Outliers, YPC, and the Cowboys running game

[TheDude]

I"d imagine we'd want fewer outliers, right? Just keep the offense in normal situations for the passing game to convert and get your big plays that way when someone in the secondary falls down or blows a coverage or commits PI. Try to avoid the 3rd-and-unlikelies that can stall your drives.

So in an effort to try to get to sleep here in NY (on LA Time), I ran a few rankings - which may not mean much, but was interesting.

To answer your first questions, not necessarily. It think more outliers is better because it should correlate with a drive having a higher probability of ending in points (and could be an outright score. BUT, you should want to be at the higher end of the median (or even mode). I dont have all RB attempt data, so I cant compare that, but it seems like a logical assertion.

However, Chris johnson could be evidence of someone that proves JUST long runs may not mean much. No Tds, few 1st downs, but good 20+ yard runs. THen again he could be an outlier due to stats, NYJ being a turd offensively, etc.

Below, I took the top 26 (because after that attempts got below <150 and ended at Kapernick in the sort) and looked at a % of runs that lead to TD, 1st Down then TD or 1st Down and considered those "effective" for lack of a better term. After that I looked at 20+yard runs (explosive). Finally I summed the Rank for Effective + Explosive. This may not be clean as I didnt have long runs that ended in TV normalized out (i.e. double count in the effective and explosive - but maybe that is ok, need to think about it).

The interesting illustration was how effective per rush Jeremy Hill, Lamar Miller, CJ Anderson and Justin Forsett were last year. Of course there are compensating factors and each back has their own story and team utilization, but for someone like Lamar miller to have the 3rd highest TD or 1st down % and 2nd highest runs >20 yards % was not what I expected.

Anyway. I draw no outcome other than it is a relative touch comparison.

BTW Russell Wilson was 2nd in >20 yards and 1st in Tds and 1st dowsn %, but I excluded QBs.

 

[FuzzyLumpkins]

I"d imagine we'd want fewer outliers, right? Just keep the offense in normal situations for the passing game to convert and get your big plays that way when someone in the secondary falls down or blows a coverage or commits PI. Try to avoid the 3rd-and-unlikelies that can stall your drives.

Well, ideally I would rather have the big plays and the down in and down out performance. Having to choose between the two I am not so sure which I would give up. Big plays lead to points where the other gives game control.

 

[Beast_from_East]

It does, an it's not, because the advanced stats show that running games mean little to winning and losing in the NFL. So people who think otherwise just ignore them.

I think what you might be missing is that the running game can help you pass more efficiently and by passing more efficiently it helps you to win more games. While there is not a direct stat for this, let me explain.

If you have a very good running game (like we did last season), then it forces teams to put 8 guys in the box and thus leaving your receivers like Dez in single coverage. This helps your passing game tremendously.

Otherwise, if you are stating that running the ball is no relationship whatsoever to winning games, then why do teams even bother running then? Why would teams not just pass every down? Additionally, why would a team spend big money on a RB if the running game matters so little?

 

[Idgit]

Well, ideally I would rather have the big plays and the down in and down out performance. Having to choose between the two I am not so sure which I would give up. Big plays lead to points where the other gives game control.

My thinking was that I'd rather have the big plays, but they're more likely to come from the passing game, so the goal would be to get as many effective passes as you can get, and that happens if you keep your downs and distance manageable, but it's tough to say.

 

[Beast_from_East]

I didn't arbitrarily do *anything.* Nobody's taking anything away from Murray, and nobody's saying he didn't have other long runs or significant plays. The 7 plays were the number of his longest runs that separated his production from the production of the average for his position. But *again,* this isn't commentary on Demarco Murray. It's commentary on what the distribution of carries for a productive NFL running back actually looks like.

One thing that has not been touched upon yet.....................if the average backs get around 4.0 and it is only 6 or 7 big runs that skew the average up to 4.5 or even 5.0 for the top level backs.............this begs the question..........

How bad does McFadden actually suck to have multiple seasons of 3.3 YPC..............clearly McFadden would be considered highly below average.

 

[TheDude]

Oh, god. She has no idea he exists. We don't talk about any of my weird addictions.

If she knew how much of my time I spent arguing with people on the internet, she'd tell me to get a life. And we've already covered that she would, of course, be right.

Did you just call @Risen Star an addiction?

]

 

[Idgit]

So in an effort to try to get to sleep here in NY (on LA Time), I ran a few rankings - which may not mean much, but was interesting.

To answer your first questions, not necessarily. It think more outliers is better because it should correlate with a drive having a higher probability of ending in points (and could be an outright score. BUT, you should want to be at the higher end of the median (or even mode). I dont have all RB attempt data, so I cant compare that, but it seems like a logical assertion.

However, Chris johnson could be evidence of someone that proves JUST long runs may not mean much. No Tds, few 1st downs, but good 20+ yard runs. THen again he could be an outlier due to stats, NYJ being a turd offensively, etc.

Below, I took the top 26 (because after that attempts got below <150 and ended at Kapernick in the sort) and looked at a % of runs that lead to TD, 1st Down then TD or 1st Down and considered those "effective" for lack of a better term. After that I looked at 20+yard runs (explosive). Finally I summed the Rank for Effective + Explosive. This may not be clean as I didnt have long runs that ended in TV normalized out (i.e. double count in the effective and explosive - but maybe that is ok, need to think about it).

The interesting illustration was how effective per rush Jeremy Hill, Lamar Miller, CJ Anderson and Justin Forsett were last year. Of course there are compensating factors and each back has their own story and team utilization, but for someone like Lamar miller to have the 3rd highest TD or 1st down % and 2nd highest runs >20 yards % was not what I expected.

Anyway. I draw no outcome other than it is a relative touch comparison.

BTW Russell Wilson was 2nd in >20 yards and 1st in Tds and 1st dowsn %, but I excluded QBs.

This is awesome. I've seen Anderson, Hill, and Miller on similar lists of RB effectiveness, and now I know why. Running on 75% of first downs probably hurt Demarco a bit. I guess that's why his first down rate is lower than I'd have guessed it was. We've got a lot of options in the passing game, and Dez in the end zone, so top 10 is about what I'd have guessed on the td rate. Knowing his counts for 20 yard runs, you can see how taking away a few big plays really affects a back's numbers.

I guess we can put an end to the idea of trading Leary for Lamar Miller, huh?

 

[TheDude]

This is awesome. I've seen Anderson, Hill, and Miller on similar lists of RB effectiveness, and now I know why. Running on 75% of first downs probably hurt Demarco a bit. I guess that's why his first down rate is lower than I'd have guessed it was. We've got a lot of options in the passing game, and Dez in the end zone, so top 10 is about what I'd have guessed on the td rate. Knowing his counts for 20 yard runs, you can see how taking away a few big plays really affects a back's numbers.

I guess we can put an end to the idea of trading Leary for Lamar Miller, huh?

I watched a couple of Dolphin games early and Miller didnt stand out. I am really curious about that one. He did have a 97yrd TD, but excluding that was still pretty darn good

 

[Idgit]

Did you just call @Risen Star an addiction?

]

I'm going to get even with you for this, McLovin, if it's the last thing I do.

 

[Idgit]

I watched a couple of Dolphin games early and Miller didnt stand out. I am really curious about that one. He did have a 97yrd TD, but excluding that was still pretty darn good

If there's one thing I've learned from this thread, you can't just go excluding people's long runs, McLovin. It's not right.

Also, I'm going to get even with you for that Brokeback picture if it's the last thing I ever do, btw. You wait.

 

[Hoofbite]

RB numbers are universally based on large chunks of yardage that make up for carries that get little to nothing. Take away Peterson's 8 best runs from 2012 and his average drops by 1.35 yards/carry. Take away the 8 best runs from every starting running back and every player's average drops. Hell, take 8 of Emmitt's best runs for every year he played and he never passes Walter Payton.

It is kind of an interesting take, but I'm not sure it's something that should lead people to think that the difference between a top back and an average back is a mere 7 or 8 carries. What about the numerous carries where one back gets 4, 5 or 6 tough yards and the alternative only gets 2, 3 or 4 tough yards? What about the carries where one back gets nothing, but a lesser guy loses yards?

If you look at the top 20 backs from last year, their average is 4.4 YPC. Remove the 7 best runs from everyone and the average drops to 3.6 YPC. Murray is at 4.15 YPC without his 7 best runs, which is "average" if you consider the entire league with their best runs, but is also over 0.5 yards better than the top 20 when you've subtracted their best runs. Out of those 20 backs, only 3 averaged more than 4.0 YPC after removing their 7 best runs. Only 1 (Murray) rushed for more than 1100 yards after removing the 7 best runs from each. Additionally, if you compare each player's average to what their average would be without their 7 best runs, Murray's difference between the two is the 2nd smallest. 18 other backs had their averages hurt more than Murray did.

Not sure I see the point in comparing a guy absent his best to the average when their best is included. The league average would drop if you subtracted everyone's best plays and you'd likely end up with a situation similar to where you started.

The best 7 runs a RB has in a season inflate their numbers, but it's the other few hundred runs that make their average.

 

[jobberone]

Sorry this is incredibly flawed.

If you want to glean something meaningful take the top 10 runs away from EVERY runner and then compare them.

Even the guys averaging pedestrian numbers had some long runs.

You boys understand the difference between average and median, right? There is a very important distinction between these two math terms and people often don't get it.

Yes it would be interesting to compare the MEDIAN runs of all backs in the league. I don't know if that stat exists.

You should look at averages and medians but it was an interesting article.

Many coaches are aware of the fact you need an effective running game, a RB that creates concern from the defense with a shift in coverage and packages, individual skill sets and the coaches ability to exploit strengths and minimize weaknesses of them, and a commitment to the run.

It is normal, as evidenced by the league wide results, to only get X numbers of big and negative plays from a RB. In fact looking at wins and losses the biggest factors are big plays both positive and negative whether run, pass or ST both O and D.

Both Walker and Dorsett changed defenses with the ability to break runs. This aspect of altering defensive schemes can't be underestimated.

Median runs gives you an idea of the consistency of handing the ball off to a RB. Average runs gives you the overall picture. If you analyze the data to give you median and look at the blips on the negative and positive side then you get more of an idea.

 

[ABQcowboyJR]

No, it wouldn't. You've got a distribution that is non-normal and will always be so as runs of 5+ yards are fairly common but the loss of 5 or more yards is rare.

All looking at SD would tell us is whether someone failed to understand what they were taught in Stat 101

I stand corrected jackwagon. I guess it would really depend on how skewed the data was.

 

[Doomsday101]

Never been concerned so much with the big run, there are many situation in a game where a RB is asked to pick up 2 yards to move the chains failure means losing possession as you are forced to punt the ball. The plays often are not going to give you 20 yards but did you pick up the 2? if you did the drive is alive and the offense chance of score increases. Yes it is a measly 2 yard gain but it is what keeps you on the field.

 

[jterrell]

I think what you might be missing is that the running game can help you pass more efficiently and by passing more efficiently it helps you to win more games. While there is not a direct stat for this, let me explain.

If you have a very good running game (like we did last season), then it forces teams to put 8 guys in the box and thus leaving your receivers like Dez in single coverage. This helps your passing game tremendously.

Otherwise, if you are stating that running the ball is no relationship whatsoever to winning games, then why do teams even bother running then? Why would teams not just pass every down? Additionally, why would a team spend big money on a RB if the running game matters so little?

Dallas lost 5 games last year.
In 1 of them Romo was injured.
In another he was out.
In the playoff game a league rule of epic football stupidity cost us a long passing TD.

Even in a historic rushing year our limited failures were tied to Tony Romo's health and the passing game.

Teams VERY seldom pay RB big money. Those that do VERY seldom win.
Adrian Peterson has played in 4 playoff games and he is a generational RB.
Barry Sanders has often been called the best back of all-time... he played 6 playoff games, and WON ONLY 1.

Asking why teams do things is pointless because it assumes they can't make mistakes yet they obviously do.
Minnesota traded us the world for Herschel Walker and Ditka traded the world for Ricky Williams. Were those smart moves?

 

[Doomsday101]

Dallas lost 5 games last year.
In 1 of them Romo was injured.
In another he was out.
In the playoff game a league rule of epic football stupidity cost us a long passing TD.

Even in a historic rushing year our limited failures were tied to Tony Romo's health and the passing game.

Teams VERY seldom pay RB big money. Those that do VERY seldom win.
Adrian Peterson has played in 4 playoff games and he is a generational RB.
Barry Sanders has often been called the best back of all-time... he played 6 playoff games, and WON ONLY 1.

Asking why teams do things is pointless because it assumes they can't make mistakes yet they obviously do.
Minnesota traded us the world for Herschel Walker and Ditka traded the world for Ricky Williams. Were those smart moves?

Commanders made a blockbuster trade with the Rams, sending three first-round picks and a second-round pick to St. Louis in exchange for the No. 2 overall pick in the NFL draft. Washington would use the pick to draft Robert Griffin III.
http://www.washingtonpost.com/blogs...ear-anniversary-of-the-Commanders-rams-trade/

Was that a smart move?

 

[jterrell]

Commanders made a blockbuster trade with the Rams, sending three first-round picks and a second-round pick to St. Louis in exchange for the No. 2 overall pick in the NFL draft. Washington would use the pick to draft Robert Griffin III.
http://www.washingtonpost.com/blogs...ear-anniversary-of-the-Commanders-rams-trade/

Was that a smart move?

Nope, it was not.
NFL teams are often stupid.
In most trades one team wins and another loses.

But the Commanders were correct to identify needing a QB as primary to winning.
They just didn't go about getting one the right way or insuring it was the right QB.

One thing they do have going for them is RG3 is still a QB and he doesn't have a 4 year shelf life via position (tho he may due to injury).
So if he ever stops sucking and gets healthy he can make a very real impact on that franchise either via wins, in trade or both.
They have had plenty of success with a bottom of the draft RB.

Wonder why they aren't winning more?

 

[BlindFaith]

Dallas lost 5 games last year.
In 1 of them Romo was injured.
In another he was out.
In the playoff game a league rule of epic football stupidity cost us a long passing TD.

Even in a historic rushing year our limited failures were tied to Tony Romo's health and the passing game.

Teams VERY seldom pay RB big money. Those that do VERY seldom win.
Adrian Peterson has played in 4 playoff games and he is a generational RB.
Barry Sanders has often been called the best back of all-time... he played 6 playoff games, and WON ONLY 1.

Asking why teams do things is pointless because it assumes they can't make mistakes yet they obviously do.
Minnesota traded us the world for Herschel Walker and Ditka traded the world for Ricky Williams. Were those smart moves?

I get what you're saying, but what has Romo won? Does that mean QB isn't important. Looking at win loss to determine a specific players effectiveness is not, well...effective.

 

[plasticman]

My advice is to beware of articles written by those that are too proud of the C- they got in "Statistics for Journalism Majors"

I'm giving Kiper the benefit of the doubt here.

Yes, take away Murray's top nine runs and his average per carry becomes 4.14.

Now do the same for those RB's with an average carry around 4.14.

Chris Ivory drops to 3.4.

Anthony Dixon drops to 2.8.

Take away every RB's top nine runs and guess what?

Murray is elite again.

 

[jterrell]

I get what you're saying, but what has Romo won? Does that mean QB isn't important. Looking at win loss to determine a specific players effectiveness is not, well...effective.

Romo is more than 25 games over .500 for his career playing largely with dregs.
He has made it to 3 divisional playoff games.

If our best player had been the running back we would've basically lived through the Dave Campo era all over again with 5 wins a year.
Even Emmitt knew you couldn't win with a diamond at RB surrounded by poo.