(11-28-2021, 09:46 PM)Shake n Blake Wrote: Yeah he was bad today, but generally speaking he revived his career in Tennessee and played great for the most part.

I'd love to see a study on good QBs with a good run game vs bad. Particularly interesting would be good QBs who went from having a great run game to a poor one.

Ben built a HOF resume with a great run game. Since the Steelers started leaning on him more over the past decade, he hasn't achieved much. Some empty calorie numbers, I guess.

Of course, I guess that's delving more into team success rather than isolating pass success.

[color=var(--darkreader-inline-color)]There is a summary at the very bottom of this post if you'd like to skip to the answer. To begin, I am using nflfastR and the programming language R to gather and analyze all of my data. nflfastR is a library of play-by-play data from 1999-current and R is a programming language used for statistical analysis. I am running my regressions using the data pack add-on within Excel. I gathered a dataset that contained every game and every pass attempt for the following QBs...[/color]

Matt Ryan
Tony Romo
Matt Stafford
Peyton Manning
Tom Brady
Drew Brees
Ben Roethlisberger
Russell Wilson
Kirk Cousins
Dak Prescott
Patrick Mahomes

I have this data grouped by QB, Season, Week so it gives per game data. Next, I gathered every teams rushing data from 2010-current and grouped it the same way, so we have per game rushing metrics going all the way back to 2010. Finally, I added a column onto my passing numbers and assigned the rushing metrics to each game so we could have all of the applicable rushing and passing data together. 

The total column is the total passing attempts for that QB in the game. Next, I ran a few regressions to find the relationship between a few metrics to see if there is any correlation. I am going to post the results below for the metrics that I tested.  Here is what we have...

YPC (Yards per carry) to Passer Rating - Very weak relationship. Over the course of our dataset, only 1% of the variation in a QBs passer rating can be explained by YPC.

YPC to YPA (Yards per attempt) - Very weak relationship. Again, over the entirety of our dataset, only 1% of the variation in a QBs passer rating can be explained by YPC. 

Rush Attempts to Rating - Weak relationship. Roughly 6% of the variation in a QBs passer rating can be attributed to rush attempts. 

Rush Attempts to YPA - Same as relationship to rating. Weak relationship, 6% of variation. 

Total Rushing Yardage to Rating - Weak relationship. Only 5% of the variation in a QBs rating can be explained by total rushing yardage. 

Total Rushing Yardage to YPA - Weak relationship. Only 4% of the variation in YPA can be explained by total rushing yardage. 


I find it interesting that the total rushing attempts and rushing yardage have a stronger relationship than YPC. One explanation I have for this is that it could just be late game rushing attempts that are salting the game away in a victory where the QB had a good game. In total, the answer appears to still be no, a strong running game doesn't seem to have an effect on a QBs performance. The running game and passing game remain independent. For every game where the QB had a good game passing and the team did well running, there is a game where the QB had a good game passing and the team did horribly running the ball. If you want me to compare other metrics or provide numbers in a different way, let me know and I will put it together. I thought about creating a thread but this isn't Bengals related, so I didn't want to clog the board. [/size][/color]

TL;DR - No, running game success doesn't correlate to passing game success. 

Thanks for doing it here. It made the thread way more interesting and gave it more worth..

(11-28-2021, 11:35 PM)Tony Wrote: Thanks for doing it here. It made the thread way more interesting and gave it more worth..

I can’t tell if you’re being genuine or facetious, but thank you.

I agree. Burrow with no run game would be an INT machine. Hed win you a few but he would be throwing INTs left and right.


Mixon and this team may be this year's Titans.



Only diff is we can actually shift into a pass first team at any time

Agree as Mixon goes the offense goes, he runs hard with passion and does not turn ball over...takes pressure off Burrow

(11-29-2021, 12:24 AM)KillerGoose Wrote: I can’t tell if you’re being genuine or facetious, but thank you.

Genuine

(11-29-2021, 12:25 AM)Housh Wrote: I agree. Burrow with no run game would be an INT machine. Hed win you a few but he would be throwing INTs left and right.


Mixon and this team may be this year's Titans.



Only diff is we can actually shift into a pass first team at any time

They were saying we are the most complete team in the AFC on Sunday night football in America.. Brees and Dungy agreed. There is no glaring weakness...

(11-28-2021, 09:41 PM)KillerGoose Wrote: Tannehill is actually an interesting point. Today, he had two 100 yard rushers and yet shit the bed. A dominating run game, but a horrific performance from him. Of course, you're right about the variables. Maybe that is something I could look into. Isolate QBs that are universally considered 'good' over the past eleven years and then study if their performances are benefitted by strong running performances. That is a study I haven't done. Not sure if you're interested, but I'll do that real quick and post the results. That could be an angle that produces something. 

In Dan Marino's entire career, he played one season with a 1,000 yard rusher(answer below). Now go back and look at his numbers in a more defensive friendly era that relied on the running game much more.



















Karim Abdul Jabar ran for 1,116 yards in his rookie season in 1996, Dan Marino's 14th NFL Season. The Dolphins went 8-8.

(11-28-2021, 09:41 PM)KillerGoose Wrote: Tannehill is actually an interesting point. Today, he had two 100 yard rushers and yet shit the bed. A dominating run game, but a horrific performance from him. Of course, you're right about the variables. Maybe that is something I could look into. Isolate QBs that are universally considered 'good' over the past eleven years and then study if their performances are benefitted by strong running performances. That is a study I haven't done. Not sure if you're interested, but I'll do that real quick and post the results. That could be an angle that produces something. 

I had a 2 point post in regards to the statistical relationship between run and pass, etc.

First.
I recall several years ago the Patriots had an almost nonexistent running game. However, several analysts stated that their short passing game to the RBs, screens and such, were basically long handoffs, and that those plays were in fact, running plays.
With that in mind, reorienting analysis to categorize pass plays behind or near the LOS as running plays (most especially if said play involves blocking for the receiver by members of the OL), does that change any of the correlations you've been mentioning? 

2nd:
Mixon has had 2 very effective games. In both games, the passing numbers were a bit pedestrian,  but efficient. You can bet that upcoming opponents will be attempting to create a game plan that will stop the run, which will probably benefit the passing game. So in upcoming games, the running game may look pedestrian,  but the passing game will improve. In this way, each type of offense will benefit one another over time. There may be a game to game " game planning effect " that leads to ultimate success.

Extra point:
How many air Coryell type,  pass almost exclusively offenses have won a superbowl? Especially keeping point number one in mind?

Sorry Tony, this may need it's own thread...

(11-24-2021, 02:40 AM)J24 Wrote: I don't agree with that I think the improved offensive line, a healthy Joe, and of course a true #1 WR in Chase.

You are right, and so is he.  It is all those things, Mixon, the defense, coaching, culture, and overall team health.  Being able to pull your starters on defense a quarter early after an already "light" day at Vegas (didn't know there was such a thing? Haha) with just 44 snaps on defense also helps to keep everyone fresher for the next game.

This team is set up to make a run and contend for the AFC North title.  Do that, and there should be a lot of awards to pass around, but I think the division crown is what they really want.   That, and a ticket to the dance.  Burrow with an offense that can beat you multiple ways and a defense that has allowed 23 points the last two games combined?  I will take that :)

(11-30-2021, 07:10 AM)SladeX Wrote: I had a 2 point post in regards to the statistical relationship between run and pass, etc.

First.
I recall several years ago the Patriots had an almost nonexistent running game. However, several analysts stated that their short passing game to the RBs, screens and such, were basically long handoffs, and that those plays were in fact, running plays.
With that in mind, reorienting analysis to categorize pass plays behind or near the LOS as running plays (most especially if said play involves blocking for the receiver by members of the OL), does that change any of the correlations you've been mentioning? 

2nd:
Mixon has had 2 very effective games. In both games, the passing numbers were a bit pedestrian,  but efficient. You can bet that upcoming opponents will be attempting to create a game plan that will stop the run, which will probably benefit the passing game. So in upcoming games, the running game may look pedestrian,  but the passing game will improve. In this way, each type of offense will benefit one another over time. There may be a game to game " game planning effect " that leads to ultimate success.

Extra point:
How many air Coryell type,  pass almost exclusively offenses have won a superbowl? Especially keeping point number one in mind?

Sorry Tony, this may need it's own thread...

To your first point, I could probably look into that. The play-by-play library I use has air_yards as a column, so I could filter a dataset on air_yards <= 1 yard and throw those in with the run data. 

To your second point, this would be something that we could see. In my dataset from 2010-current, I am working with roughly 400k plays. We would be able to see a stronger relationship one way or another. I could break it down into groups based on run game effectiveness and study passing numbers for the game immediately after a big rushing game. Regardless, I don't think we will find anything there, either. A defensive coordinator is much more afraid of Burrow and co. than Mixon. Run games, by nature, tend to be inconsistent. Now, some of them are more consistent than others (Taylor, Chubb). A strong passing game also has a higher correlation to winning games than rushing does.

I'm sure the DC for the Chargers will have situations to handle the run, that will be normal, but I would also wager a bet that the gameplan is going to be centered around slowing the passing attack down. Passing is just significantly more dangerous than running. If a team can hold Burrow to a game like these past two games, they will take their chances with the running game. They will have installed run fits to help contain the rushing attack, but bringing eight men into the box isn't as common as many think. 

EDIT - What specifically are you meaning by Air Coryell? Are you genuinely talking about Air Coryell passing concepts, or just heavy passing teams?

EDIT #2 - I ran the numbers for your first point. The correlations actually become weaker. I took all passes that traveled one yard or less in the air and gathered metrics on those, adding them to the rushing data. Then, I ran the same tests and here were the results...

New YPC to Passer Rating - Very weak relationship, less than 1% of the variation in passer rating can be explained here.

New YPC to YPA - Genuinely ZERO relationship. Correlation coefficient of 0.01. Zero percent of the variation in YPA numbers can be explained by this new YPC number. 

(11-30-2021, 07:10 AM)SladeX Wrote: I had a 2 point post in regards to the statistical relationship between run and pass, etc.

First.
I recall several years ago the Patriots had an almost nonexistent running game. However, several analysts stated that their short passing game to the RBs, screens and such, were basically long handoffs, and that those plays were in fact, running plays.
With that in mind, reorienting analysis to categorize pass plays behind or near the LOS as running plays (most especially if said play involves blocking for the receiver by members of the OL), does that change any of the correlations you've been mentioning? 

2nd:
Mixon has had 2 very effective games. In both games, the passing numbers were a bit pedestrian,  but efficient. You can bet that upcoming opponents will be attempting to create a game plan that will stop the run, which will probably benefit the passing game. So in upcoming games, the running game may look pedestrian,  but the passing game will improve. In this way, each type of offense will benefit one another over time. There may be a game to game " game planning effect " that leads to ultimate success.

Extra point:
How many air Coryell type,  pass almost exclusively offenses have won a superbowl? Especially keeping point number one in mind?

Sorry Tony, this may need it's own thread...

I went ahead and pinged a buddy of mine who has been a DC at the NCAA level and currently coaches for a big 6A school down in Texas (DC there as well). I was curious about how defensive gameplans are drawn up, and here is what he had to say...
 
"The NFL is very personnel heavy as far as their tendencies are concerned. They will say 'okay, they are in this personnel grouping, here are the three formations they get in and here are the plays ran out of it' and make a call that way. The play call sheets I used when I was at the college level were significantly different than what we use here in HS because everything is broken down by personnel groupings."

I think this is interesting. It sheds more light on how an NFL DC creates a gameplan. They don't tailor it to one thing or another, they tailor it based on personnel groupings and go from there. So, really, they aren't gameplanning to stop the pass OR run specifically, they just focus on who is on the field and what you do when you're out there. 

Lots of factors effect a QB's passing numbers that are hard to quantify.

-A team with a weak defense will be behind more. A QB who is trailing will have to take more chances and throw more interceptions. A QB with a lead can play it safe more often. Also when a team is behind late in a game the defensive line can concentrate on just rushing the passer instead of stopping the run.

-Poor pass blocking will effect a QBs numbers. Even if there are not a lot more sacks there will be a lot more short passes or passes thrown away to avoid sacks.

-QBs that play in domes on turf have an advantage.

But even with all of these variables many times a really good QB that plays outside with a poor defense and weak O-line will still outperform a poor/average QB who has everything going for him.

(11-24-2021, 01:23 AM)Tony Wrote: He's 4th in rushing, and I know people will say the defense is the reason they are better than last year. I think It changed the whole dynamic of this offense... The Bengals record has to be high when he rushes around 100 yards... If they can keep sustaining that, they will be hard to stop.. No wonder Mixon loves Pollack...

The O-line is improving every week, in fact, I think they are starting to gel together.  I think they are a much talented bunch than the last 3 season squads.  The biggest different this year is DISCIPLINE and staying HEALTHY.   

(11-30-2021, 02:40 PM)fredtoast Wrote: Lots of factors effect a QB's passing numbers that are hard to quantify.

-A team with a weak defense will be behind more.  A QB who is trailing will have to take more chances and throw more interceptions.  A QB with a lead can play it safe more often.  Also when a team is behind late in a game the defensive line can concentrate on just rushing the passer instead of stopping the run.

-Poor pass blocking will effect a QBs numbers.  Even if there are not a lot more sacks there will be a lot more short passes or passes thrown away to avoid sacks.

-QBs that play in domes on turf have an advantage.

But even with all of these variables many times a really good QB that plays outside with a poor defense and weak O-line will still outperform a poor/average QB who has everything going for him.

Yeah, at the very least we can quantity the domes/no domes aspect of things. We have access to all passing metrics broken down into stadium type. The best that I can think of to do for #1 there is grabbing passing metrics based on differential, which isn't super difficult to do. The database that I use has a differential column that allows you to pull data based on differentials. 

The second one would be tougher because how do you quantify a 'good' line? Pass protection is both line and QB controlled. Maybe use PFF grades. Your final point is spot on, though. 

(11-30-2021, 11:06 AM)KillerGoose Wrote: To your first point, I could probably look into that. The play-by-play library I use has air_yards as a column, so I could filter a dataset on air_yards <= 1 yard and throw those in with the run data. 

To your second point, this would be something that we could see. In my dataset from 2010-current, I am working with roughly 400k plays. We would be able to see a stronger relationship one way or another. I could break it down into groups based on run game effectiveness and study passing numbers for the game immediately after a big rushing game. Regardless, I don't think we will find anything there, either. A defensive coordinator is much more afraid of Burrow and co. than Mixon. Run games, by nature, tend to be inconsistent. Now, some of them are more consistent than others (Taylor, Chubb). A strong passing game also has a higher correlation to winning games than rushing does.

I'm sure the DC for the Chargers will have situations to handle the run, that will be normal, but I would also wager a bet that the gameplan is going to be centered around slowing the passing attack down. Passing is just significantly more dangerous than running. If a team can hold Burrow to a game like these past two games, they will take their chances with the running game. They will have installed run fits to help contain the rushing attack, but bringing eight men into the box isn't as common as many think. 

EDIT - What specifically are you meaning by Air Coryell? Are you genuinely talking about Air Coryell passing concepts, or just heavy passing teams?

EDIT #2 - I ran the numbers for your first point. The correlations actually become weaker. I took all passes that traveled one yard or less in the air and gathered metrics on those, adding them to the rushing data. Then, I ran the same tests and here were the results...

New YPC to Passer Rating - Very weak relationship, less than 1% of the variation in passer rating can be explained here.

New YPC to YPA - Genuinely ZERO relationship. Correlation coefficient of 0.01. Zero percent of the variation in YPA numbers can be explained by this new YPC number. 

Air Coryell is just an example of a pass heavy offense. So Marino Era Dolphins may not have been a Coryell offense, but would also be what I'm driving at, which is that to achieve the ultimate goal, an offense must be competent in the run game. 

(11-28-2021, 06:55 PM)Rubekahn29 Wrote: That line was moving dudes. Easily the best run game I’ve seen from the O-line since ZT has been here.

Yes and I am liking Adeniji a lot lately, especially the last 2 games. Spain has been good all year pretty much.

If this OL keeps this up we will be a tough out in the Playoffs with our RB's, namely Mixon of course.