Writers of Pro Football Prospectus 2008

printable version

FO Mailbag

We've got mail

Need Help with KUBIAK or Premium?

blog advertising is good for you


26 Jan 2010

FO Mailbag: Run-Pass Ratio in Short Yardage

Craig Cunningham: Do you track how often a team runs or passes by down and distance. I am specifically interested in how often the Steelers ran on third- or -fourth-and-short. Steelers fans seem to hate the Offensive Coordinator and one of the common complaints is that he doesn't run enough on third-and-short. I would be interested to know how the Steelers rank against the rest of the league in this situation.

Aaron Schatz: Sure, I went and looked at the run-pass ratios for every team in what we call power situations: third down, fourth down, or the goal line with 1-2 yards to go. The average NFL team ran in 63 percent of these situations (including DPI calls as passes). The Pittsburgh Steelers ran in 66 percent of these situations. So Bruce Arians may not be grinding it as much as Steelers offensive coordinators of the past, but this isn't 1975 anymore. He's not particularly pass-wacky compared to other NFL coordinators.

Seven teams ran the ball in at least 70 percent of power situations, in order: Jacksonville, New England, New York Jets, New York Giants, New Orleans, Carolina, and Cleveland. Apparently, if you want to run in a power situation, it helps to have "New" in your name. Notice that the pass-heavy Patriots and Saints are run-centric in only situation on the field where the run is generally more successful than the pass.

Six teams ran the ball in fewer than 55 percent of power situations, in order from highest to lowest: Tampa Bay, Kansas City, Washington, Arizona, Chicago, and Green Bay. Most of these teams didn't run in power situations because they were awful in power situations, but Green Bay is a total mystery. The Packers ranked third in success running in short yardage, yet they were the only team in the league to pass more often than run in these situations.

Another interesting stat: Miami (85) and New England (82) were in the most power situations this year. Buffalo (40) and Tennessee (42) were in the fewest.

Posted by: Aaron Schatz on 26 Jan 2010

50 comments, Last at 27 Jan 2010, 10:41pm by Dork

Comments

1
by Nathan :: Tue, 01/26/2010 - 3:38pm

Are there stats for Philly this year vs last year? I'd be interested in seeing the numerical degree of the Weaver effect.

2
by andrew :: Tue, 01/26/2010 - 3:41pm

Now only if we knew the % of plays that were called runs but audibled to a pass...

49
by commissionerleaf :: Wed, 01/27/2010 - 3:51pm

Cough, Vikings, cough.

3
by ten.gtd (not verified) :: Tue, 01/26/2010 - 3:44pm

Don't we have to go one level deeper and look at the correlation between expected vs actual yds gained in these situations to understand whether or no teams are making the right decisions for their team at that time?

8
by Pat (filler) (not verified) :: Tue, 01/26/2010 - 4:13pm

whether or no teams are making the right decisions for their team at that time?

There's no way to know that anyway. You have no way of knowing whether the success of a given play (say, a run) was set up by the prior plays. As a simple example, consider personnel - if the prior plays forced the defense to stack the line, running is a bad option. If the prior plays forced the defense into nickel, passing is a bad option.

Those are obvious extremes, but you see the point. Judging coaching decisions is really, really hard.

41
by Kulko :: Wed, 01/27/2010 - 3:52am

Quote: Judging coaching decisions is really, really hard.

No its really realy easy, thats why everybody is doing it.

Now Judging coaching decisions rightfully is extremely hard, but who wants to do complicated things.

4
by Theo :: Tue, 01/26/2010 - 3:49pm

"The average NFL team ran in 63 percent of these situations (including DPI calls as passes). "

Pass interference is included with the runs?

6
by Marko :: Tue, 01/26/2010 - 4:02pm

I assume the parenthetical refers to the fact that penalties are officially "no play," so they are not recorded as passes in official stats.

9
by Aaron Schatz :: Tue, 01/26/2010 - 4:17pm

Right, so the only penalty that I included in the count of "short yardage plays" was DPI. But they count as passes, not runs.

5
by JasonK :: Tue, 01/26/2010 - 3:59pm

Another complicating point is the granularity of the official measurements. 37 inches and 72 inches both go in the books as "3rd & 2," but that gap makes a huge difference in the prospect of a "power run" being successful. Running is probably the best move in the former situation, but for a lot of teams, 3rd & 72 inches is a passing down.

Tangentially related: I just looked over the "power success" stats, and the bottom of the league sure has some outliers there. Teams ranked 20-30 are all between 60% and 56%, then it drops to Buffalo at 50%, and San Diego at 46%!

29
by Arkaein :: Tue, 01/26/2010 - 5:46pm

Although I agree with you point in general, 37 inches to go will be called 1 yard, at least going by how it's called on TV. I think they usually round to the nearest yard greater than zero.

7
by countertorque :: Tue, 01/26/2010 - 4:07pm

My complaint with Arians is how often he passes in power situations, but the high percentage of times that he lines up in shotgun or empty backfield in these situations. Isn't that the last situation where you want to show pass prior to the snap?

10
by Jmagik (not verified) :: Tue, 01/26/2010 - 4:23pm

Eh, maybe. Cowboys seemed to have more success in short yardage lining up in shotgun or taking long drops to delay handoffs with Jones/Choice... versus handing to Barber and watching him fall down at the line. Granted, this has to do with (a) Cowboys o-line's strengths/weaknesses and (b) Tony Romo being undoubtedly one of the best QBs in the league at selling the draw/play-action (if you think I'm being a homer here, offer evidence... the guy has flaws, but he is incredible at this one thing).

12
by drobviousso :: Tue, 01/26/2010 - 4:46pm

I believe the Steelers ran a draw that one time. And that's about it.

EDIT: I should specify, this isn't nesisarily a bad thing. A good draw play requires your line to pass block for a couple of seconds, and your RB has to be patient have have good vision. This has not been what you would call a Steeler's strength for the past couple of years.

46
by Mister Asterisk (not verified) :: Wed, 01/27/2010 - 1:58pm

My biggest beef with Arians is that when the Steelers spread four or more wide, they almost always seemed to pass the ball (often with an empty backfield), whereas when they only split two wide, it was almost always a run. I would be curious to know if others saw the same thing, esp. the game charters. There seemed to be an almost unforgivable lack of guile in using formation to disguise intent, particularly given the general level of success when play-action was called. Watching on television, it certainly seemed easier to predict the playcall from looking at the formation than during the Mularkey and Whizz years.

48
by CraigInDC :: Wed, 01/27/2010 - 3:40pm

It would be great to get real numbers on this sort of thing. I never trust my own perception of such things because I feel it is too easily swayed by the exceptional cases. I'm not sure if it is always four wide, but they run the 3 WR bunch formation a lot, and they run out of that formation. I think that formation, at least, gives the offense some deception.

15
by CraigInDC :: Tue, 01/26/2010 - 4:48pm

I don't know if it's the last situation that you want to show pass prior to the snap or not. The point here is that the Steelers are often successful when they run (5th in the league in Power Success), and they choose to run more often than average. Perhaps they should run even more often than they do, but it seems a little strange to complain about the strategic choices of a team that is the 5th most successful in the league at the situation that you are complaining about. Perhaps they are so successful because they show pass?

16
by drobviousso :: Tue, 01/26/2010 - 4:55pm

Game theory suggests they should run more (until their success @ passing and running from that distance are = ). They are good now, but they could be gooder.

19
by CraigInDC :: Tue, 01/26/2010 - 5:05pm

I don't know what there success rate when passing in these situations is, so I can't tell if they should run more often. Maybe their success rate when running in this situation is equal to their success rate when passing.

26
by drobviousso :: Tue, 01/26/2010 - 5:16pm

Fair point. I don't know either.

20
by CraigInDC :: Tue, 01/26/2010 - 5:05pm

I don't know what there success rate when passing in these situations is, so I can't tell if they should run more often. Maybe their success rate when running in this situation is equal to their success rate when passing.

33
by Pat (filler) (not verified) :: Tue, 01/26/2010 - 9:27pm

That's not really true. It'd be true if the entire point of the game were to convert third downs, and all you had were third downs, but that's not the game. Close to the goal line, converting a third down may not be nearly as important as avoiding a turnover, for instance, especially if they're moving the ball well or you're not.

There are subtler problems as well - namely, the fact that the success rate of a third down play is likely strongly dependent on what happened the previous plays, so even if running is a better option on an average third down and short, if you've run, say, twice before, the success rate of running on that play might be far worse.

The behavior of stochastic repeated games is a lot different than one-shot games.

11
by milo (not verified) :: Tue, 01/26/2010 - 4:34pm

Notice that the pass-heavy Patriots and Saints are run-centric in only (sic) situation on the field where the run is generally more successful than the pass.

Notice that the Patriots are the 21st most likely team to run in all situations and that the Saints are the 10th most likely team to run in all situations. Neither team is especially pass-heavy. They are better described as pass-successful. The Saints could also be described as a good running team, especially as there is a website on the intertubes that has a statistic called DVOA and there the Saints are ranked number one in rushing DVOA.

30
by tuluse :: Tue, 01/26/2010 - 7:39pm

Patriots are 21st most likely to run, and you wouldn't call that pass-heavy?

You have strict standards.

34
by milo :: Tue, 01/26/2010 - 9:47pm

The league average is to pass 54.8% of the time. The median is 55.7%. Standard deviation is 4.8%. NE passed 56.0% of the time. NO passed 53.8% of the time.

The NYJ are not pass happy at 39.3%. Nor are CAR at 47.0%.
IND at 62.2% and ARI at 61.9% are the most pass happy.

35
by Brendan Scolari :: Tue, 01/26/2010 - 9:50pm

Just wondering, but where did you find those stats? I've always wondered how you could see pass/run ratios and the like.

38
by milo :: Tue, 01/26/2010 - 10:28pm

Copied the team rushing and team passing stats from nfl.com to excel. Combined the tables with VLOOKUPs and wrote a couple math formulas. Yes, I've wasted too much of my life as an excel jockey.

45
by Mr Shush :: Wed, 01/27/2010 - 10:56am

But both those teams are good, and therefore frequently in a position to favour the run in order to use more time, rather than because they think it is the best way to gain yards, first downs or scores. I think when people talk about a team being run/pass-heavy, they're likely to mean it in a situation-neutral sense. FO probably has better and more sophisticated ways to filter out the effect of situation, but a probably adequate rough-and-ready approach would be to consider run:pass ratio in the first half only.

13
by drobviousso :: Tue, 01/26/2010 - 4:47pm

Steelers nation is a bit divided on why we don't like Arians (assuming they don't). One camp doesn't like him because he doesn't run enough. One camp doesn't like him because he runs too often. One camp doesn't like his choice of run/pass balance. I think the first two camps are silly. He ran when he had a much better O-line, and passed more when his O-line was bad and had a better WR corp. Incidently, Rooney is in the first camp, FWIW.

The last dislike camp thinks he just telegraph's his play calling, and doesn't adjust to opponents strengths and weaknesses. The Cleveland game this year, on a frigid windy night vs a bad run team featured a lot of passing. That God-(^&*(^*(%^%^&& game against Denver two years ago was the same thing. Also, he loves the "two runs up the middle for 2 yards, then third and long" pattern, which may not be the best use of the Steeler's offense. How we beat the Ravens 4 out of the last 5 times with that, I'll never know.

Most of the camp that likes Arians usually cites Ben's 4K season and the 1K season of three 'skill' position players as proof that his system is a very good one. I like to call them "statistically illiterate."

18
by CraigInDC :: Tue, 01/26/2010 - 5:03pm

I don't particularly like Arians, but I haven't heard a complaint that I agree with either. I agree that the first two camps that you name are silly. However, I think that the third camp is just as silly. It is based on a few bad games which can be countered with a few great games. I can't defend anything any Steeler player or coach did in the Browns game. However, I think complaint that he likes to run, run, pass too often would not be backed up by the facts if anyone chose to tally the plays. I've heard this complaint about specific games, and when I looked at the play-by-play of those games, they didn't actually have a run, run, pass pattern at all. The contention that he doesn't adjust to opponents is pretty hard to quantify, so it is an easy thing to fall back on when things didn't work out.

I think fans don't like Arians because they have to have someone to hate. They can't hate Ben because he's a god. They can't Tomlin because he's just so darn cool. They can't hate Lebeau because he's a legend. They can't hate the Rooneys because they are such great people. That leaves Arians and Limas Sweed to hate.

23
by drobviousso :: Tue, 01/26/2010 - 5:14pm

I read somewhere before the season started that in the couple of years previously, Pittsburgh ran more in second and long more often than most other teams. Trying to locate it, but I don't remember where it was.

For the record, I don't hate Arians, I just think he's a better position coach than play caller. The WR corp played very good fundamental ball when he was WR coach, and the rest of the 'skill position' players did as well when he became OC.

25
by DGL :: Tue, 01/26/2010 - 5:15pm

And Sweed was inactive for most of the second half of the season, so there you go.

42
by Jerry :: Wed, 01/27/2010 - 4:41am

His hands, of course, were inactive even longer.

43
by Jerry :: Wed, 01/27/2010 - 4:51am

The offense, aside from the one game a year where an opponent's defense comes up with something different and gets eight sacks, has been good for the last couple years, and BA deserves credit. Two things frustrate me: (1) Coming out in an empty set on 3rd and 1. Even if you want to go shotgun, have a running back you can fake a handoff to. (2) The inability to sustain a running attack when leading in the fourth quarter. I don't know how much it's just nostalgia for "riding the Bus" in that situation, but I really miss watching them grind out first downs; when they actually did in Miami, it was great.

47
by rk (not verified) :: Wed, 01/27/2010 - 2:45pm

I am baffled by the people that complain about Arians. Last year the Super Bowl champs scored 347 points and allowed 223. This year the team that missed the playoffs scored 368 and allowed 324. It is absurd to blame the offense. Even without those numbers it was abundantly clear that the Steelers lost games because of defensive collapses and horrible special teams.

14
by Packerpalooza (not verified) :: Tue, 01/26/2010 - 4:47pm

Green Bay's number is likely skewed because early in the year the offensive line was a mess and couldn't push through a marshmellow fence plus Ryan Grant wasn't running with either forward lean or aggressiveness. After Tauscher showed up things got better though Daryn Colledge remains a disaster at guard.

17
by Packerpalooza (not verified) :: Tue, 01/26/2010 - 5:01pm

Ahman Green rejoined the club and took over as the short yardage back. Grant took that as a challenge and when given the opportunity seemed to take on a more attacking style.

28
by Arkaein :: Tue, 01/26/2010 - 5:42pm

One other thing to look it is that the coaches might have felt that pass plays really provided a pass-run option, since Rodgers was effective at running for first downs on called pass plays if nothing was open.

It would be nice to know what GB's success passing in these same situations was, since overall GB was a very good 3rd down team across all distances, and also to know how many of the successful runs were Rodgers scrambles instead of called runs.

EDIT: oops, somehow missed some of the previous comments.

22
by ammek :: Tue, 01/26/2010 - 5:05pm

I don't think Grant got many of the carries in power situations at any point in the year. It's just not his thing. When the Packers run, it's either a QB sneak, or the fullback John Kuhn. I suspect the Packers' above-average success at power conversions owes a great deal to Aaron Rodgers.

Oh, one other thing: McCarthy just luurves to throw deep on third-and-1. He regards it as a two-down situation, so he sometimes takes a shot on third down, and runs a sneak on fourth.

24
by Packerpalooza (not verified) :: Tue, 01/26/2010 - 5:15pm

Point taken on Rodgers....

21
by Nuk (not verified) :: Tue, 01/26/2010 - 5:05pm

I think calling those situations "power" situations is misleading. Power is force times speed, and you don't need a lot of speed to gain a yard. To draw an analogy with motorcycles, the high-power bikes are the sport bikes - i.e., a Chris Johnson situation. 3rd and 1 would call for a Harley. So I'd call 3rd and 1 a "torque" situation.

31
by tuluse :: Tue, 01/26/2010 - 7:42pm

Power is work/time.

32
by MJK :: Tue, 01/26/2010 - 8:38pm

work = force x distance

speed = distance / time

power = work / time = (force x distance) / time
= force x (distance / time) = force x speed

So you're both right.

39
by RickD :: Tue, 01/26/2010 - 10:51pm

Glad somebody had high school physics.

40
by DGL :: Wed, 01/27/2010 - 1:50am

But time is money.

And money is power.

So money = work/money, or work = money squared.

Therefore, money is the square root of work.

But money is the root of all evil.

So work is all evil.

Q.E.D.

44
by Bobman :: Wed, 01/27/2010 - 6:23am

First you get the money.

Then you get the power.

Then you get the women.

Tony f-ing Montana didn't need no f-ing HS physics course!

50
by Dork (not verified) :: Wed, 01/27/2010 - 10:41pm

Knowledge is Power
Time is Money

Power = Work/Time
Knowledge = Work/Money

Money = Work/Knowledge

As Knowledge approaches zero, Money approaches infinity --
i.e. the less you know, the more you make.

36
by Brendan Scolari :: Tue, 01/26/2010 - 9:57pm

I think it's more referring to the footballism of "power football". When you think of a "power running team" you think of a bunch of big, fat O-lineman blocking for a Jerome Bettis/Brandon Jacobs type back. You don't think of a team of fast guys (or at least I don't). And traditionally "power running teams" were also supposed to be teams that could churn out a couple yards on the groudn when they needed too, hence the name of the stat.

27
by MJK :: Tue, 01/26/2010 - 5:28pm

I don't think the data's out there, but I'd actually be interested to know what percentage of the time a team LINES UP in a run or pass-oriented formation in power situations. That might go a bit towards knowing when they had a run called but audibled to a pass (or vice versa). I know one of my pet peeves is when I see a team in 3rd and 2, say, come out either in a heavy I formation with no WR's to even threaten a pass, or an empty backfield shotgun formation so that the run is never even an option*. The glory of 3rd and 2, as opposed to say 3rd and 6, is that you're right on the tipping point, and neither running nor passing is especially favored.

* The one exception is the rare occasions where a team comes out in heavy I personnel that unexpectedly line up in a 4- or 5-wide shotgun to trick the defense into having linemen and linebackers being forced to cover athletic TE's and RB's in space. That's just fun.

37
by Brendan Scolari :: Tue, 01/26/2010 - 9:59pm

Well, you could look it up pretty quickly for just one team by just going through the game charting at PFF, but it would be time consuming to do that for the entire league so that you coudl set up an average. I agree that it'd be interesting to see what formations teams lined up in in different situations.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

More information about formatting options

CAPTCHA
To skip this, please log in.