A Brief Look at the Impact of Scheme in Punting

Submitted by steve sharik on

There was a recently deleted thread about concern of the kicking game.  I am not in the least concerned about our kickers and punters.  What I'm concerned about is coverage.  Kickoff coverage once again looked a little suspect.  And although Chesson made a nice play on punt coverage, he was the only guy there and if the ball is one of those Hagerup boomers and the returner makes the first man miss, he'll get 10+ yards before he meets the next defender.  That is simply atrocious.  

The goal of punt return is to get an extra first down; i.e., get at least 10 yards.  The NFL style tight punt makes it easier to get that.  I mean, how hard is it to look at the statistical leaders in net punting over the last five years and the schemes they run and draw the obvious conclusion?

For example, Alabama runs NFL style tight punt.  No team has the depth of talent as Alabama.  If anyone should be good at covering punts out of an NFL style scheme, it's the Crimson Tide.  Yet in 2013:

RK TEAM AVG PUNT
3 Alabama 46.9
     
RK TEAM RET AVG
112 Alabama 13.4

I know that some of you are thinking, "This doesn't include punts that aren't returned, and so really it's net punting average that counts."  There's some truth to that, so in the interest of fairness, the Tide was #1 in the land in net punting, mostly due to the high punting average and the fact that they only had 25 of their 40 punts returned.  

One team that jumped off the page was Purdue, who punted 73 times and only had 10 returned on them. The Boilers averaged 43.3 yards per punt, and despite giving up a meaty 9.7 yd/return, was #2 in net punting average.  Purdue even had a punt blocked by Wisconsin.  Purdue has a very low talent level, as well as depth of talent.  Three guesses as to their punting scheme.

Also, in the interest of fairness, Michigan averaged giving up 6.3 yd/return, good for #37 in FBS. We were pretty meh in net punting b/c we averaged 40.7 yd/punt.

It is my opinion that with a spread punting scheme, very athletic punters who can launch in not only Hagerup but also Kenny Allen, that we could gain a lot of hidden yardage by switching schemes. 

I know, I know; preaching to the choir.

Comments

grumbler

August 31st, 2014 at 4:38 PM ^

I must confess that I have no idea what you are arguing here.  Are you arguing for, or against, tight punting formations?  I haven't a clue as to which scheme Michigan or Purdue uses, and the Alabama example means nothing because there is no context or counter-example here.

aiglick

August 31st, 2014 at 6:03 PM ^

He's referring to the rugby style of punting or the spread punt but for some reason our staff wants to use the pro style punt which has been shown to be to the team's detriment. Not that big a deal until a team breaks a punt for a touchdown. Let me be clear that I wish we made this change but for some reason we l have stuck with the pro formation.

Yeoman

August 31st, 2014 at 6:24 PM ^

...because it's supposed to be an argument against the NFL-style protection they use, but they were #1 in the country in net punting.

There might well be a cogent statistical argument to make in favor of shield-punts. This wasn't it.

steve sharik

August 31st, 2014 at 8:23 PM ^

It's clear from the data that Alabama led the nation in net punting not b/c of their coverage (13.7 yd/return).  That they were #112 in the nation at average punt return yds allowed with one of--if no the--most athletic units shows pretty cogently that the way they cover punts hurts them.

I'm guessing 'Bama was #1 in net punting b/c they usually had great field position and their punter had great hang time on his punts.

The problem with the NFL style in college is not with the gunners, but the other guys.  The gunners will get there quickly but the interior guys can easily be held up.  With the spread punt (2 yd gaps b/w players) and athletes at those positions, they can easily evade blockers and get down field as quickly as gunners.

Yeoman

August 31st, 2014 at 10:13 PM ^

Or maybe they forced all those fair catches because their coverage was terrific, and the occasional long return happens when their punter, who's obviously got a big leg, hits one so long and low it'd be impossible to get it covered no matter what system they used.

Who knows, without looking at film? Giving all the credit for the zeroes to the punter but all the blame for the return yards to the coverage seems like handwaving to me.

And whatever they did doesn't seem to have hurt them all that badly because at the end of the day they were #1 in the country in net punting. BYU was, if memory serves, #107. That doesn't prove Alabama's system is superior but it sure as hell doesn't prove the reverse.

Or you could go back to 2012, when Alabama's net numbers were mediocre and BYU's were tops in the country...and with the same pattern as Alabama's last year, lots of fair catches but a pretty respectable average on the few that were returned.

When you know in advance what you think is right, it's easy to find something in the numbers to back it up. But since the other side can do the same thing just as well, none of it's terribly convincing.

steve sharik

September 1st, 2014 at 3:02 PM ^

How can the coverage be good when they give up 25 returns for 13.7 per? What, the coverage was really good 15 times but dudes were held and clipped the other 25 times?

Sorry that I didn't have a linear regression with a nice R^2.

Alabama punt coverage:

  • 2012: 9.9 yd/return allowed, 17th in net punting
  • 2011: 13.2 yd/ret, 57th net
  • 2010: 12.5 yd/ret, 68th net
  • 2009: 15.1 yd/ret, 81st net

I think two things are clear:

  1. Alabama has improved consistenly in net punting and average return allowed
  2. They've been a very poor coverage unit under Saban, armed with the best athletes in the land and the worst punting scheme.

I'd like to see the stats on shield teams getting punts blocked and giving up scores, too.  I'm not MCalibur or Mathlete, with archives of play-by-play data, however.

Reader71

August 31st, 2014 at 6:24 PM ^

Any data on blocked punts? I think Hoke and the staff are more concerned about the shield punting formation yielding more blocked punts. I think I remember seeing that there was either no correlation or that teams using the shield actually had fewer blocks.

Yeoman

August 31st, 2014 at 6:45 PM ^

Here's a list of every team that had more than one punt blocked last year.  I won't pretend to know what every team in the country is running. It's suggestive, at least, that BYU tops the list (they're one of the more famous spread-punt teams, referenced by S.S. above), but anecdotes aren't data.

Maybe somebody else can make something of this.

3 punts blocked:

  • BYU
  • Tulane

2 punts blocked:

  • UTEP
  • Arizona St.
  • Penn St.
  • Indiana
  • Georgia
  • Kansas
  • Tulsa
  • Louisiana-Monroe
  • New Mexico St.
  • Mississippi St.
  • Colorado St.
  • Oklahoma
  • Buffalo
  • Marshall

36 teams had one punt blocked, 71 didn't have any.

Raw data here, including the number of punts for each team.

Both versions of this seem a bit simplistic to me. It's not obvious to me that maximizing net punting average is really the ideal goal; it's pretty obvious that minimizing punt blocks shouldn't be (if that's really your goal, don't punt in the first place). This seems like a task the football-outsiders methodology is well equipped for--how much field-position value did you gain with your punt, on average? Isn't that what we want to maximize?

Reader71

August 31st, 2014 at 9:11 PM ^

My comment wasn't a defense of Hoke or spread punting. It's just that from what I know of Brady Hoke, he seems to be a guy who would rather die than lose a game on a blocked punt. Not that he is risk-averse in general, but I wouldn't be surprised if his mentality is that he'd rather give up a few yards per punt in the name of not having a calamitous, tragic loss via blocked punt. Huge assumption obviously, that the spread punt is safer than the shield. I'm not sure it is.

Yeoman

August 31st, 2014 at 9:52 PM ^

Yes, that's a fair assessment of Hoke. It also might be right--if there's a tradeoff to be made between a bit of net average and a lot of TD-producing variance (long returns or blocks) it might be rational to opt for safety (and like you I don't know which method is safer).

I have no dog in this fight either. Most of the arguments I've seen put forward (I don't mean just here, anywhere I've seen it discussed) have seemed pretty poor, in both directions, and I'm hoping for something better.

Reader71

August 31st, 2014 at 11:34 PM ^

Again, this is just pure abstract speculation, but if it is true that shield punting leads to more return team scores, than I understand the reluctance to go with it. I'm not going to lie, I prefer the spread punt (which is what we run), but I also know that this is due to an irrational fear caused by Michigan's 2003 season with the shield/rugby punt experiment. I hate this fear and realize it is probably irrational, so I'd love to be convinced.

grumbler

August 31st, 2014 at 9:04 PM ^

Any special training or personnel needed for the spread punt formation, or is it something a team  can switch to whenever (before offseason practices start, obviously) they decide to?

Yeoman

September 1st, 2014 at 11:10 PM ^

To do this we need to know who's using which system.

Mathlete's article a while back on shield punts ran into the same problem. He thought shield punting was probably better on circumstantial evidence--net punting was improving, and more teams were using the shield. But he didn't know which teams were using what, so he couldn't prove his point the way he wanted to.

No one of us is likely to see every team ...but all of us put together might come close. We'd at least have enough coverage to have some usable data. If you're interested in helping get this together, take a note whenever you're watching a game of which team used which style of coverage and post it here. We probably won't hit every team in the Sun Belt but I'm guessing by mid-season somebody here will have seen pretty much every major-conference team.

It might be useful to know who's rugby-punting, too.

wolfman81

September 4th, 2014 at 1:37 AM ^

Let the crowdsourcing begin.

Seriously, there are football fans of every team (or enough of them anyway).  Label teams as spread/shield punting teams or pro-style punting teams by having fans classify them.  There are only millions of us.  Then the t-tests or other non-parametric two sample rank sum tests can begin so that we may end this debate...or muddle it further.  Isn't college football great?

Space Coyote

September 2nd, 2014 at 9:56 AM ^

Shield (Called spread above and often called that, to everyone's confusion)

Strength: Long snapper mirrors, two OGs get spill, two ends still act as gunners, and OTs contain. Shield players and punter are safeties. It forces the defense to declare its intention. It gets more players down in the first wave. Is more spread out (similar to a KO) and therefore is easier to define coverage lanes. Simplified release for the players on the LOS (step outside into gap and release upfield instead of going back to block).

Weakness: 4 guys in the two A-gaps can block a punt directly up the middle (shortest route to the football); rushers are typically going up against OL and FBs in space for the "shield" aspect of it, and it's hard for those shield players not to chase. Besides the rugby scramble, it's also more difficult to fake because penetration is allowed. After the first wave you have less support because you have OL types chasing the play and acting as safeties. Punt generally needs to get off quicker and cleaner because block angle is shortest route to football and there is less emphasis on protection.

Spread (Pro-style)

Strength: Can pick up any 8-man rush, and forces rush to come from the outside (don't let anything inside of you). With more defenders covering the gunners, you can release different players at the snap and leverage the football. Can still use LBs, safeties, as wings and OTs if need be to get down field quicker in first wave. Protects against the block first. Learn an NFL style scheme (good for getting players to the NFL).

Weaknesses: First wave is essentially the gunners and maybe the long-snapper mirroring the returner. Tends to put up to four OL instead of three OL on the field. Isn't spread out and therefore is more difficult to get into coverage lanes.

Conclusion

The shield punt is better at getting the intial wave down field, meaning that it typically will limit return yardage. The issue comes if the coverage gets too spread out or out of their lane, and then the safeties are a few OL and a punter, meaning that it is dangerous if it can be blocked initially. It also means that a lack of discipline is a bit easier to exploit, because penetration is allowed and four guys in the A-gaps can't all be picked up.

Conversally, the pro-style punt forces the rush to come from the edges and can pick up 8 men rushing the punt without problem. The protection is simplified (rather than the release), meaning it's easier to prevent the block. However, the first wave is typically only two players. The return against this formation is more likely (brining in variability), however, the return yardage is typically less variable (because there are more waves).

What it comes down to, is that a shield punt is often kind of all or nothing. Your punter's distance may drop a little because he has to focus on punting in the direction of the shield and has to get it off quicker and cleaner; if he doesn't get adequate hang time, the returner must only break through the first wave and bust a long one, however, that first wave is larger. A shield punt is more conservative. Your variable comes in allowing a return more often, but there will be less drastic shifts in field position. 

So what you'll see is a lot of defensive minded coaches do the pro-style punt as it tends to give the defense a chance to get off the field, rather than saying they need more yards in order to do so. Offensive minded coaches think about winning some field position or getting the ball back quickly if they don't. It's really a mind set thing.

My feelings

In general, I prefer the shield punt myself. There's a reason you spread out on KOs, you might as well utilize that advantage on punting. I think blocked punts are a rare thing anyway, as are fakes, so my preference is the shield. But I see it both ways.

 

Marley Nowell

September 1st, 2014 at 11:24 PM ^

On a long punt that Hagerup booms downfield we are look at an average of 10 yards/return? So on the shorter punts coverage will get there quicker so that should be reduced correct?  I am not against the spread punt but I am just curious as to how Hoke seees it.  GIving up 10 yards in the other teams redzone area or even less than that on shorter punts doesn't seem like it matters that much, but I don't know jack.

steve sharik

September 2nd, 2014 at 1:26 AM ^

...Mesko punted pro-style for 2 years, averaging 41.6 and 41.1 yd/punt, then rugby/shield (mostly shield after the Sparty fake debacle) and averaged 43.0 and 44.5 yd/punt.

This doesn't prove anything (/glares at Yeoman), but it suggests that shield punting doesn't affect performance.  If that's actually true, the 10 yards certainly matters.