Grantland: The coach who never punts

[Spaceman Spiff]

http://www.youtube.com/watch?v=AGDaOJAYHfo&feature=youtu.be

 

Thought this could make for some good discussion here.

[dfitzo53]

I think the punt is overused.

 

However, I also think simplifying the entire thing to basic percentages is possibly misleading.  If you have a stellar defense and a poor offense, which clearly happens, your opponent is not going to score a touchdown 77% of the time from midfield.

 

If you have a bad defense and an above average offense you should absolutely be going for it most of the time.

[s0crates]

I have long thought you should never punt in your opponent's territory, and I have often wondered whether or not it would be worthwhile to onsides kick more frequently.

I find the coaches philosophy here intuitively plausible, but I would be more convinced by an actual analysis of the numbers. The coach mentions a Harvard statistician who studied this, but unfortunately the study goes unnamed. I would be interested in seeing that study.

If the numbers really do work, then it is only a matter of time before you see more coaches doing it. Look at what the stats guys have done for baseball.

[dfitzo53]

I have long thought you should never punt in your opponent's territory, and I have often wondered whether or not it would be worthwhile to onsides kick more frequently.

I find the coaches philosophy here intuitively plausible, but I would be more convinced by an actual analysis of the numbers. The coach mentions a Harvard statistician who studied this, but unfortunately the study goes unnamed. I would be interested in seeing that study.

If the numbers really do work, then it is only a matter of time before you see more coaches doing it. Look at what the stats guys have done for baseball.

I've heard of the study before.

I do think we're going to see fewer and fewer punts in the next five years or so.

[twa]

Punting done well is a valuable tool....besides blasting the idiot that doesn't fair catch is fun.

 

it can of course be overused,but not utilizing it at all is a disservice to the team imo

[CRobi21]

Interesting strategy.....for the high school level.

[twa]

Interesting strategy.....for the high school level.

Yep, onside kick receiving alone is problematic for most.

[balki1867]

I think the stats guys are already starting to show up in the NFL, especially late in games.  Doesn't anyone remember Ahmad Bradshaw falling into the endzone in the Super Bowl as he tried to go down at the 1 while the opposing Pats players were trying to let him score?  Statistically, it was better to run the clock and attempt a field goal than score a touchdown and give the Patriots offense 45 seconds.

 

http://newsfeed.time.com/2012/02/06/giants-running-back-accidentally-scores-game-winning-touchdown/

[skinswillwin06]

I think the punt is overused.

 

However, I also think simplifying the entire thing to basic percentages is possibly misleading.  If you have a stellar defense and a poor offense, which clearly happens, your opponent is not going to score a touchdown 77% of the time from midfield.

 

If you have a bad defense and an above average offense you should absolutely be going for it most of the time.

I know it's a video game, but playing Madden has gotten me to buy into the no punt theory if you have a bad defense. A few years ago, when I last bought the game, the defense was atrocious and any if you played anyone remotely good they could score every time.

 

In this situation, when your defense truly can't do anything worthwhile, but your offense is great, it makes perfect sense to go for it on fourth downs. If you are losing, it also makes sense to kick onside kicks every time. The opposing offense will have less field to go and you will get the ball back sooner when they inevitably score again.

 

If your defense is adequate, the no punt theory starts to lose support.

[Spaceman Spiff]

It'll be interesting to see if anyone at the NFL level has the balls to try this.  Too much money, too much attention at stake.  

[twa]

It'll be interesting to see if anyone at the NFL level has the balls to try this.  Too much money, too much attention at stake.  

The numbers change at that level...just as the talent and coaching does.

even at the HS level adjustments to counter can be made...if you want to take the time and effort.

[thebluefood]

Like with most things, the solution is somewhere in the middle. Never going for it on fourth down is far too conservative, but never punting can put you in some very dangerous territory from a field position standpoint.

 

If you're deep in your own territory and it's fourth and forever, you're better off booting it and at least putting some distance between your opponent and the goal line.

 

Having said that, I, too, don't understand punting when you're in your opponent's territory. Chances are, unless you have a Sammy Baugh clone on your team that's a top-notch coffin corner kicker, it's going to be a touchback and the net difference of the kick is negligible. Furthermore, I don't understand not going for it on fourth and short when you're inside your own 40-45 yard line.

[Elessar78]

I don't know why someone wouldn't try it in preseason during scrub time? Those guys are still trying their hardest (to make the cut) and I think it'd give you a good gauge if this stuff would work or not. 

[Elessar78]

The thing is you need an outsider to implement this in the NFL, not someone who has built his career on something else. First, 2nd, and 3rd down playcalling would change a lot in the hope of getting to third and short. 

 

Below is the link for fourth down conversion attempts in college football this year:

http://espn.go.com/college-football/statistics/team/_/stat/downs/sort/fourthDownAttempts

 

They're around 45-60%. I think that's pretty good. I would think that that number would be even higher if teams planned on going for it on fourth down. I can imagine that a lot of these 4th down attempts are in the fourth quarter and most of the teams probably aren't trained for the psychology of routine fourth down conversions. 

[dfitzo53]

I don't know why someone wouldn't try it in preseason during scrub time? Those guys are still trying their hardest (to make the cut) and I think it'd give you a good gauge if this stuff would work or not. 

I don't know, do you really think third or fourth stringers going against third or fourth stingers is an accurate model?  How often do we see guys come in and light it up at the end of preseason games only to ride the pine the rest of the year (or even get cut?)

 

Like with most things, the solution is somewhere in the middle. Never going for it on fourth down is far too conservative, but never punting can put you in some very dangerous territory from a field position standpoint.

 Well part of the point of the "no punt" school of thought is that the field position aspect may be overrated.  If your opponent is likely to score anyway, it doesn't really matter whether they do it from your 25 or their 45.  In fact, it's better for you if they get it over with and leave you with plenty of clock.

 

So the question then becomes at the pro level, what's your probability of scoring from any particular point on the field?  The problem is that it's not a static percentage.  It's not the same for every team, and it's not even the same for the same team on different weeks.

[Elessar78]

I don't know, do you really think third or fourth stringers going against third or fourth stingers is an accurate model?  How often do we see guys come in and light it up at the end of preseason games only to ride the pine the rest of the year (or even get cut?)

It's apples to apples though right? There's a big jump between these guys and the starters (sometimes not) but, as a coach, you can gauge the effectiveness of such an approach. 

[Riggo-toni]

Why is this approach considered so novel? Teams that play against Haslett's defenses have almost never had to punt?

[Lombardi's_kid_brother]

This isn't a new story. It's pretty well-established that - based on math - teams should punt a lot less.

 

The problem is, not even Belichick is immune from crazy talk radio rantings the second the 4th and 1 play from your own 40 gets stopped.

[Redskins Diehard]

Interesting strategy.....for the high school level.

I've heard of little league baseball coaches use a "run until they tag you" strategy....I can't believe nobody has tried that in MLB yet

[twa]

reminds me of the why design plays for less than 10 yd gains theory

[TD_washingtonredskins]

This isn't a new story. It's pretty well-established that - based on math - teams should punt a lot less.

 

The problem is, not even Belichick is immune from crazy talk radio rantings the second the 4th and 1 play from your own 40 gets stopped.

 

And I think that's the major obstacle right now in the pro game. Someone would almost have to come into the NFL with the primary goal of implementing this strategy. That's dangerous since, as a coach, you are calling your priorities into question (winning the game vs. proving a point).

 

Then again, how dumb would a group of critical reporters look when the coach can whip out a binder and quantify exactly how much better his team's chances of winning were based on his decision? I'm sure it would hold up just fine to some beat writer's opinion that this strategy can't work "on Sundays" or "up here" if a coach said that he believes his moves provided his team with a 18% better chance to win the game.

 

I'd love to see it, quite honestly.

[Redskins Diehard]

And I think that's the major obstacle right now in the pro game. Someone would almost have to come into the NFL with the primary goal of implementing this strategy. That's dangerous since, as a coach, you are calling your priorities into question (winning the game vs. proving a point).

 

Then again, how dumb would a group of critical reporters look when the coach can whip out a binder and quantify exactly how much better his team's chances of winning were based on his decision? I'm sure it would hold up just fine to some beat writer's opinion that this strategy can't work "on Sundays" or "up here" if a coach said that he believes his moves provided his team with a 18% better chance to win the game.

 

I'd love to see it, quite honestly.

It would probably take quite a bit of time to develop enough data to prove that it is mathematically the best option.  Especially if there is only one coach trying it. 

[TD_washingtonredskins]

It would probably take quite a bit of time to develop enough data to prove that it is mathematically the best option.  Especially if there is only one coach trying it. 

 

Some data points are available now. Let's say Shanahan was going to start doing this in 2014 (assuming he sticks around, of course). He could probably look right now at how often our defense yields points based on our opponent's starting field position. That would be at least a start. Then, he could use our current fourth down conversion percentage which would actually represent a worst-case scenario.

 

So, if the decision in question occurs at our 30 yard line (on 4th and 5) and we average a net punt of 35 yards, we can balance the likelihood of our opponent scoring from their own 35 vs. our 30. Somehow you factor in that we had a X percent chance of converting the 4th and 5 and you have your equation.

 

Obviously, I would assume that before any coach implemented this strategy, he would run the numbers to make sure it would actually benefit the team.

[Elessar78]

There's lots of data now. All NFL teams have huge, complete databases on plays going back who knows how long now. They can sort to an unimaginable degree. 

 

I don't see what a team that's completely out of it has to lose either. I mean, it would put you into a much better slot to draft high if it bombs, right? Plus, no one can criticize you for doing it since you're bad anyway and they'll think you're tanking for the draft pick. 

[Redskins Diehard]

Some data points are available now. Let's say Shanahan was going to start doing this in 2014 (assuming he sticks around, of course). He could probably look right now at how often our defense yields points based on our opponent's starting field position. That would be at least a start. Then, he could use our current fourth down conversion percentage which would actually represent a worst-case scenario.

 

So, if the decision in question occurs at our 30 yard line (on 4th and 5) and we average a net punt of 35 yards, we can balance the likelihood of our opponent scoring from their own 35 vs. our 30. Somehow you factor in that we had a X percent chance of converting the 4th and 5 and you have your equation.

 

Obviously, I would assume that before any coach implemented this strategy, he would run the numbers to make sure it would actually benefit the team.

I don't think the equation is as simple as that but I am sure there is someone much more familiar with decision theory that could weigh in on it.  It isn't simply analyzing the likelihood we convert on 4th down to the likelihood that they score in both scenarios.  I believe it is a comparison of the likelihood we score vs. the likelihood they score given the 2 different decisions.