Stat Analysis
Advanced analytics on player and team performance

To Return or Not to Return

To Return or Not to Return
Photo: USA Today Sports Images

Guest column by Jack Duffin

During the Week 8 matchup between Green Bay and the Los Angeles Rams, Packers returner Ty Montgomery ran a kickoff out of the end zone. He proceeded to fumble the ball away, costing Aaron Rodgers, one of the best quarterbacks in the league, what could have been a game-winning drive. While the team had instructed Montgomery to take a touchback, he decided he knew best and took a risk that backfired, and led to Green Bay trading him. Inspired by Montgomery, I wanted to look much deeper into whether teams should ever take the risk of running the ball out of the end zone on kickoffs, and whether the risk is worth the reward.

In both 2016 and 2018 the rules in the NFL regarding kickoffs changed, and questions were raised over how teams would react. With touchbacks moved to the 25-yard line (As of 2016), why would teams ever leave the end zone when they could get a free 25 yards without having to make a play? I looked up every kickoff during the 2018 regular season that was returned out of the end zone to see how successful teams were. I have used the data from NFL.com's play-by-play game center. I included all the plays that landed 1 yard or more into the end zone and were returned. I am also measuring each play by the yard line to which the ball was returned, not the distance of the return. If a ball is caught 3 yards deep in the end zone and recorded as 22-yard return, this means the ball is spotted at the 19-yard line, and that's what I'm looking at. I'm not interested in how far the ball has travelled, only where it ended.

I measured 251 kickoffs that were returned out of the end zone during the 2018 season; the results are in the table below. The aim of every return at a minimum is to get to at least the 26-yard line, as you can get to the 25-yard line just by taking a knee. Unfortunately, NFL teams only managed to make the 26-yard line on 29.3 percent of their attempts in 2018. This means that in the vast majority of cases, teams would have been better suited taking the touchback for the free 25 yards.

Kickoffs Returned out of End Zone, 2018
Team Frequency Avg. Yd
Line
26-Yd
Line
35-Yd
Line
50-Yd
Line
TDs TOs
MIA 20 27.7 35% 5% 5% 1 0
LAR 16 21.2 13% 6% 0% 0 0
NYG 15 22.8 27% 13% 0% 0 0
LAC 14 21.4 29% 0% 0% 0 0
ATL 13 25.8 31% 15% 8% 0 0
SF 10 28.2 20% 10% 10% 0 0
NE 9 21.6 22% 11% 0% 0 0
NO 8 26.4 25% 25% 0% 0 0
TB 8 25.5 50% 13% 0% 0 0
BAL 8 24.4 50% 13% 0% 0 0
MIN 8 22.5 25% 13% 0% 0 0
CLE 8 21.9 13% 0% 0% 0 0
CAR 8 19.1 13% 0% 0% 0 0
NYJ 7 31.3 71% 29% 0% 0 0
SEA 7 29.7 29% 14% 14% 0 0
TEN 7 24.4 43% 0% 0% 0 0
KC 7 24.3 43% 0% 0% 0 0
DET 7 23.1 29% 0% 0% 0 1
PIT 7 23.1 29% 0% 0% 0 0
CHI 7 21.3 29% 0% 0% 0 0
DEN 7 20.7 14% 0% 0% 0 0
CIN 6 29.2 60% 17% 0% 0 0
HOU 6 27.5 83% 0% 0% 0 0
GB 6 22.7 17% 17% 0% 0 2
ARI 6 18.7 0% 0% 0% 0 0
JAX 5 29.0 40% 40% 0% 0 0
PHI 5 25.0 20% 0% 0% 0 0
DAL 5 22.6 20% 0% 0% 0 0
BUF 3 23.0 33% 0% 0% 0 0
IND 3 20.0 0% 0% 0% 0 0
WAS 3 19.3 0% 0% 0% 0 0
OAK 2 21.5 0% 0% 0% 0 0
Total 251 24.1 29% 8% 2% 1 3

There were actually 256 kickoffs that were returned out of the end zone in 2018. Of these I have excluded four that were either the final or penultimate play in the half; teams might have treated those differently, as a big return might have given them a chance to score and field position wasn't as important. I also excluded one attempt during the Week 5 game between the Rams and Seahawks as the kick resulted in a flag and was then retaken. For penalties on the receiving team, I measured the return to the point at which the infringement occurred. For penalties on the kicking team, I measured the point where the return finished, not where the foul took place. I wanted to do everything I could to minimize the impact of penalties, as they can happen on plays where a touchback occurs too.

I separated fumbles into two categories. Fumbles during or directly after the catch where the receiver has regained the ball and made a return have been treated as part of the return yardage. When a fumble has occurred further up the field and the ball has been collected by a teammate or the opposition, the point where the player lost control of the ball has been used as the yardage for the return.

Only five teams (Texans, Jets, Bengals, Buccaneers, Ravens) managed to make the 26-yard line on even 50 percent of their attempts. Four teams (Raiders, Colts, Redskins, Cardinals) failed to make the 26-yard line even once during the season, though the Cardinals were the only team in that four to have more than three attempts. The Raiders had the fewest returns out of the end zone with two, followed by the Colts, Redskins, and Bills with three each. Maybe these teams were quickest to learn that taking a touchback was a sensible strategy?

The Dolphins had 20 returns out of the end zone, most in the league, followed by the Rams, Giants, Chargers, and Falcons. I'm not sure what special teams coordinators are trying to achieve by continuing to do something that isn't working. Maybe if they evaluate a full season of data they will make the right decision to make on a regular basis. Still, sometimes the decision is solely on the player, as we saw with Montgomery.

Average Return

The Jets had the best results on returns out of the end zone, reaching the 31.3-yard line on average. Only nine other teams (Seahawks, Bengals, Jaguars, 49ers, Dolphins, Texans, Saints, Falcons, Buccaneers) averaged better than the 25-yard line on returns out of the end zone.

Great Plays

I considered a great play to be a team getting to the 35-yard line -- 10 yards more than what a touchback would have given them. We saw this play achieved eight percent of the time by teams across the entire season, showing the chances of having a great play on a return out of the end zone was very low. Only three teams (Jets, Jaguars, Saints) made the 35-yard line on 25 percent of their attempts.

Houston had interesting results -- they were the most consistent team at getting to the 26, doing so 83 percent of the time, but they didn't reach the 35-yard line even once.

Flipping the Field

I also decided to check how often teams flipped the field by returning the ball to the 50-yard line. I was not expecting this to be happen very often, but this is the sort of thing that would lead me to believe that returns out of the end zone could be worth the risk of losing yardage on some plays, because your chances of scoring on shorter drives would be greatly improved. Only four plays -- two percent of all end zone returns -- managed to get to the halfway point of the field.

Touchdowns vs. Turnovers

The most dramatic possible results of a return on a kick-off are that a team scores a touchdown or turns the ball over. In 2018 we saw one touchdown and three turnovers. The only touchdown scored was by Jakeem Grant of the Miami Dolphins against Tennessee in Week 1. The Packers had two of the turnovers -- the one by Montgomery against the Rams, and one by J'Mon Moore against the Jets in Week 11. The other was suffered by Detroit's Ameer Abdullah, who fumbled the ball away against Seattle in Week 8.

Key Takeaways

I think it is important for all teams to seriously consider whether they should be returning kickoffs out of the end zone. With both the chance of getting to the 26-yard line as low as it is and the average return failing to reach the 25, it would be better to take a knee and start drives as near to the opponent's goal line as possible.

If I was a special teams coordinator, I would want to reach the 26-yard line at least 75 percent of the time, the 35-yard line 25 percent of the time, and midfield 5 percent of the time before I considered taking the ball out of the end zone. None of the 32 NFL teams hit all three of those targets, although each was met by at least one team.

With so little value coming from returns out of the end zone, it wouldn't surprise me if the NFL moves in the future to consider all kicks into the end zone as touchbacks and continue to reduce the amount of returns that occur in the league in an effort to reduce concussions.

It will be interesting to see what the data says in 2019. We have a solid sample of more than 250 plays, but more data would give a better indication of what is happening. The 2019 data will allow us to see if teams change their style this offseason and attempt fewer returns. With the league making changes to increase the amount of scoring, we are likely to see more and more kickoffs, giving teams the chance to make more returns if they want.

Jack Duffin is the co-host of the Paul Brown Podcast, a Daily International Browns Podcast based out of London, England. You can follow him on Twitter @JackDuffin.

Comments

27 comments, Last at 23 Feb 2019, 5:24am

1 Re: To Return or Not to Return

"In 2018 we saw one touchdown and three turnovers. "

Am I missing something here?

Play Index shows five returned kickoffs for touchdowns:

9/9 Jakeem Grant
9/9 Darius Jennings
10/21 Cordarrelle Patterson
12/16 Richie James
12/23 Andre Roberts

15 Re: To Return or Not to Return

Limiting this particular stat to just those returns from the end zone seems a bit over-constrained, I'd agree.

The general point remains the same: on kickoff returns, turnovers outnumber TDs by about 3 to 1. (for all returns this past season, it's 17 to 5)

Returning the ball out of the end zone is a triumph of hope over experience.

3 Re: To Return or Not to Return

Interesting. I think this confirms what most thought. Not that I want to assign more work, but have you considered the companion piece - which teams kick the most touchbacks (so they don't have to do coverage), and which teams kick to 0 - 10, forcing a return (so they can cover within the 25).

8 Re: To Return or Not to Return

It is something I do want to work on at some point, I just didn't have the time while writing this.

Some places I looked recorded a touchback and a kick out of the back of the endzone as a touchback as well so I need to find the right place to get the data.

27 Re: To Return or Not to Return

In that vein, I'd love to see which teams gave up the most positive returns (it might correlate with teams that gave up the biggest returns overall. Or not, since the sample sizes are so small). Surely that plays a part in the return team's strategy - facing some special teams, you might warn your returner to never risk it. Or conversely, you might notice a weakness in a kickoff unit and greenlight your returner. Even ten percent greater chance of breaking a long return would be a major incentive.

4 Re: To Return or Not to Return

"If I was a special teams coordinator, I would want to reach the 26-yard line at least 75 percent of the time, the 35-yard line 25 percent of the time, and midfield 5 percent of the time before I considered taking the ball out of the end zone."

Seems reasonable, if a little arbitrary. Could you explain the thought process behind this benchmark?

9 Re: To Return or Not to Return

It's arbitrary. I was bouncing some ideas around with friends after writing the piece and wanted to come to a threshold that would encourage me to return kicks into the endzone. Those were the numbers I would look at personally but everyone can come to their own conclusion.

21 Re: To Return or Not to Return

The question for the coach is whether they want best average position or want to try for the rare return that really flips the field. I suspect for most drives there really isn't a huge amount of difference between starting at the 20 or the 30.

Generally speaking, people are willing to accept lower expected return if the probability of a jackpot is there.

5 Re: To Return or Not to Return

When I saw that Miami was leading the league with 20 returns and they were averaging 27.7 yards per return, at first I thought that Miami was correctly taking advantage of an enhanced ability to return kickoffs - gaining 2.7 yards per return from the end zone. Plus, they scored the only TD of the season. But, that 27.7 yard average is hugely impacted by that one TD. Assuming my math is correct, they had 20 returns for 554 yards. If we remove that TD return, they had 19 returns for 454 yards - an average of 23.9 yards per return.

I wonder if scoring a TD in week 1 made the coaching staff and/or players think they had a good return game, so they spent the whole season trying to recapture lighting in a bottle.

6 Re: To Return or Not to Return

I think that would absolutely affect your team's thinking. As rational as I generally am, it would take me at least 6-8 weeks to figure out the returns weren't worth it.

(Interestingly if I recall correctly when the Pats gave up the Miracle in Miami play they kicked off short to about the 15 and it got returned to the 30. That info plays out in two ways, Miami returning kicks out of endzone probably means you don't go for the endzone because they're going to return it and their average suggests they get it out past the 25. The mortar kickoff seems to gives you a better chance of bottling them up inside the 25)

7 Re: To Return or Not to Return

"With so little value coming from returns out of the end zone, it wouldn't surprise me if the NFL moves in the future to consider all kicks into the end zone as touchbacks and continue to reduce the amount of returns that occur in the league in an effort to reduce concussions."

Wondering what would happen if the NFL allowed teams to kick off out of bounds without penalty? (Just like punts are allowed to go out of bounds)

Would it encourage kicking teams to do that and discourage returns as the NFL would like?

11 Re: To Return or Not to Return

That would be a really good idea, it would certainly make kick offs more important meaning teams can get a massive advantage off of them without any serious contact between players.

I think the AAF model might be the way the NFL goes as it will have been tried and tested first.

18 Re: To Return or Not to Return

I am in favor of the AAF model, but after watching one game like that, it's a little weird to get used to. After the score, the other team just lines up on the 25. It lacks a sense of "closure" that kickoffs have.

13 Re: To Return or Not to Return

Wondering what would happen if the NFL allowed teams to kick off out of bounds without penalty?

Two Problems:

Normal kickoffs would likely cross the sideline high in the air making getting the spot right difficult.

Currently an untouched kickoff is a live ball. A low kickoff that bounced in-bounds instead of out could create a more injury prone situation. Fun, but probably not what the league wants.

16 Re: To Return or Not to Return

Being from Canada, I watch a lot of CFL and the kicking game in the CFL is much more impactful and entertaining. There are a lot more kickoff and punt returns because they have to give the player a five yard cushion to receive the ball. If you have a quick and shifty returner, they can make that first tackle attempt miss and then they can usually get significant yardage. It might also be enhanced by the wider field which can allow the returner to string out the coverage and find a lane.

If a player intentionally violates the 5 yard rule it is a 15 yard penalty and in some cases if they just blast the guy they could also be given an roughing penalty which is another 15 yards.

Frankly, i would rather have the CFL rules or the AAFL rules than any of the current NFL punt or kickoff rules. What fun is there is watching a team run down the field when it is almost always a touchback or a fair catch. Even downing near the goaline isn't very exciting, nor is it watching a bunch of guys stand around the ball until it stops moving.

17 Re: To Return or Not to Return

i dio like punting and feel punting is still good in the nfl. watching guys down punts neasr the goal line is usually interesting. Gerat one was done by Pates this past season. Was in Novem,ber, maybe early December. Could look it up if I have time. Probably best downing of a punt ever.

Kickoffs in NFl really blow now. most of time it is touchback or penalty. youi see guy return bal for 40 yards and most of time you can;'t get excited becuause you see a flag waa thrown on play.

10 Re: To Return or Not to Return

This is a nice start.

The question that occurs to me, though, is how much loss will a team be willing to accept for the occasional big return. To take kind of an extreme example, imagine a team that either returns the ball to its 24 or scores a touchdown. How often do they have to score to make those returns worthwhile? What if the unsuccessful returns are to the 20? Maybe this involves using expected win probabilities, and maybe it still turns out that taking the knee is better, but I think there's a little more nuance to coming up with the answer.

12 Re: To Return or Not to Return

I agree, this is good stuff. Yes, the key investigation here would be to understand under what circumstances the risk is worth it. If I am down by 4 points and it's inside 2 minutes should I run it out from 1-yard deep. How about 5 yards deep? Win probability based on expected return outcomes and so forth.

19 Re: To Return or Not to Return

That's basically the Dolphins. They had 19 kick returns averaging 24 yards, and a 20th that scored a TD. So they lost 1 yard 19 times, yet gained a TD. That seems worth it.

Though, in that average of 24 yards, I wonder how many they ended up with the ball inside their 10. I think "Hidden game of football" said having the ball on your own 25 gives you 0 expected points. On your own 12.5 gives you -1 expected points. So if the Dolphins ended up with the ball inside their own 12.5 seven or more times, they lost expected points.

25 Re: To Return or Not to Return

I'd love to see the end of game/time remaining considered as well. Nothing drives me crazy like seeing a returner bring the ball out of the end zone when there's less than 2 minutes to go while losing/tied.
Put a table of how quickly the ball gets advanced from scrimmage in a 2 minute drill vs the results from that return out of the end zone table and I would think it's a complete no-brainer.

From pro-football-reference.com I can find 6 kickoffs taken out of the end zone with under 2 min left by a team either losing or tied with time wasted on the return and net yards:
9/23 Saints/Falcons 6 seconds +1 yard
10/14 Steelers/Bengals 6 sec -2yds
10/15 Pack/49ers 6 sec +22 yds (includes 15yd penalty on kicking team)
10/21 Panthers/Eagles 5 sec +5 yds
11/19 Rams/Chiefs 5 seconds +1 yard

Then the NFC Title game...Rams tie it with 15 seconds left (after Saints burn the inevitable 'ice the kicker' timeout). Kickoff goes 7-8 yards deep, they bring it out to...the 29, in 7 seconds...7 seconds of clock for 4 yards gained when all you need is a field goal. Better chance of that offense going 35 yards (to approx FG range) in 14 seconds with a timeout left (or 2 timeouts if you hadn't burned one to ice the kicker), or getting a return into FG range or TD? Makes no sense.

All told these returns burned 35 seconds of clock to gain 16 yards and 15 more by penalty. Only one truly positive outcome in 6 tries (and that one was only positive because of a 15 yd penalty, otherwise it's 7yds in 6 seconds), weighed against the possible negative outcome (turnover) seem like taking a knee should be standard operating procedure, especially when the ball flies up and down the field from scrimmage.

I understand with under 10 seconds left, try for the return TD which is a greater possibility than the 'Miami Miracle' type play from scrimmage, but with 30 or 40 seconds or more left (especially with timeouts in hand), why would you burn clock to get what is most likely the same field position, quite possibly worse? It always seems completely pointless to me.