Herm And The Yards-Per-Play Phenomenon

[FAX]

My crack research team of rocket engineers, nuclear scientists, and painter monkeys ran across an interesting article tonight. Albeit dated, the article raises a point of interest to me considering our offense has recently taken a healthy dose of Hermicide.

Basically, the article cites Elias Sports Bureau statistics in respect to the passing game and two primary factors that dictate wins/losses:

First, the average pass <b>(even counting incompletes)</b> outgains the average run by 50-100%.

Second, the team that scores first wins most NFL games.

Keeping in mind that the article is somewhat dated (August 31, 2005), the statistics they use are pretty compelling in terms of winning percentage by lead in a game (previous five years).

<b>Lead Pct. Games Won</b>
3-0 59%
7-0 71%
10-0 82%
14-0 87%

The premise of the article is that, if you score first, you dramatically increase your odds of winning the game and passing teams (because the average yards per play is greater) have the best chance of scoring first. The following is quoted from the article ...

"Last year, again according to Elias Sports Bureau statistics, the average NFL offense gained 4.1 yards per run and 6.1 yards per pass (even factoring in incompletes) — roughly a 50% advantage for pass plays.

For good passing teams like New England and St. Louis, the average gain was about 7 yards per pass — a 75% advantage over an average running play. For a transcendent passing team, Peyton Manning's Colts, the average gain was 8.5 yards per pass — double the 4.3 yards averaged by Edgerrin James and other Indianapolis runners."

http://www.latimes.com/sports/la-093005oates,1,500987.story?coll=la-util-sports-football-nfl

Anyway, I googled up one of John Clayton's articles (writing for ESPN in 2005) and his statistics seem to support this theory (at least in part). Basically, his stats show that, in 2004, the Colts averaged 6.7 yards per play followed by the Vikes, Packers, Chiefs, Eagles, Goats, Rams, Eggos, etc. And, 10 of the 13 teams who led the league in YPP made the playoffs.

http://sports.espn.go.com/espn/print?id=2145710&type=story

My crack research group loves defense as much as the next homer, and certainly, the best of all worlds is a balanced, high quality performance on both sides of the ball that not only defeats the enemy but shames them into oblivion and sends them home to their mothers. But, one has to admit that there's no substitute for points on the board when you need them and YPP (and by extension) an effective passing offense appears to be a major factor in accomplishing that goal.

I'm interested in the Planet's take on this.

FAX

Disclaimers: Sorry if repost.

[Direckshun]

Fascinating point of view there, FAX. You exhaust my rep button sometimes.

I have to step in here and defend the SchottenHermers here, though, because that's my type of football. I say Nay! to your Steve Young west coast offense. Bring me a power run game and everything else flows from that. LJ up the center's ass? 'Til Wiegmann's ass bleeds, baby.

A better predictor of victory in the NFL is not passing yards, it's rushing yards. I'm not going to go deep into NFL history to support that theory because hell, we're just two guys on the couch and it's not worth researching that.

You bring up compelling stats though. Passing yardage in its worst case still trumps rushing yardage in its best -- but it also gives up three times as many turnovers in a dozen more different ways.

And you don't get to be the first team to score unless (a.) you have an equally effective defense, or if you can't get one of those, let me suggest (b.) you win the coin toss every single time for the rest of human existence. I can't attribute "scoring first" stats to be a strong supporting point of passing > rushing. In my opinion, if you want to score first it's more important to have a good defense than a good offense.

I guess I should conclude by saying that therefore, I approve of the team Herm's trying to build. Whew. I've done my thinking for the night.

[FAX]

That makes perfect sense, Mr. Direckshun. For the record, I'm not saying that a pass-happy, wacky, circus offense with clowns and a fire-eating bearded lady is the best offense. It is, however, interesting to me that YPP is such a major factor in dictating wins/losses and that the pass (even when factoring in incompletions) translates into a much higher YPP.

Like you, I have neither the time nor inclination to dive into this up to my nose, but I did check another source that breaks down these kinds of things. It's TwoMinuteWarning.com (whom I know nothing about, although I'm sure someone here does) and they say that Offensive - Pass Overall Effectiveness equates into a win/loss factor of 0.65. The only higher offensive correlation in their system is Points Per Drive.

As for scoring first, the only argument that can be made in that department is that good passing teams eat up chunks of yards and aggressively go for the end zone rather than try to patiently "wear down" the opponent over the course of the game.

I like smashmouth football and I like defense, but when you consider that the teams that are going to the SB lately pass, pass well, and carry a high YPP average, it might be worth thinking about.

FAX

[FAX]

In a last-ditch effort to foster conversation on this topic, here's a quote from the LA Times article ...

"It need hardly be said that running teams don't think of the first quarter as a race for the lead. To a running team, the goal at the start of every game is to "establish the run." This means running on nearly every first down as well as most second and third downs. Their intent, often stated, is not to score quickly, but rather to soften up an opponent for scores that will come later.

Instead of aiming for a lead, the running team aims for eventual physical and emotional dominance — against defensive players who are, at the start of a game, fresh, motivated, and focused on the run. The result in the first two or three offensive series, even for the best running teams, is often three-and-out and an appearance by the punter.

Passing teams, by contrast, and particularly "attack passing" teams — those throwing aggressively on first down — open typical games looking to score in the fewest plays possible. Moreover, the yards-per-play advantage gives passing teams the lead in big plays — those of 20 yards and more that pump up an offense and generate momentum. As a result, a proficient passing team can frequently drive to a touchdown in a handful of plays, while the typical running team labors slowly down the field. Attack passing teams are, literally, racing for the lead."

Sound like anybody you know?

FAX

[TinyEvel]

You guys are both indicative of what makes this place a great FOOTBALL forum.
Fax, your opening post reminds me somewhat of a "how to win at BlackJack" or "How to win at Craps" book.
Plays the odds, the percentages. But how many of us have gone to Vegas with that kind of book memorized, only to leave having paid our contribution to the construction of the next new wing of said hotel.
I would be more prone to believe in these passing numbers if associated defense was a factored in. Obviously, we all have firsthand witness of how a great passing/offense cannot win championships.
Believe me, I am NOT waiting for Herm's offensive scheme with baited breath. Let's say I await it like waiting for an m-80 to blow up in the public school toilet. One hand over the eyes, one over an ear, not knowing if this will be everythig I wanted, or if I'll end up in deep sh*t.
Last season has me worried.
We could show Herm a friggin Price Waterhouse bonafide concrete proof that passing wins games, and he'll just go with what he wants. So I'll continue to be frustrated, and hope the m-80 blows in the direction of the other teams and we come out clean.

[FAX]

You guys are both indicative of what makes this place a great FOOTBALL forum.
Fax, your opening post reminds me somewhat of a "how to win at BlackJack" or "How to win at Craps" book.
Plays the odds, the percentages. But how many of us have gone to Vegas with that kind of book memorized, only to leave having paid our contribution to the construction of the next new wing of said hotel.
I would be more prone to believe in these passing numbers if associated defense was a factored in. Obviously, we all have firsthand witness of how a great passing/offense cannot win championships.
Believe me, I am NOT waiting for Herm's offensive scheme with baited breath. Let's say I await it like waiting for an m-80 to blow up in the public school toilet. One hand over the eyes, one over an ear, not knowing if this will be everythig I wanted, or if I'll end up in deep sh*t.
Last season has me worried.
We could show Herm a friggin Price Waterhouse bonafide concrete proof that passing wins games, and he'll just go with what he wants. So I'll continue to be frustrated, and hope the m-80 blows in the direction of the other teams and we come out clean.

ROFL

It's funny you mention the "How to win .." books, Mr. TinyEvel. That's exactly what I was thinking of when the article was forwarded to me.

But then, I started thinking. Most coaches (if not all) study all kinds of statistics all the time. Then they take those statistics into account when implementing "their" approach, philosophy or strategy. What interests me is that stats don't lie and yet a lot of teams think "defense first", "run first", etc even when the winning percentage is improved when you fling it, fling it well, and fling it often. Additionally, the fact remains that the majority of the teams that have made it to the SB lately are teams that have high YPP averages and can pull the trigger on an aggressive passing game if and when it's needed.

I'm not promoting a particular offensive philosophy here, Mr. TinyEvel. I just thought this was an interesting data point that deserved conversation. And while we're at it, here's another item from TwoMinuteWarning.com ... as mentioned earlier in the thread, Pass Overall Effectiveness (Offense) represents a 0.65 win/loss correlation factor in their system. However, Pass Overall Effectiveness (Defense) equates to only 0.57 ... just under offensive TOP.

Just sayin'

FAX

[Rasputin]

I think you've gotta have a good ballance of run and pass. 49rs had a brutell running attack with Roger Craig and we know they had Jerry Rice.
Cowboys had Emit Smith, Michal Erving, The Colts had a running and passing game last year the list can go on, the Patriots could run and pass, even the Stealers could do both in their SB.

Can any body name a team that has won the SB with out haveing Both attacks or able to do both?

If you stuff the run then you can back off to defend the pass. But if you burn a team with a deap pass then that can open up the running game. It just works both ways. If a team knows you can't pass then they can stuff the run that much easyer. If you can run the ball then that should open up the passing game.

I think Dan Marino was one of the best passers ever but he never had the luxury of haveing a great running attack to compliment the pass.

Most successfull teams have both running and passing.

[Direckshun]

*scratches head*

Well... first of all, and I may be revealing my ignorance here, but I have no idea what "Pass Overall Effectiveness equates into a win/loss factor of 0.65" means. I do know, kind sir, that you're incorrect when you say the only teams that go to the SB these years are the wide-open passing offenses -- the Bears operated primarily through Benson and Jones and the Steelers won the Super Bowl the year before with a one-two punch of Parker and Bettis.

That said, that's an awfully sexy breakdown you posted a couple posts ago of the Smashmouth Run Team. That's exactly what we were in 2005, that's what we were more often than not in 2006, and that's exactly what this Herm Edwards team wants to be this season and beyond. It's a formula that has won Super Bowls under the right guidance.

[FAX]

I think you've gotta have a good ballance of run and pass. 49rs had a brutell running attack with Roger Craig and we know they had Jerry Rice.
Cowboys had Emit Smith, Michal Erving, The Colts had a running and passing game last year the list can go on, the Patriots could run and pass, even the Stealers could do both in their SB.

Can any body name a team that has won the SB with out haveing Both attacks or able to do both?

If you stuff the run then you can back off to defend the pass. But if you burn a team with a deap pass then that can open up the running game. It just works both ways. If a team knows you can't pass then they can stuff the run that much easyer. If you can run the ball then that should open up the passing game.

I think Dan Marino was one of the best passers ever but he never had the luxury of haveing a great running attack to compliment the pass.

Most successfull teams have both running and passing.

This Post Is For Discussion Purposes Only.

In respect to the need for balance between the run and the pass, I completely agree with you, Mr. KC Tattoo. No question. The interesting part about this subject to me is the issue of "priority" in offensive approach on game day.

What the Times article is saying is basically this; the "old school" of establish the run first, wear down the enemy defense, play field position, etc. does not translate into a higher winning percentage compared to teams that strike quickly and establish a lead. Further, he's arguing that teams that pass effectively have a higher YPP and, therefore, are more likely to get a lead. The final piece of the puzzle is that teams who have a lead at any time in any game tend to come out on top when the cheerleaders head for the showers.

Given that, and assuming the statistics are accurate, I think he has a pretty good point.

FAX

[cdcox]

Last year our yards per pass attempt were 1.72 times our rushing per attempts.

We scored first in 9 of our games.

[FAX]

*scratches head*

Well... first of all, and I may be revealing my ignorance here, but I have no idea what "Pass Overall Effectiveness equates into a win/loss factor of 0.65" means. I do know, kind sir, that you're incorrect when you say the only teams that go to the SB these years are the wide-open passing offenses -- the Bears operated primarily through Benson and Jones and the Steelers won the Super Bowl the year before with a one-two punch of Parker and Bettis.

That said, that's an awfully sexy breakdown you posted a couple posts ago of the Smashmouth Run Team. That's exactly what we were in 2005, that's what we were more often than not in 2006, and that's exactly what this Herm Edwards team wants to be this season and beyond. It's a formula that has won Super Bowls under the right guidance.

I was a huge Nigerian Nightmare fan, if that tells you anything, Mr. Direckshun. I loved watching him humiliate defenses in the fourth quarter.

But on to your points ...

Apparently, these guys define "Pass Overall Effectiveness" as part of their "Play Success Rate" quotient which they claim is a statistic that represents how often a team achieves "success" on a given play.

From what I understand from their system, a play's "success" is defined by the type of play and its result. For example, a successful first down play is one that picks up 40% or more of the yardage needed to make another first down (so 4+ yards on 1st and 10, 8+ yards on 1st and 20, and so on). On second down, it's 60% of the remaining yardage needed to get the first (so 6+ yards if it's 2nd and 10), etc. The 0.65 number thingy is simply a factor representing the "weight" of that statistic against win/loss. If it's an offensive factor, you have the ball. If it's a defensive factor, they have the ball. The highest one I found was 0.69. The lowest was 0.10 - Rush For Big Gain (Defense). I'll dig around and find some more information on this, 'cause I think it may be the actual secret to the universe.

As for Da Bears, you're right. But, please note that, during their SB run, their average points per game went up signficantly. If I recall correctly, they were one of a handful of NFC teams to average over 20 points per game and their average points per game jumped probably 10 points from the previous year. They made huge offensive strides. Some of that was the return game, of course. But, nonetheless, they scored points.

The Steelers averaged 8.19 YPA in the passing game when they won it all. Not bad, really.

FAX

[FAX]

Last year our yards per pass attempt were 1.72 times our rushing per attempts.

We scored first in 9 of our games.

I was hoping you would show, Mr. cdcox. You can solve this mystery in a single post.

How many of those 9 games did we win? Do you know off hand?

FAX

[cdcox]

I was hoping you would show, Mr. cdcox. You can solve this mystery in a single post.

How many of those 9 games did we win? Do you know off hand?

FAX



We scored first against Cincinnati and lost.

We scored first in Denver and lost.

We beat Arizona, despite falling behind.

Oakland scored first, but we still won.

In the remaining games, the team that scored first won.

So we went 7-2 in the games we scored first and

2-5 in the games that we allowed the other team to score first. That stat held up pretty well for us last year.

[FAX]

Thanks, Mr. cdcox. I'm so glad you're here. I dug around in TwoMinuteWarning.com for a while and these are their top correlation factors ...

O/D Stat Correlation
Offense Points per drive - 0.69
Defense 1st Down Rushing play ratio - 0.68 #
Defense Points per drive - 0.65
Offense Pass Overall Effectiveness - 0.65 *
Offense Touchdown % - 0.61
Offense Pass Play Success Rate - 0.61 *
Offense Pass Average Yards per Attempt - 0.60 *
Defense 2nd Down Rushing play ratio - 0.59 #
Offense Time of Possession per drive - 0.57
Defense Pass Overall Effectiveness - 0.57 *

* Pass-related offense or defense.
# Rush-related offense or defense.

I note with interest that, of these ten factors, four are pass-related. Either offense or defense. Two are rush-related. The rest are points, man. Except for TOP which is one of Herm's main objectives in order to keep the D off the field.

What's your take on this?

FAX

[Direckshun]

What's your take on this?
I don't know, man, I think the way they break those stats down, you can legitimately argue that a good pass game > good run game.

Brand me a frigid coot, however, because I just can't allow it to change my mind. No set of statistics can take into account the psychological factor of pounding a defense right in the nose with your run game. As you described earlier, if you have an effective run game it essentially wears the defense down physically and psychologically -- which, if you noticed several times it has appeared as if LJ was getting stronger in the second half and the fourth quarter. That wasn't magic, it was an effective run game by a good OL and an elite RB.

Not to mention the strategic importance of sucking a defense in against the run so that your QB can pick them apart on the wings or downfield for chunks of yardage. That's perfect football.

What's my take? My take is that football, on paper, is a game of gaining significant yards to score. And that's where passing looks the sexiest. In practice, however, there's an assortment of factors such as four-quarter strategies, field positioning, and psychological toughness that the run game lends itself to.

You may take my statistical support, BUT YOU CAN NEVER TAKE MY FREEDOM.

[Redrum_69]

Its clear that the statistics are going to be way off balance comparing this years total to last years total if Larry Johnson isnt signed. Brodie Croyle is not going to step in and throw a 4000+ yard season, hes not Manning-esque. If Croyle doesnt step his game up, he'll be pulled as a starter and replaced with Huard, but what if Huard had a freak season last year? I can see Huard stepping in and throwing 4 INTS in a game destroying any credibility he had built from last year. This is where Johnson steps in and carries the team once again. Will Larry work on his blocking skills? Nope..he'll just run more angry. But, without Larry, our hopes will diminish. Bennett cannot carry the team. He just doesnt have the work horse attitude. Wasnt this the same guy who hurt himself in a foot race after practice last year?

With Larry and with the addition to the wide receiver corps we'll have equal stats to last year.

Without Larry....we are screwed

So DAMMIT CARL SIGN LARRY!1

[noa]

What the Times article is saying is basically this; the "old school" of establish the run first, wear down the enemy defense, play field position, etc. does not translate into a higher winning percentage compared to teams that strike quickly and establish a lead. Further, he's arguing that teams that pass effectively have a higher YPP and, therefore, are more likely to get a lead. The final piece of the puzzle is that teams who have a lead at any time in any game tend to come out on top when the cheerleaders head for the showers.

With coaches like Bill Belichick and Andy Reid, I think we are seeing a lot of conventional football wisdom consistently being challenged.
It would have been nice to take an unconventional approach at our playoff game in Indy last season. Everyone in the stadium knew we were going to try to wear the Colts down with LJ. I'm not sure that our passing game was strong enough to justify coming out that aggressively with the pass, but it certainly would have kept Indy honest if we had mixed it up more.

Just look at this play-by-play: http://scores.espn.go.com/nfl/playbyplay?gameId=270106011&quarter=0

We led off one drive with a deep pass, and it was our last possession of the game. Other than that, we led off every drive with a run or a short pass that was basically equivalent to a run.