Bills offense compared to 2020 by the numbers, and they may shock you.

[Hapless Bills Fan]

33 minutes ago, FireChans said:

Your solution to one set of data being incomplete is to make both incomplete? 
Incredible statistic work.

 

I understand that you're being sarcastic here, but it may well be appropriate to do this, from a statistical standpoint.  At best, you can't say a priori it's more appropriate to compare a full and a partial data set.  And by the way, if I'm remembering correctly, @billsfan1959 has some good statistical background and actually knows more about this than the Average Bear, maybe more than I do since for much of my professional life I had the support of professional statisticians.

 

There are a number of factors that go into sound statistical practice.

One of them is comparable size data sets.

One of them is comparable data collection conditions

 

You can compare data sets of unequal sizes, but then you have to use special statistical tools and tests to ask "are these truly different, or are the differences I'm seeing the result of a smaller sample size in one set?"

 

If you have reason to believe that conditions varied during the course of the data collection (as occurs during a football season) it may be statistically appropriate to limit the size of the larger set to correspond to the smaller set.  Is there reason to believe that conditions vary during the season?  Yes, on the one hand weather may worsen, injuries may mount up.  On the other hand, the team may "click" and bond more as the season goes on, or find what they're best at and focus there, or just "practice makes perfect".

 

McDermott has openly said on many occasions that it's his goal for the team to play their best football in Nov and Dec, indicating he is trying to skew the full-season data set towards improvement at the end.

 

So it may well be the best "apples to apples" strategy to compare first 11 games last season to first 11 games this season.  It would probably be inappropriate or cherry picking to perform other comparisons, such as picking the best or worst 11 games from last season to to compare to this season.  It might be less cherry-picking to make a random selection of 11 games, but only if you believe there are no systematic factors that enter into the course of the season (which is probably true - see above).

[clayboy54]

16 hours ago, Virgil said:

Amazingly similar season for Josh. He's slightly better in a few key areas which is extremely impressive considering the drop-off in OL play. A little improvement every game and every season is what you're looking for. We're getting that, it appears.

[Alphadawg7]

26 minutes ago, FireChans said:

How can I be ignoring those games when I included every game he’s played? 

Your solution to one set of data being incomplete is to make both incomplete? 
 

Incredible statistic work.

 

We are comparing Josh’s numbers here. Last season compared to this season so far. Sorry I’m not cherry picking his first 11 games last year, you’re welcome to do so but that’s stupid.

 

The only one “manipulating data” is you. I have posted the numbers from every game our QB has played in the regular season in the last two years. Sorry it doesn’t fit your agenda.

 

Dude you are just being foolish now.  So one data set INCLUDES 6 more games that IMPROVED Josh season end numbers...and the other does NOT.  So you want to compare 11 games to 16 games instead of comparing where Josh was at the SAME time last year?  

 

You clearly do not understand statistical analysis to even the slightest.  Or you do, and just want to cherry pick things that don't make sense because doing a FAIR and ACCURATE analysis doesn't prove your biased.  

 

You are literally the definition of confirmation bias.  Where you search for data to support your biased view...and when its not there, you search for different data until you find something you hang your hat on.  

 

Bottom line, Josh is on par with last year through 11 games.  Nothing you can say can change those FACTS.

[FireChans]

27 minutes ago, Alphadawg7 said:

 

Dude you are just being foolish now.  So one data set INCLUDES 6 more games that IMPROVED Josh season end numbers...and the other does NOT.  So you want to compare 11 games to 16 games instead of comparing where Josh was at the SAME time last year?  

 

You clearly do not understand statistical analysis to even the slightest.  Or you do, and just want to cherry pick things that don't make sense because doing a FAIR and ACCURATE analysis doesn't prove your biased.  

 

You are literally the definition of confirmation bias.  Where you search for data to support your biased view...and when its not there, you search for different data until you find something you hang your hat on.  

 

Bottom line, Josh is on par with last year through 11 games.  Nothing you can say can change those FACTS.

And Josh this year is worse than Josh last year when including all of the games. Fact.

 

Sorry Charlie.

[Motorin']

1 hour ago, FireChans said:

How can I be ignoring those games when I included every game he’s played? 

 

[billsfan1959]

1 minute ago, FireChans said:

And Josh this year is worse than Josh last year when including all of the games. Fact.

 

Sorry Charlie.

 

You have had some pretty knowledgeable posters in this thread point out the flaws in your conclusions based on "statistics." You haven't bothered to respond to most of those points because it is apparent you are unable to do so.

 

At this point, it is clear you are interested in nothing more than manipulating numbers to pass off your opinions as "facts."

[FireChans]

2 minutes ago, billsfan1959 said:

 

You have had some pretty knowledgeable posters in this thread point out the flaws in your conclusions based on "statistics." You haven't bothered to respond to most of those points because it is apparent you are unable to do so.

 

At this point, it is clear you are interested in nothing more than manipulating numbers to pass off your opinions as "facts."

I didn’t even intend this topic to discuss Allen as much as how the offense is still really good, despite the eye test indicating that our offense is worse than last year.

 

You and your cabal tried to make it an “Allen sucks” topic because that’s all you have interest in discussing. No one worth responding to thinks Allen sucks. I refuse to discuss anything with you or Alpha further because that’s all your interested in talking about. I won’t defend a strawman I never stated or implied. Good day.

[billsfan1959]

2 minutes ago, FireChans said:

I didn’t even intend this topic to discuss Allen as much as how the offense is still really good, despite the eye test indicating that our offense is worse than last year.

 

You and your cabal tried to make it an “Allen sucks” topic because that’s all you have interest in discussing. No one worth responding to thinks Allen sucks. I refuse to discuss anything with you or Alpha further because that’s all your interested in talking about. I won’t defend a strawman I never stated or implied. Good day.

 

Actually, I never once even implied that you said "Allen sucks." I even stated that I had no problems saying the offensive efficiency numbers or statistics from this year might not be as good as last year at this point. My issues were simply with how you went about your statistical analyses and your conclusions that you tried to pass off as facts based on "statistics." Your reaction was to simply dig your heels in and double down on your flawed analyses.

 

And you have made it about Allen, from your first post to your last. Your quotes: "Conclusions? Josh is just not as sharp as he was last year."  "Josh this year is worse than Josh last year when including all of the games. Fact."

 

Those are your own words.

[Alphadawg7]

34 minutes ago, FireChans said:

And Josh this year is worse than Josh last year when including all of the games. Fact.

 

Sorry Charlie.

Not fact.  I’m convinced you don’t know what the word “fact” actually means

[FireChans]

2 minutes ago, billsfan1959 said:

 

Actually, I never once even implied that you said "Allen sucks." I even stated that I had no problems saying the offensive efficiency numbers or statistics from this year might not be as good as last year at this point. My issues were simply with how you went about your statistical analyses and your conclusions that you tried to pass off as facts based on "statistics." Your reaction was to simply dig your heels in and double down on your flawed analyses.

 

And you have made it about Allen, from your first post to your last. Your quotes: "Conclusions? Josh is just not as sharp as he was last year."  "Josh this year is worse than Josh last year when including all of the games. Fact."

 

Those are your own words.

 

1 minute ago, Alphadawg7 said:

Not fact.  I’m convinced you don’t know what the word “fact” actually means

I’m really not interested anymore guys. You’re pathological. Move along. 

Edited December 1, 2021 by FireChans

[billsfan1959]

Just now, FireChans said:

 

I’m really not interested anymore guys. You’re pathological. Move along. 

[Alphadawg7]

17 minutes ago, billsfan1959 said:

This literally made me laugh out loud

[FireChans]

1 hour ago, Alphadawg7 said:

This literally made me laugh out loud

It was incredible lol

[Hapless Bills Fan]

6 minutes ago, FireChans said:

It was incredible lol

 

I thought you had moved along?

[Generic_Bills_Fan]

20 hours ago, FireChans said:

So 2020 was a great offensive season by the Bills. Wonderful, exciting, electric, whatever you want to call it.

 

We can all agree the 2021 offense has its moments where it feels like a continuation of 2020 and moments where it feels like we are back in 2019. But what do the numbers say?

 

1. We are actually better at running the ball this year.

 

Last year, the Bills were 17th in attempts, 20th in yards, 15th in TD's and 20th in YPA.

 

This year, we are 17th in attempts, 11th in yards, 9th in TD's and 9th in YPA.

 

Some of this may be explained by Josh running a bit more (he's averaging 34 YPG compared to 28 last season, but on a similar number of attempts per game 6.4 to 6.3), but the fact remains, we are better at running.

 

2. We are worse through the air this year.

 

Last year, 11th in attempts, 3rd in TD's, 3rd in yards, 9th in INT's and 4th in NY/A

 

This year, 8th in attempts, 3rd in TD's, 7th in yards, 24th in INT's, and 10th in NY/A.

 

A lot of this is Josh, statistically. He is across the board less efficient and more TO prone compared to last season. Still playing great, still one of the best young QB's in the NFL, but not 2020 Josh.

 

3. On the whole, we are not much different than last year.

 

I have long held the belief that the best statistic to judge a unit to their peers is their "per drive" stats. Not as prone to confabulation due to the defense creating opportunities for the offense etc. 

 

Last year, Bills offense 2nd in score percentage, 15th in TO percentage, 5th in time of possession, 10th in plays per drive (an admittedly stat missing in utility due to scoring fast not being particularly bad), 4th in yards per drive, and 3rd in points per drive.

 

This year, Bills offense 2nd in score percentage, 12th in TO percentage, 14th in time of possession, 10th in plays per drive (an admittedly stat missing in utility due to scoring fast not being particularly bad), 4th in yards per drive, and 2nd in points per drive.

 

Outside of the huge difference in time per drive (which may be related to sheer number of possessions due to a great defense), we are still a very similar offense per drive. We score often, we are efficient, we get a lot of yards and we get a lot of points. Our TO% is actually better despite our INT number climbing, likely due to less fumbling overall.

 

Conclusions? Josh is just not as sharp as he was last year, but overall he is still playing great football and leading a great offense, an offense that by the numbers is just as good as the one last year.

 

The variance we are seeing from Josh is not really uncommon, even among the historically great. Last year, Aaron Rodgers threw for 48 TDs. This year, through 11 games, he has thrown for only 23, which will likely place him in the mid 30's as a season total trendwise. His passer rating is a full 17 points less this year compared to last year.

 

The great QB's have good years, and great years, and insane years. We aren't in an insane year for Josh. We are in a great year. And our offense continues to remain great.

I still swear we are in an insane year for josh and everything will go more smoothly with Feliciano and brown back but maybe I’ll have to eat my words haha we’ll find out soon.  every game he plays his oline is getting whooped like in the afc championship last year. Not to mention all our wins have been blowouts and most of which he has missed time in the 4th quarter…there haven’t been nearly as many shootouts where a late score puts the game away. Our pass blocking was outstanding at times last season 

Edited December 1, 2021 by Generic_Bills_Fan

[billsfan89]

I am not shocked, the Bills offense in the first 11 games of 2020 produced three stinkers (Titans, Chiefs and the second Jets game) the defense bailed out the field goal kicking offense against the Jets but the other two games were losses. In 2021 the Bills offense has had three stinkers the Steelers, Jags and Colts games. The rest of the games in 2021 the Bills have scored 26 or more points with all but one (the second Fins game) being a 30 or more point affair. 

 

The Bills in 2020 offensively turned in six good to great performances to end the season they scored 27, 34, 26, 48, 38 and 56 points in the last six games of the season in 2020. Hopefully the Bills in 2021 can find more consistency.

[FireChans]

18 minutes ago, Hapless Bills Fan said:

 

I thought you had moved along?

Funny deserves to be recognized.

[SoCal Deek]

Here’s how I would summarize it:

First, the Bills blowout wins came in the final six games of last season, not in the first eleven. Second, the 2021 Bills have been feast and famine, piling up stats in a number of blowout wins against bad opponents. Let the season play itself out. NE, TB, and NE all on tap. 

[HappyDays]

16 hours ago, GunnerBill said:

I wonder @FireChans if you have looked at consistency game to game? My perception is that the offense has been more boom or bust in 2021 than in 2020.... but that is perception I haven't crunched the numbers on it. 

 

Football Outsiders measures variance. This year our offense has by far the most variance in the league so we are extremely inconsistent. Last year we were the 13th most consistent. Definitely a big difference. Your perception is accurate.

[Buffalo716]

10 hours ago, Hapless Bills Fan said:

 

I understand that you're being sarcastic here, but it may well be appropriate to do this, from a statistical standpoint.  At best, you can't say a priori it's more appropriate to compare a full and a partial data set.  And by the way, if I'm remembering correctly, @billsfan1959 has some good statistical background and actually knows more about this than the Average Bear, maybe more than I do since for much of my professional life I had the support of professional statisticians.

 

There are a number of factors that go into sound statistical practice.

One of them is comparable size data sets.

One of them is comparable data collection conditions

 

You can compare data sets of unequal sizes, but then you have to use special statistical tools and tests to ask "are these truly different, or are the differences I'm seeing the result of a smaller sample size in one set?"

 

If you have reason to believe that conditions varied during the course of the data collection (as occurs during a football season) it may be statistically appropriate to limit the size of the larger set to correspond to the smaller set.  Is there reason to believe that conditions vary during the season?  Yes, on the one hand weather may worsen, injuries may mount up.  On the other hand, the team may "click" and bond more as the season goes on, or find what they're best at and focus there, or just "practice makes perfect".

 

McDermott has openly said on many occasions that it's his goal for the team to play their best football in Nov and Dec, indicating he is trying to skew the full-season data set towards improvement at the end.

 

So it may well be the best "apples to apples" strategy to compare first 11 games last season to first 11 games this season.  It would probably be inappropriate or cherry picking to perform other comparisons, such as picking the best or worst 11 games from last season to to compare to this season.  It might be less cherry-picking to make a random selection of 11 games, but only if you believe there are no systematic factors that enter into the course of the season (which is probably true - see above).

Exactly 

 

The goal is to be the best team in December and January... When it's high stakes football 

 

Lots of teams have been amazing in September and fluttered