The Most Important Passing Stat - 2nd Down Yardage?

[IC Khari]

On 2017-05-07 at 10:02 PM, BigBlue said:

AND THIS PROVES WHAT?

Does it change the position of the horse behind the cart?

That some sites copy and paste real well.

[Jpan85]

Matt Nichols I don't care the numbers is the most Winnipeg guy QB since Dunnigan. Just something about him says while his production might not say it he will be very successful. He is a Mike O'Shea guy when your QB is the mirror image of what your head coach wants all his players to be you have something.

[Eternal optimist]

On 5/7/2017 at 10:02 PM, BigBlue said:

AND THIS PROVES WHAT?

Does it change the position of the horse behind the cart?

It proves that wins is not an input variable in the QB efficiency rating (which was the "horse behind the cart" dilemma you had mentioned). I'm not sure you're understanding what I'm saying, but I don't know how better to explain it.

Bottom line is the best predictor for determining the number of wins a quarterback is likely to have is the QB efficiency rating. It is a calculation designed to predict the output variable (wins) based on the given input variables as discussed in my prior post.

[BigBlue]

28 minutes ago, Eternal optimist said:

It proves that wins is not an input variable in the QB efficiency rating (which was the "horse behind the cart" dilemma you had mentioned). I'm not sure you're understanding what I'm saying, but I don't know how better to explain it.

Bottom line is the best predictor for determining the number of wins a quarterback is likely to have is the QB efficiency rating. It is a calculation designed to predict the output variable (wins) based on the given input variables as discussed in my prior post.

I am just trying to isolate predictors instead of coincedental and lagging indicators ... I don't believe efficiency rating is a predictor .... I am suggesting 2nd down performance is one, even when things may not be turning out very well at the moment .... QBs with good 2nd down stats are eventually going to have a great deal of success in the future .... I don't know what other stat predicts as well as that one

Edited May 14, 2017 by BigBlue

[Eternal optimist]

2 hours ago, BigBlue said:

I am just trying to isolate predictors instead of coincedental and lagging indicators ... I don't believe efficiency rating is a predictor .... I am suggesting 2nd down performance is one, even when things may not be turning out very well at the moment .... QBs with good 2nd down stats are eventually going to have a great deal of success in the future .... I don't know what other stat predicts as well as that one

Did a bit of digging, and there doesn't appear to have been any analysis between 2nd down conversions with wins for CFL quarterbacks. However, I did find comparative analysis using a similar thought process with 3rd down conversions in the NFL. The correlation coefficient between 3rd down conversions and wins in the NFL is about 0.43 which is pretty good for as a single-variable predictor. The QB passer rating still surpasses it though and has a correlation with wins of about 0.51.

That said, the NFL/CFL are completely different games and the shortcomings of the QB passer rating are well documented. For those who want a bit of reading, here's the articles that I found on the subject:

Source regarding 3rd down conversions: http://archive.advancedfootballanalytics.com/2008/01/is-3rd-down-conversion-percentage-good.html

Source regarding QB passer rating: http://www.footballperspective.com/correlating-passing-stats-with-wins/

 

 

[BigBlue]

3 hours ago, Eternal optimist said:

Did a bit of digging, and there doesn't appear to have been any analysis between 2nd down conversions with wins for CFL quarterbacks. However, I did find comparative analysis using a similar thought process with 3rd down conversions in the NFL. The correlation coefficient between 3rd down conversions and wins in the NFL is about 0.43 which is pretty good for as a single-variable predictor. The QB passer rating still surpasses it though and has a correlation with wins of about 0.51.

That said, the NFL/CFL are completely different games and the shortcomings of the QB passer rating are well documented. For those who want a bit of reading, here's the articles that I found on the subject:

Source regarding 3rd down conversions: http://archive.advancedfootballanalytics.com/2008/01/is-3rd-down-conversion-percentage-good.html

 

3 hours ago, Eternal optimist said:

Source regarding QB passer rating: http://www.footballperspective.com/correlating-passing-stats-with-wins/

 

 

"That means that we can predict a teams's future 3rd down percentage with almost crystal ball-like accuracy using passing efficiency stats.... So if we want to know a team's ability to covert 3rd downs, we're far better off looking at passing stats than previous 3rd down conversion rates. And a prediction model is far better off using those passing stats (pass efficiency, interception rate, sack rate) and excluding to-date 3rd down percentage."

(your http://archive.advancedfootballanalytics.com/2008/01/is-3rd-down-conversion-percentage-good.html)

[BigBlue]

3 hours ago, Eternal optimist said:

Did a bit of digging, and there doesn't appear to have been any analysis between 2nd down conversions with wins for CFL quarterbacks. However, I did find comparative analysis using a similar thought process with 3rd down conversions in the NFL. The correlation coefficient between 3rd down conversions and wins in the NFL is about 0.43 which is pretty good for as a single-variable predictor. The QB passer rating still surpasses it though and has a correlation with wins of about 0.51.

That said, the NFL/CFL are completely different games and the shortcomings of the QB passer rating are well documented. For those who want a bit of reading, here's the articles that I found on the subject:

Source regarding 3rd down conversions: http://archive.advancedfootballanalytics.com/2008/01/is-3rd-down-conversion-percentage-good.html

Source regarding QB passer rating: http://www.footballperspective.com/correlating-passing-stats-with-wins/

 

 

"Since passer rating is just an average of four statistics, there’s a better way to analyze the four inputs. You can run a multiple regression analysis to see how much weight should be placed on each variable, with future wins (i.e., wins in the other half-season) as the output. The P-values on the completion percentage and interception rate variables were not significant at the 1%, 5% or 10% levels. Essentially, this means that for predictive purposes, two of the four inputs in passer rating are meaningless.... What does make the formula predictive? Using net yards per attempt — which deducts sacks from a passer’s production — is the simplest and best way to predict future performance." 

(your http://www.footballperspective.com/correlating-passing-stats-with-wins/ )

[Eternal optimist]

15 hours ago, Eternal optimist said:

Did a bit of digging, and there doesn't appear to have been any analysis between 2nd down conversions with wins for CFL quarterbacks. However, I did find comparative analysis using a similar thought process with 3rd down conversions in the NFL. The correlation coefficient between 3rd down conversions and wins in the NFL is about 0.43 which is pretty good for as a single-variable predictor. The QB passer rating still surpasses it though and has a correlation with wins of about 0.51.

That said, the NFL/CFL are completely different games and the shortcomings of the QB passer rating are well documented. For those who want a bit of reading, here's the articles that I found on the subject:

Source regarding 3rd down conversions: http://archive.advancedfootballanalytics.com/2008/01/is-3rd-down-conversion-percentage-good.html

Source regarding QB passer rating: http://www.footballperspective.com/correlating-passing-stats-with-wins/

 

 

I'm not disagreeing with you regarding using 2nd-down conversion as a predictor, in fact it looks like it has some merit. I was just saying that nobody (as far as I can tell) has run the numbers on it for the CFL. As for your other post regarding the QB efficiency rating, its shortcomings are well known and documented.

[blue_gold_84]

QB efficiency rating seems like most accurate metric, IMO. Wins are a team statistic.

[Mike]

What they really need to do is develop a QBR type formula that uses advanced statistics to come to it's conclusion. Then you'll be getting somewhere.

Combine (Completion % - Expected Completion %), At-fault interception throws, yards per attempt, sacks per snap, etc.

It's nearly impossible to develop a formula to rate quarterbacks based on the raw, preliminary numbers we accept as the standard. A quarterback's individual success is too dependent on the result at the other end of his throw and without coming up with a way to define that, you're not going to get anywhere.

Quarterback throws the ball, it's tipped by the receiver into a defender's hands. QB's fault? No, but it has a negative impact on his statistics.

Quarterback takes the snap, LT blows the edge and allows the rush to get through, QB sails it out of bounds. QB's fault? No, but it has a negative impact on his statistics.

It works in the other direction too. 

Quarterback throws the ball right to a DB because he misread a coverage. DB drops it. QB's fault? Yes, but no negative impact.

TLDR: Football is the toughest sport to analyze statistically. We're nowhere even close to being able to do it.

[17to85]

14 minutes ago, Mike said:

TLDR: Football is the toughest sport to analyze statistically. We're nowhere even close to being able to do it.

And for these reasons maybe we need to stop trying. I mean it's a position that really does rely on so many other people doing their job correctly it seems like a lot of wasted effort trying to distil an individual into a catch all stat that tells how good or bad they are. Can't we just watch the games and judge who is good and who isn't?

[The Zipp]

13 minutes ago, Mike said:

What they really need to do is develop a QBR type formula that uses advanced statistics to come to it's conclusion. Then you'll be getting somewhere.

Combine (Completion % - Expected Completion %), At-fault interception throws, yards per attempt, sacks per snap, etc.

It's nearly impossible to develop a formula to rate quarterbacks based on the raw, preliminary numbers we accept as the standard. A quarterback's individual success is too dependent on the result at the other end of his throw and without coming up with a way to define that, you're not going to get anywhere.

Quarterback throws the ball, it's tipped by the receiver into a defender's hands. QB's fault? No, but it has a negative impact on his statistics.

Quarterback takes the snap, LT blows the edge and allows the rush to get through, QB sails it out of bounds. QB's fault? No, but it has a negative impact on his statistics.

It works in the other direction too. 

Quarterback throws the ball right to a DB because he misread a coverage. DB drops it. QB's fault? Yes, but no negative impact.

TLDR: Football is the toughest sport to analyze statistically. We're nowhere even close to being able to do it.

And that is what makes it the most fun and sometimes the most frustrating to watch...great team with "bad" QB...bad team with a "great" QB...so many permutations and outcomes that are unpredictable...a team sport dictated by individual performances or individual performances working as a team??

I love the game even with all the warts and challenges (like guys getting their brains injured) - that's another topic though.

[Mike]

1 minute ago, 17to85 said:

And for these reasons maybe we need to stop trying. I mean it's a position that really does rely on so many other people doing their job correctly it seems like a lot of wasted effort trying to distil an individual into a catch all stat that tells how good or bad they are. Can't we just watch the games and judge who is good and who isn't?

I've long given up. I have better things to do. Like drink.

[17to85]

plus with all the ambiguity in what stats matter for a qb it makes it a hell of a lot easier to argue about who is better, which pairs nicely with drinking. 

[blue_gold_84]

3 hours ago, 17to85 said:

And for these reasons maybe we need to stop trying. I mean it's a position that really does rely on so many other people doing their job correctly it seems like a lot of wasted effort trying to distil an individual into a catch all stat that tells how good or bad they are. Can't we just watch the games and judge who is good and who isn't?

But analysis to that degree is important and does provide entertainment for some. It can also help players understand how to improve/fine tune their personal game.

Mike does make a really good point about drinking, though.

[Noeller]

over analyzing stats is just a way for nerds to try and fulfill their dreams of fitting in with the cool people....

[Noeller]

Btw, for you stat nerds...here's a thing from Derek Taylor......

 

Derek Taylor  ?? ?‏Verified account @DTonSC

37 days to the @CFL. Love this graph: % of 2nd-down passes thrown at or past the 1st-down mark. @BCLions Jennings leads all. #Details

 

[17to85]

3 hours ago, blue_gold_84 said:

But analysis to that degree is important and does provide entertainment for some. It can also help players understand how to improve/fine tune their personal game.

 

No the only reason people want these stats is so that they can make definitive claims without having to watch every minute. Players don't need them to know what they need to improve on, it's purely so you can make a judgement call on someone without having spent an exhaustive amount of time watching the player. 

[blue_gold_84]

25 minutes ago, 17to85 said:

No the only reason people want these stats is so that they can make definitive claims without having to watch every minute. Players don't need them to know what they need to improve on, it's purely so you can make a judgement call on someone without having spent an exhaustive amount of time watching the player. 

Uh... No, that's not it.

What about advanced statistics is so offensive to you? And for anyone compiling advanced statistics, there is a requirement to spend time watching the sport - exhaustively. It's how statistics of any kind are compiled in the first place.

[Atomic]

23 minutes ago, blue_gold_84 said:

Uh... No, that's not it.

What about advanced statistics is so offensive to you? And for anyone compiling advanced statistics, there is a requirement to spend time watching the sport - exhaustively. It's how statistics of any kind are compiled in the first place.

Not true, the people compiling the stats are not the same ones analyzing them

[Noeller]

I'd listen to a case that most advance stats nerds don't actually understand what they're watching......you can watch all the tape you want, but it doesn't help you understand it if you've never really played the game, or at the very least have been around competitive sports and spent time "in the room". I'm sure there's some overlap, but it's the exception rather than the rule...

[blue_gold_84]

9 minutes ago, Atomic said:

Not true, the people compiling the stats are not the same ones analyzing them

That's not necessarily true. Anyway, my point is compiling stats still requires watching the game. And in order to compile advanced statistics - that is to say, statistics beyond "standard" numbers we see now in football - one has to actually watch the game.

Edited May 16, 2017 by blue_gold_84

[blue_gold_84]

2 minutes ago, Noeller said:

I'd listen to a case that most advance stats nerds don't actually understand what they're watching......you can watch all the tape you want, but it doesn't help you understand it if you've never really played the game, or at the very least have been around competitive sports and spent time "in the room". I'm sure there's some overlap, but it's the exception rather than the rule...

How did you come to that conclusion?

[Noeller]

1 minute ago, blue_gold_84 said:

How did you come to that conclusion?

As I said, all I'm saying is "I'd listen to a case for" that theory.....it would make sense......nerds like to think they understand sports, but unless you've been there, in the room, on the bench/sideline.....you don't really "get it". I would guess that most advance stats nerds are guys who suck at sports but want to be cool and say they're involved in sports, so here's there way to put nerd skills to use. 

Again, just saying I'd listen to the case for that......

[Floyd]

3 hours ago, Noeller said:

Btw, for you stat nerds...here's a thing from Derek Taylor......

 

Derek Taylor  ?? ?‏Verified account @DTonSC

37 days to the @CFL. Love this graph: % of 2nd-down passes thrown at or past the 1st-down mark. @BCLions Jennings leads all. #Details

 

This is more of a commentary on Offensive Coordinators than anything...