. As you mentioned it’s all about getting the ball where it needs to be. coaches could see a lack of precision as evidenced by a change in or sloppy mechanics. That would lead to more reps for muscle memory. Or they could see his front foot not pointed correctly thus not being accurate despite consistent mechanics. Maybe they don’t talk about accuracy or precision but I think they talk in this way. If that’s the case I would think they are having an accuracy/precision discussion without knowing it. 

[transplantbillsfan]

3 hours ago, formerlyofCtown said:

Judging by the numbers it possibly keeps his passes from being batted.

 

That was what I saw.

 

Maybe unsurprisingly Baker Mayfield was the one with the most batted balls at the LOS.

 

I don't have the breakdown of tipped at the LOS vs mid air on the way to the WR, but Mayfield had the most at the LOS and Darnold had by far the most mid-flight.

[Foxx]

21 hours ago, transplantbillsfan said:

 

It's because I actually watched every single pass of all 5 rookies along with Watson--so far---and as I watched every single pass I identified each pass simply as catchable or uncatchable and what I saw was that Allen's accuracy was at least similar to all the other rookies this year and Watson last year.

 

That's why.

 

It's not that their observations don't align with my confirmation bias.  It's that their system is clearly overly intricate and impossible to accurately  chart with NINE different categories each throw would fall into.

 

And how is each of those categories defined?

 

For example, are passes 20 yards down the field to a WR who has to stop in the middle of a crossing route to catch the football in his chest still one of those throws that they categorize in that "Accurate+" circle in the middle of the body?

 

Seems like my "confirmation bias" is no stronger than theirs, honestly.

and i would counter that the PFF guys/gals watched every single pass as well. 

 

as far as their system/analysis being overly intricate (according to you), i doubt that it would be impossible to chart, being as, you know, they charted it. as an addendum here.. is it any more overly intricate than your ridicules charting of Tyrods passing? i mean, you did have the field broken down into NINE different sections, right?

 

how are each of their categories defined? i don't know to be honest, but i am willing to bet that they are defined.

 

as for confirmation bias, you do know what that term means, right? it means that you seek confirmation of your bias. being that PFF is a major analysis outfit, my money is on it being a hell of a lot more objective than your analysis. sure i could be wrong but i don't think so. i think the odds of their objectivity being innocently objective is considerably higher than yours.

 

again, i respect the time and effort you have put forth here, i'm just not real convinced of the conclusions being drawn. 

 

i want JA to be the man, i just don't know that he is or isn't the man at this point. after this coming year, i think we all will have a better idea just where he might stand on that desired want.

Edited January 30, 2019 by Foxx

[transplantbillsfan]

1 hour ago, Foxx said:

and i would counter that the PFF guys/gals watched every single pass as well. 

 

as far as their system/analysis being overly intricate (according to you), i doubt that it would be impossible to chart, being as, you know, they charted it. as an addendum here.. is it any more overly intricate than your ridicules charting of Tyrods passing? i mean, you did have the field broken down into NINE different sections, right?

 

That's weird.

 

Why are you bringing up Taylor?

 

What does he have to do with this?

 

1 hour ago, Foxx said:

 

how are each of their categories defined? i don't know to be honest, but i am willing to bet that they are defined.

 

That's nice.  You can trust them if you want.

 

I got no problem with that.

 

But you're trusting them blindly.

 

1 hour ago, Foxx said:

 

as for confirmation bias, you do know what that term means, right? it means that you seek confirmation of your bias. being that PFF is a major analysis outfit, my money is on it being a hell of a lot more objective than your analysis. sure i could be wrong but i don't think so. i think the odds of their objectivity being innocently objective is considerably higher than yours.

 

Well, we can disagree.  And we do.  I think these "experts" and pundits are often victims of their own analysis, and what did the analysis about Allen coming out of college overwhelmingly state?

 

"Josh Allen has an area code accuracy problem and should sit on the bench for a year or two to learn the game."

 

Now that Allen's rookie year is over and it was for a team like the Bills that just doesn't get a lot of national attention who finished with a worse record than last year and Allen's numbers in the passing game look mediocre to poor, it's pretty easy to gloss over that with an "I was right" and move onto the next guy.

 

You can trust their system, but their system involves nine different categories.  If you trust that someone from PFF has accurately divided up Allen's 320 throws into those 9 categories--and it's actually probably significantly less than the 320 total throws because I would assume they would discard throwaways or batted/tipped passes, but since there's nothing to explain that part, who knows?--and is left with an average of less than 40 throws for each category, which means that a single ill-defined pass moves the needle more than a couple percentage points one way or the other....

 

then...

 

well...

 

you go ahead and do it.

 

I'm not even saying my numbers should be 100% trusted.  I did the exercise, was aware of potential subjectivity and tried to absolutely throw it out the window, but also acknowledge that that would be completely understandable for someone not to trust me.

 

But that's why I think you should try it yourself rather than blindly trust someone like PFF.

 

1 hour ago, Foxx said:

 

again, i respect the time and effort you have put forth here, i'm just not real convinced of the conclusions being drawn. 

 

You don't have to be.

 

Try it yourself and draw your own conclusions if you're that skeptical of mine.

 

If you distrust them that much, though, and just don't want to try the exercise, then what's the point of your arguing with me over something you refuse to do?

 

1 hour ago, Foxx said:

 

i want JA to be the man, i just don't know that he is or isn't the man at this point. after this coming year, i think we all will have a better idea just where he might stand on that desired want.

 

Agreed.