Other Critical Analysis of 4th Downs

[Exe Gesis]

Here's a number of great articles and analyses on the topic of 4th Downs in football.

http://www.nflbowl.com/re-analysis-of-4th-down-attempts-the-impact-of-momentum/
http://mgoblog.com/diaries/fourth-down-decisions-never-punt-tebow
http://www.advancednflstats.com/2010/05/4th-down-briefs.html

(Romer's famous paper on 4th Downs)
http://www.econ.berkeley.edu/users/dromer/papers/PAPER_NFL_JULY05_FORWEB_CORRECTED.pdf

(referencing and critiquing Romer's paper)
http://www.pro-football-reference.com/blog/?p=40
http://www.pro-football-reference.com/blog/?p=41
http://www.pro-football-reference.com/blog/?p=42
http://www.pro-football-reference.com/blog/?p=46

(a poker genius' take on 4th Downs)
http://web.archive.org/web/20050301053618/www.twoplustwo.com/magazine/current/sklansky0305.html

(another famous theory on never punting at all)
http://www.footballoutsiders.com/stat-analysis/2006/never-punting

(and an article about a team that tried the never punt offense)
http://highschool.rivals.com/content.asp?CID=892888

 

[JeffDunne]

Well that was a good way to do my head in first thing in the morning.

Interesting reading but I still think you need to base calls on the game & the opposition tendancies rather than a stastical analysis.

The statistics do show one thing though - common sense is often a distorted perception.

 

[greennick]

Information overload! If you were to only read one article, which one would it be??

 

[drd23]

the advancednflstats one is the simplest to understand.

 

[Exe Gesis]

I would read the 4 articles that critique Dave Romer's paper.
I'd also read the "never punt" one for a bit of bizarrity.
There's also that advancedflats one which has a slideshow presentation way of explaining it.

 

[JeffDunne]

What people often fail to comprehend is the impact on play calling if you abandon the option to punt.

If you adopt 4 downs to get 10 rather than 3, you open up your playbook on the first 3 downs (i.e. less than 10 yard plays are still an option on 3rd down).

Can't see you'd ever abandon punting completely, but I'd love to see someone try it.

 

[Exe Gesis]

There are some famous cases of teams who have tried it.
Mike Leach's Texas Tech often abandoned punting.
And there was a high school that also tried it more rigidly and they won their championship.

 

[Exe Gesis]

They followed it to a T, including going for an onside-kick every time.

http://highschool.rivals.com/content.asp?CID=892888

Arkansas coach punts traditional game plan
by Jeff Fedotin

Kevin Kelley decided to flip football convention on its head after Pulaski (Ark.) Academy's second game of the 2007 season.

Never a fan of taking his offense off the field, the coach became miffed when his Bruins punted away to Pine Bluff (Ark.) Dollarway High only to see Pulaski allow an 80-yard touchdown on the return.

"That was stupid," Kelley said. "We should've gone for it."

As a result, his 2008 team did not punt during 14 games. Such an unorthodox strategy may seem like lunacy, but it was successful: Pulaski won the 5A state title on Dec. 6.

Kelley's team only punted twice in 2007 − once as an act of sportsmanship to prevent running up the score − and never after that Dollarway game.

Kelley has reasons to go for it.

Keeping the offense on the field on fourth down allows for more creative play-calling. Third-and-long does not have to be a passing down. The Little Rock school can run the ball, throw a screen pass or use any number of formations. Defenses do not know whether to use a nickel or dime defense. And Pulaski's offense has less pressure on third down.

"We don't really worry too much about it," quarterback Spencer Keith said. "We just get as many yards as we can. We don't have to go for the first down."

If Pulaski converts on fourth down, it creates a momentum change similar to a turnover. Other high school coaches have told Kelley they would rather see his team punt.

The Bruins even avoid punting when the defense has stopped them inside their own 10-yard line.

"You can just tell people are in the stands thinking, 'You're an idiot,' " Kelley said.

Kelley supports this rationale with numbers analysis.

If Pulaski has a fourth-and-8 at its own 5-yard line, Kelley said his explosive offense likely will convert a first down at least 50 percent of the time. If it fails to convert, statistical data from the college level shows that an opponent acquiring the ball inside the 10-yard line scores a touchdown 90 percent of the time. If Pulaski punts away (i.e., a 40-yard punt with a 10-yard return) the other team will start with the ball on the 38-yard line and score a touchdown 77 percent of the time. The difference is only 13 percent.

An innovative and statistics-minded coach, Kelley had tinkered with eschewing the punting game since winning his first state championship in 2003. He became further emboldened after reading several studies, including "Do Firms Maximize? Evidence from Pro Football," by University of California-Berkeley economics professor David Romer. Kelley also examined ZEUS, a computer program developed by Chuck Bower, who has a doctorate in astrophysics, and Frank Frigo, a game theory expert, to model and predict football outcomes.

The Pulaski coach has adopted an unusual approach to kickoffs as well. About 75 percent of the time, he uses an onside kick instead of a standard kickoff. To illustrate why, Kelley again relies on numbers.

If his team does not recover the onside kick, the opponent likely will field the kick around its own 47-yard line. On a typical kickoff, the other team usually starts around the 33-yard line.

"You're only giving up 14 yards," Kelley said. "And you get a chance to get the ball."

Pulaski features seven different kinds of onside kicks, including bunching eight players on one side of the field and three on the other; faking the kick with one kicker while another player shifts over to kick to a vacated spot in coverage; clustering all 11 players before spreading out just as the ball is kicked; bouncing a hard kick off the turf for a jump ball and launching a "helicopter kick" by kicking a ball placed on the ground against the tee. The latter strategy causes the ball to spin like a helicopter's propeller and move like a curveball.

"Much like the punting situation, [the onside kick] becomes something the other team has to work on a lot during the week," Kelley said. "That's taking time from their preparation against your offense or defense. So it all works towards the common goal."

For Kelley's objective of winning games with a risky but aggressive offense, Pulaski had the perfect quarterback. Keith, who has received major interest from Louisiana Tech, Arkansas State and several Ivy League schools, could make defenses pay for not stopping the Bruins on fourth down. And if the other team scores off a short field because of a missed fourth-down opportunity, the unflappable passer could compensate by scoring points in a hurry.

Kelley called him the most athletic quarterback he has ever coached, and Keith set the state record with 5,308 passing yards this season. He also possesses the requisite accuracy for an offense that threw on about 45 of its 75 plays a game and averaged 570 total yards.

"It's a really fun offense," Keith said. "I wouldn't change it for anything."

A possible pre-med or pre-engineering student who scored a 30 on the ACT, Keith has the intelligence to master an intricate scheme, which features pre- and post-snap reads with receivers making adjustments based on coverage.

With Keith and several other impact players returning from his 2007 team, Kelley said he knew his team had the potential for a state title this year. When the media asked for his pick of the No. 1 team in Arkansas before the season, Kelley chose Pulaski and consequently received some heat.

"It was just confidence in my guys," he said. "I thought this might be one of the better teams we've ever had."

His prediction proved to be on the mark. Although Pulaski lost its first game of the season, 46-29, to West Helena, it reeled off 13 consecutive wins and avenged that Week 1 defeat with a 35-32 state title victory against the Cougars.

During the offseason Kelley will begin investigating different football strategies. He also plans to further study the punting game by analyzing specific instances where punting may prove statistically superior.

After Kelley searches through data, Pulaski may tweak its approach next year. His 2009 team could punt on occasion, or he may develop a new tactic that defies the norms of football but gives his team an edge.

"Just because something's always been done that way," Kelley said, "doesn't mean it should continue to be done that way."

 

[TheGreatBarryB]

Outside the Square HS Coach - No Punts or Kicks

Off season so thought a good time to post this.

Reading a great US book called Scorecasting which discusses the hidden influences behind how games are played and won.


Anyway it discusses a high school coach named Kevin Kelley whose Arkansas HS team Pulaski Academy Bruins have no punter or kicker on their team.


Kelley’s thought process is quite interesting.


Kelley has used HS football stats to back his theory that punting and kicking are a waste of time. They punt even if backed in their own 20.



They always go for the 2 pt conversion rather than PAT.


They will onside kick rather than just kick off to restart. They also never return punts.


Kelley says it is all based on rational thinking not madness as most would think. The guy is a trained Economist so when he first started the job he amassed lots of stats and based on results formed his game plan.
He found that in Arkansas HS football teams tend to average a TD on 1 of every 3 possessions. So he reasoned by punting the ball away three times when he didn’t have to, he would essentially be giving up a TD every game. Also his teams averaged 5-6 yds per play.


His 4th down plays succeed roughly 50% of the time. His logic on punts is that in HS game the average punt nets around 30 yds it doesn’t make sense to punt if you can convert half the time.


His logic to punting when deep in his own 10 if they punt opposing teams that take possession inside the 40 yd line (after a net punt of 30 yds) will score 77% of the time. If they take over from unsuccessful 4th down inside 10yd line they will score 92% of the time. He therefore thinks its worth the extra 15% risk by going for it on 4th down.


For onside kicks his team recovers between 25-33%. With a normal kickoff on average a team takes over on own 33 yd line. With unsuccessful onside kick it assumes possession on own 48 yd line. So Kelley’s logic is you’re giving up 15 yds for a roughly 30% chance of getting the ball back. Probably worth the risk when thought about that way.


He doesn’t return punts because his school never has any speedy players and punts in HS travel 30-35 yds on average. More likely they will be called for a penalty or simply fumble the ball. So Kelley thinks it’s not worth the risk.


The guy also runs various trick plays, runs on 3rd and long , passes on 3rd and short, basically drives defences crazy. Opposing coaches say they need to practice more when coming up against his teams and admit their D can get demoralised conceding so many 4th downs. So psychologically it gives Kelley’s team a big advantage.


Pulaski HS is mainly made up of wealthy white kids in small community so Kelley hasn’t got much of a player base to pick from. He’s 77-17 with a few State Championships so he’s been highly successful.


Obviously it’s different in College and NFL where punters can kick 50-60 yds and there are plenty of speedy PR. But in the book they argue that NFL coaches are naturally risk averse and in many 4th down situations going for it would be more rational decision.



They even defended BB’s call a couple of seasons back when Pats went for it on 4th down on their own 28 against the Colts.

 

[bjtoby]

Re: Outside the square HS coach - no punts of kicks

I remember hearing about this on one of the espn shows in '09 after the pats 4th and 2 attempt against the colts. I think it's rather interesting, I'd say why it would work in some High School divisions, in College and NFL it would be a lot harder to convert it. I would wonder what he would do if down by less then three 10-30 yards out. Would he go for it or take the easy fg for the draw/win?

 

[Exe Gesis]

Merged.

 

[Dirty Bird]

http://sportsillustrated.cnn.com/2011/writers/scorecasting/09/15/kelley.pulaski/index.html

Kelley doesn't believe in punting. His Bruins teams go for it on fourth down, even in the most extreme situations. His playbook is filled with tricks and gimmicks. He often forbids his players to return punts, reckoning that the odds of a fumble outstrip the incremental yards that can be gained from a return. After his team scores, it almost always attempts an onside kick. There are 12 varieties in the playbook

http://sportsillustrated.cnn.com/2011/writers/scorecasting/09/15/kelley.pulaski/index.html

Think we'll ever see it in the NFL?
There's reports the San Diego State Aztecs in college are seriously contemplating it. So it is creeping up a little.

 

[drd23]

No it wont because NFL HCs tend to be very risk adverse

 

[Blacky]

Ahhhhh wouldn't have thought so, you WILL get punished by a good opponent for trying those things, and as soon as the opposition knows you're going to attempt onsides, 4th down conversions etc it will be a lot harder to pull them off

 

[Southerntakeover]

No, because it's a terrible idea which disrespects the game.

 

[Broken]

NFL (and NCAA) coaches will always lean to the conservative side to protect their jobs. The majority would rather punt (or FG) on fourth and 1 than go for it and have it fail in turn copping the wrath of the decision.

 

[Dixie Flatline]

There was a newspaper report (possibly from the NY Times or from Grantland, can't now remember which) discussing a statistical analysis of teams that go for it on fourth down compared to those that punt on fourth down. It turns out that, statistically, teams are better off going for it on fourth down, but most head coaches in the NFL are terrified of going for it and failing in bad field position for their defensive units that they will almost always go for the punt.

ETA - found a link to a Fifth Down blog post (via the NY Times) from 2009 discussing the statistical analysis I referred to above.

http://fifthdown.blogs.nytimes.com/2009/09/17/a-new-study-on-fourth-downs-go-for-it/

 

[larkis]

Would of been great if the Rams did this last season while we had McJackass. Might of won a few more games ...

On a serious note this is more than possible to do. With the expanded preseason rosters a team can easily plan to do offside kicks and fourth down plays or the two point conversion. Won't be surprised to see the Jets see how lucky Tebow is and use him a lot like this.

 

[drd23]

There was a newspaper report (possibly from the NY Times or from Grantland, can't now remember which) discussing a statistical analysis of teams that go for it on fourth down compared to those that punt on fourth down. It turns out that, statistically, teams are better off going for it on fourth down, but most head coaches in the NFL are terrified of going for it and failing in bad field position for their defensive units that they will almost always go for the punt.

ETA - found a link to a Fifth Down blog post (via the NY Times) from 2009 discussing the statistical analysis I referred to above.

http://fifthdown.blogs.nytimes.com/2009/09/17/a-new-study-on-fourth-downs-go-for-it/

There are many analyses that show that it is statistically superior to go for it in most 4th down situations, and also to go for a 2pt conversion every time.

The problem is that those statistics apply to a large number of plays and don't apply as well to any given play between any 2 teams, so when a team follows that analysis but fails and it costs them the game, the HC will cop a beating from the media despite the decision being based in fact.

As such, most coaches won't do it - although we are seeing HCs acknowledge these statistical analyses more and more (I believe that Mike Smith noted that going for it was the right decision according to the statistics in their play-off loss last year for example)

 

[Exe Gesis]

The stats certainly didn't help the falcons a few times or the patriots. In the NFL what will work more is trick punt and fg plays.

 

[Exe Gesis]

Btw there's already the same thread about this very topic. Merge eventual

 

[drd23]

The stats certainly didn't help the falcons a few times or the patriots. In the NFL what will work more is trick punt and fg plays.

The problem in those instances were play calling (the Falcons last year) and execution (the Pats 4th & 2), not necessarily the decision to go for it itself.

Who on earth thought it was a good idea to run a QB sneak with an empty backfield or to shift the entire line?

 

[TheGreatBarryB]

Are all these stats based on past 4th down decisions? If so aren't they flawed? The resaon is teams would have gone for 4th downs thinking they would be a good chance to get it based on situation at the time. Also the fact that it would have been more of a "shock" to D.

Think it's a different story if everyone in NFL just goes for it on 4th downs regardless of the situation. Those stats would look a lot different.

 

[Burningleviathan]

Failing on fourth down at your own 15 yrd line will have far more consequences than if you punted. The risk is beyond absurd.

That being said, most teams now acknowledge going for it on 4 down on the other teams 35 yrd line has far more benefits than a punt that starts them at the 20 (if they don't trust the kicker from about 52yrds).

 

[Dixie Flatline]

I noted the Cowboys tried for it on fourth down and short early in the game against the Giants. From memory, they tried the good ol'-fashioned run up the middle, which the Giants were alive to and squashed at the line of scrimmage pretty easily.