Are punters getting better?
Google Visualization API Sample
Punts
The second most unloved play in the NFL (behind the extra point and, possibly, the Blaine Gabbert pass). When they show a punt on DirecTV's Redzone Channel, Andrew Siciliano actually apologizes, as if it were an errant f-bomb that made it past their tape delay.
The purpose of this post is to show some statistical love for this less glamorous element of the game. Much has been made of the improvement in field goal accuracy over the years. Do we see a similar improvement in punting? What I found is that punters are improving, having added about 4 net yards to their punts over the past ten years.
Data
For this analysis, I used the play by play data Brian Burke posts on his site, looking at seasons 2002 through 2011. From the fields and the play description, I was able to parse out which plays were punts, as well as the following data fields:
- Field Position (where the team punted from)
- Gross Yards (how far the kick went)
- Net Yards
- Result - Did the play end in a touchback, fair catch, downed by the kicking team, muff, return, or block?
- Penalty Yards (if any) assessed on the punt
I only focused on "normal" punts for this analysis, meaning I excluded any punts that were blocked or that somehow resulted in the punting team retaining possession. If a punt had to be rekicked due to penalty, I am only counting the second "official" punt.
A note on penalty yards: Net Yards in my analysis below is also net of any penalty yards assessed on the punt. In general, penalties favor the punting team as the most common penalty called is holding on the return team. Average penalty yards in favor of the punting team has actually declined very slightly over the past ten seasons, so it is not a contributor to the increase in net punting yards shown below.
Net Yards Per Punt
The first stat I looked at was net yards per punt, controlled for field position. Punts 40 yards from the end zone tend to net fewer yards than punts from 80 yards away. If teams have been changing where they punt from over time (e.g. punting less when in their opponent's territory), that will show up as an increase in net yards per punt, but doesn't really say much about the punters themselves.
Net Yards Per Punt
The first stat I looked at was net yards per punt, controlled for field position. Punts 40 yards from the end zone tend to net fewer yards than punts from 80 yards away. If teams have been changing where they punt from over time (e.g. punting less when in their opponent's territory), that will show up as an increase in net yards per punt, but doesn't really say much about the punters themselves.
So, for each season, I took the average net yards from each yardline and weighted them by a common distribution, representing the total distribution of punts by field position from the 2002-2011 seasons. Here are the results:
In 2002, the average punt netted 34.4 yards. By 2011, that had increased to 38.2 yards, where once again, I am controlling for differences in field position (i.e. where teams punted from). Here is another way of looking at the data which doesn't rely on any manipulation:
As you can see, no matter where punters are kicking from, they appear to be netting more yards now than they have in the past (some statistical noise not withstanding).
From Long Distance
From Long Distance
I decided to look more closely at this change in punting yards, with a focus on where the team is punting from. Here are results for punts where the kicking team is greater than 65 yards away from the endzone:
From this area of the field, there is little risk of a touchback, so punting is more about raw leg strength. At this distance, gross yards per punt have been increasing at a rate of 0.53 yards per season, and net yards have increased by about 0.40 yards per season. So, kickers appear to be getting stronger, but perhaps the slightly lower gain in net yards is due to the phenomenon of "outkicking your coverage"? Time permitting, I would like to take a closer look at this and see of "outkicking your coverage" is a real thing, or just something that gets repeated by NFL announcers.
Out of your goalpost's shadow
What about when you get closer to the endzone? The graph below shows net yards when punting between 50 and 65 yards from the endzone.
Out of your goalpost's shadow
What about when you get closer to the endzone? The graph below shows net yards when punting between 50 and 65 yards from the endzone.
We see a similar improvement from this area of the field as well. Although net yards appear to be slightly outpacing gross yards here, with net yards improving at a rate of 0.44 yards per season, but gross yards improving by 0.40 yards per season.
Pinning them deep
Pinning them deep
Inside the 50, punting becomes less about strength, and more about finesse and accuracy. The previous results have clearly established that punters' legs are getting stronger. Are they also getting better at finding that sweet spot between the 20 and the goal line? Here are the results for punts from between 35 and 50 yards from the endzone:
While gross yards per punt has stayed relatively flat, net yards still shows improvement, at a rate of about 0.26 yards per season. I decided to dig into this a bit more and see what was driving this improvement in net yards.
Touchback Probability
Touchback Probability
If punters are improving their net yards when kicking from inside the 50, that probably means they are getting better at avoiding touchbacks. The graph below shows the percentage of punts that resulted in touchbacks by field position, split by the 2002-2006 and 2007-2011 seasons.
This graph is telling for two reasons. One, between 40 and 50 yards from the endzone, touchbacks are becoming significantly less likely, which should help explain the improvement in net yards per punt mentioned previously. Secondly, between 55 and 65 yards from the endzone, touchbacks have become more likely, probably as a result of improved leg strength.
So, if punting teams are avoiding touchbacks more when punting from inside the 50, that must mean they are downing more punts, right? Here is the percentage of punts downed by field position, once again split by the 2002-2006 and 2007-2011 seasons.
So, if punting teams are avoiding touchbacks more when punting from inside the 50, that must mean they are downing more punts, right? Here is the percentage of punts downed by field position, once again split by the 2002-2006 and 2007-2011 seasons.
The data is noisy, but there doesn't appear to be any clear shift to more downed punts when punting from inside the 50. If anything, teams appear to be downing punts less often, not more. So, if teams aren't downing the ball more, what else could explain this decrease in touchbacks?
Fair Catch Probability
The graph below shows the percentage of punts resulting in a fair catch, split between seasons 2002-2006 and 2007-2011.
Fair Catch Probability
The graph below shows the percentage of punts resulting in a fair catch, split between seasons 2002-2006 and 2007-2011.
Between 40 and 50 yards, there appears to have been fairly significant increase in how often a fair catch occurs. Which raises an interesting question: Is this because punters are getting better at finding that sweet spot on the field where the returner is forced to fair catch the ball? Or have returners been getting more conservative when making that decision?
Conclusion
Conclusion
Punters are getting better, most clearly when it comes to leg strength, improving by about a half yard per punt per season since 2002. Punting teams have also gotten better at playing the finesse game inside the 50. My analysis above shows that touchbacks are becoming less frequent and that fair catches are becoming more frequent. But I think this may be an area for further investigation.
Appendix
Appendix
There's no real point behind the following, but I'm sharing because I found it interesting. Here are graphs showing the probability of a touchback, fair catch, and a downed punt by field position (where field position signifies where punting team is kicking from).
2 comments:
This is a great writeup, thanks!
Thanks Andrew. It was interesting playing around with the detailed play by play data.
Post a Comment
Note: Only a member of this blog may post a comment.