Jump to content

Interesting ESPN article about skewed stats to tell stories. Allen used as example

loyal2dagame
 Share

Recommended Posts

Ronin Veteran

Ronin

Posted
Posted (edited)
50 minutes ago, leh-nerd skin-erd said:

 

It wasn't not always because he had open receivers.  It wasn't not always because he was impatient.  Sometimes he ran because he could, and sometimes he was patient in the pocket.

 

He does need to improve to be better though....and we can all agree he on that.  It looks like the coaching staff was not happy with some of the other pieces around him and have made attempts to improve his opportunity to improve.  What would make you happy to see out of JA in year two?  Not ecstatic, btw, just cautiously optimistic that he has grown and has potential to step forward yet again in year 3?  

 

Of course, it was for a variety of reasons, but it would be foolish in the analysis to discount the first two, eh.  

 

IMO if he can achieve average NFL QB passing it will be a huge step in the right direction, including minimizing his TOs.  Last season he had 8 fumbles but was fortunate to only have lost two.  My biggest concern insofar as his passing goes is his riskiness which was manifested, at least largely, but INTs.  I've said it often, what killed Kizer was his INTs, he had 22 or 1.5/game, his INT% was 4.6%.  

 

Allen had 12 INTs in 11 starts, his INT% was 3.8%.  The only one higher was Fitzpatrick who was 36 and who hasn't played well in several seasons.  If he continues to average an INT+/game and has 11 or 12 fumbles which would likely result in more than 3 lost-fumbles, then his TO issues would remain significant.  

 

If he can overcome the INTs IMO he'll have corrected, by default, a number of the issues that are plaguing him.  If he can do that then his passing TD production will be there as well.  

 

 As to numbers, as I've maintained, average would be a significant stride in the positive direction.  Average last season was about 22 TDs and 11 INTs.  I don't think that the yards will matter as much as that ratio and Compl.% will factor in as well as will YPA, which should also be in the average range, approx. 66%/7.5.  I don't think that happens unless all of those metrics move together, they're at least somewhat related, as it pertains to him.  That would put him in the company of a Stafford, Manning, Prescott, Dalton, or Trubisky for last season, which is far from too lofty a goal, none of those QBs had fantastic seasons last year.  

 

 

 

Edited by Ronin
Link to comment
Share on other sites
Thurman#1 Hall of Famer

Thurman#1

Posted
Posted (edited)
12 hours ago, loyal2dagame said:

https://www.espn.com/fantasy/football/story/_/id/27054210/100-facts-2019-fantasy-football-season

 

Finally someone admits all the talk about players is made up and for show. 

 

 

Like stats are the key point here. They're not. Of course TV shows are for show.

 

And it doesn't take stats to switch your opinion on cue. It's just as easy to do with "the eye test," or anything else, really.

 

And while I'm sure there are plenty of shows out there that work like that, I'm equally sure that there are plenty of shows where the pundits say just what they feel and believe.

 

The Allen example was pretty funny, but even as I read it was obvious that the writer was straining like crazy to tell each story, the story of QB A and QB B.

 

I mean, to tell the story in stats for QB B, he had to use "Tristan H. Cockcroft's great Consistency Ratings [in which] our guy was tied for sixth in 'star' games." What? Please. Then he goes to "aDOT," Average Depth of Target. Which tells you virtually nothing usable about whether a guy is good or bad. Then he went to how many "fantasy points" that QB scored. And finally some good run stats. That's all the stats he had. Puh-leeze. What you've got there that means something is probably the good run stats.

 

He's very right that confirmation bias twists more thinking than alcohol and mental illness combined, and that once you commit yourself to a viewpoint on a guy, you'll only see the stats, plays, comments, stories and "eye test" results that back up your opinion. He's very very right about that.

 

 

 

 

Edited by Thurman#1
Link to comment
Share on other sites
Drunken Pygmy Goat Veteran

Drunken Pygmy Goat

Posted
Posted (edited)
Quote

These facts tell only part of the story. The part of the story I want you see. The part that supports whatever opinion I have of a player. The opinion I want to try to convince you to share with me.

 

No matter what you read, hear, write or watch -- it's not the WHOLE story.

 

One of my favorite sayings is "stats never lie, but they never tell the whole truth either"...it's pretty easy to point out an agenda driven statement or viewpoint, be it in sports, politics, etc., so long as you're aware and looking for them.

 

This forum is proof. People do it all the time to make a point, and aware people call them out for it. 

 

Way to take a simple, rather well known sports phrase and turn it into a long, offseason filler article, Mr. Berry! ?

Edited by Drunken Pygmy Goat
  • Like (+1) 1
Link to comment
Share on other sites
Thurman#1 Hall of Famer

Thurman#1

Posted
Posted (edited)
2 hours ago, BADOLBILZ said:

 

Numbers.....slanted as "statistics"...... are fake news to the ignorant.  

 

 

 

Yeah. That's why so very very many really excellent players have terrible statistics. And so many awful players have Hall of Fame type stats. It's so common that ...

 

Oh, wait.

 

 

 

 

For those saying that stats don't tell the whole story ... very true. Nothing does, but you can certainly include stats in that.

 

For those saying stats are nonsense and fake news ... that statement is nonsense.

Edited by Thurman#1
  • Like (+1) 1
Link to comment
Share on other sites
oldmanfan All Pro

oldmanfan

Posted
Posted
4 hours ago, Thurman#1 said:

 

 

Yeah. That's why so very very many really excellent players have terrible statistics. And so many awful players have Hall of Fame type stats. It's so common that ...

 

Oh, wait.

 

 

 

 

For those saying that stats don't tell the whole story ... very true. Nothing does, but you can certainly include stats in that.

 

For those saying stats are nonsense and fake news ... that statement is nonsense.

Depends on the particular stat and how it's calculated.

Link to comment
Share on other sites
row_33 Hall of Famer

row_33

Posted
Posted
7 hours ago, Thurman#1 said:

 

 

Yeah. That's why so very very many really excellent players have terrible statistics. And so many awful players have Hall of Fame type stats. It's so common that ...

 

Oh, wait.

 

 

 

 

For those saying that stats don't tell the whole story ... very true. Nothing does, but you can certainly include stats in that.

 

For those saying stats are nonsense and fake news ... that statement is nonsense.

 

You seen what geek baseball are doing lately?

 

 

Link to comment
Share on other sites
CincyBillsFan Veteran

CincyBillsFan

Posted
Posted
On 6/30/2019 at 1:35 PM, oldmanfan said:

This is why I am generally skeptical of the statistical analyses promoted by many.  I have a research background and when planning studies you decide what question you want to ask, determine variables that can be controlled so you can measure between control and experimental groups, and define sample sizes and statistical methods that are appropriate for data analysis.   What I tend to see with info such as referred to in the story is you can have an opinion, then choose statistics to buttress your opinion.  The term used for the latter is confirmation bias and it's rampant in studies.  When I review manuscripts for publication I reject 80-90% of them because of inappropriate study design and thus inaccurate conclusions.

 

Exactly right oldmanfan. 

 

The extreme variability and relatively small sample size in the game of football does not lend itself to the use of analytics to assess performance.  And this is particularly true for a QB.  A few simple stats like games won, playoff games won, TD passes thrown, INT's thrown,  yards per pass and completion percentage can all help describe a QB's performance.  But even here they're not of much use until a QB has completed several seasons. 

 

Think about how poorly analytics does in predicting a rookie QB's progression from year 1 to year 2.  If statistical analyses was an appropriate tool for evaluating QB's we should be able to predict, after a single season, a QB's next season outcomes.  Of course even a casual glance at the stats will tell you that they can't predict year 2 & 3 performance based on year 1 numbers. 

 

This is why the critical stats for rookie QB's likely involve subjective things like "leadership", "grit" and whether they improved over the course of their rookie year.  I know it's frustrating to the numbers jocks but the "eye test" by folks with a lot of experience assessing NFL QB's is probably the best way to evaluate young QB's and determine their potential. 

 

 

Link to comment
Share on other sites
CincyBillsFan Veteran

CincyBillsFan

Posted
Posted
15 hours ago, Thurman#1 said:

 

 

Yeah. That's why so very very many really excellent players have terrible statistics. And so many awful players have Hall of Fame type stats. It's so common that ...

 

Oh, wait.

 

 

 

 

For those saying that stats don't tell the whole story ... very true. Nothing does, but you can certainly include stats in that.

 

For those saying stats are nonsense and fake news ... that statement is nonsense.

 

In defense of Badolbilz I think he's referring to how statistics can be manipulated intentionally or unintentionally to give us almost any answer we want.  From what I can see much of the analytics used in football does not properly account for the extreme variability, subjective nature of the observations and small sample sizes involved in the game.  This often makes the results of the statistical analyses irrelevant or flat out wrong.

 

But as you point out analytics can be useful after a QB has accumulated a long enough track record.  Towards the end of their career a QB's stats are a great way to capture his performance relative to his peers.  As you note, HOF's don't have crappy stats.

 

My issue is the use of analytics to assess very young NFL QB's.  I just don't think the numbers gathered during a QB's first 2 to 3 seasons are very useful in predicting the steady-state performance of a QB over their career. 

  • Like (+1) 1
Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...