Writers of Pro Football Prospectus 2008

Most Recent FO Features

ThomasEar10.jpg

» 2013 Play-Action Defense

Are the best defenses against play action the best against regular passes too? How much impact does play action really have in an NFL game, and does it correlate from year to year?

19 Oct 2007

How Good Is Devin Hester?

by Vince Verhei

Few plays in football can top a kick returned for a score for instant excitement and dramatic impact on a game. NFL fans these days get to watch two of the best returners of all-time as Dante Hall's career winds down and Devin Hester's blossoms.

But are Hester and Hall the greatest kick returners playing now, or are they the best of all-time? Or is the opposite true -- do their accomplishments pale when compared to legends of the past?

I looked at this issue in a recent Extra Point article, and found that by one measure, Devin Hester was the greatest kick returner of all time. Some readers agreed, some were skeptical, and one was content to point out that I was misusing the term "touchdown rate." A lot of questions were raised, and I'm going to address them here.

Before we can get into that, we need to start at ground level for those who missed the original article. Anyone who did read the original piece can skip down to the "New Research" subhead.

I was first moved to do this research by one Gregg Easterbrook's TMQ columns on ESPN.com, when he dropped this nugget of info:

"Devin Hester has fielded 93 kicks in his career and run eight back for touchdowns."

It turns out that Easterbrook's math was slightly off, but two weeks' worth of games have been played since then anyway. Hester's career numbers as of Sunday, October 14, 2007 -- after returning another kick for a score against Minnesota -- include three touchdowns on 35 kickoff returns, five touchdowns on 67 punt returns, and one touchdown on one return of a missed field goal. That's a grand total of nine touchdowns on 103 returns, an average of one touchdown per 11.4 returns. How does that compare to the greatest kick returners in league history?

To answer that, we have to decide what makes a great kick returner. The NFL Record and Fact Book lists leaders for combined kick returns on its kickoff returns page. There is also a page for punt return records.

The all-time leader in combined kick return touchdowns is Brian Mitchell with 13 (nine punts, four kickoffs). Tied for second with 12 apiece are Eric Metcalf (10 punts, two kickoffs) and Dante Hall (six punts, six kickoffs). But since those are career totals, players with shorter careers (such as Hester) obviously don't compare. Mitchell played 13 years; Hester has played just 22 regular-season games.

We should look at each player's touchdown rate, not touchdown total. And while we're figuring numbers for Mitchell, Metcalf, and Hall, we may as well look at other great players from the kickoff and punt returns pages.

Our list looks like so:

  • The all-time record holders in combined kick return touchdowns in a single season. Hester actually set this record with five last year (He is not given credit for returning the missed field goal -- more on this later). Nine men (Jack Christiansen, Emlen Tunnell, Gale Sayers, Travis Williams, Cecil Turner, Billy "White Shoes" Johnson, Rick Upchurch, Hall, Eddie Drummond) have scored four times on kick returns in a season.
  • The top three men in total career kickoff return yardage: Mitchell (14,014 yards), Mel Gray (10,250), and Glyn Milburn (9,788).
  • The three all-time leaders in kickoff return average: Gale Sayers (30.56), Lynn Chandnois (29.57), and Abe Woodson (28.69).
  • The all-time leaders in kickoff return touchdowns. The NFL lists several men here in three groups: Those with four career scores on kickoff returns (Turner, Ron Brown, Jon Vaughn, Andre Coleman, Tamarick Vanover, Tony Horne, Mitchell, Darrick Vaughn and Terrence McGee); those with five (Bobby Mitchell, Woodson, Timmy Brown, and Michael Bates) and those with six (Ollie Matson, Sayers, Williams, Gray and Hall).
  • The career leaders in punt return yardage: Brian Mitchell (4,999), David Meggett (3,708) and Darrien Gordon (3,601).
  • The career leaders in punt return average: George McAfee (12.78), Christiansen (12.75) and Claude Gibson (12.55).
  • The career leaders in punt return touchdowns. Metcalf holds the record with 10. Brian Mitchell has nine. Three men have eight: Christiansen, Upchurch and Desmond Howard.

Counting Hester, that's 32 of the greatest kick returners in NFL history. I added one more name to the list: Deion Sanders. I included his name for two reasons: 1) With six career touchdowns on punt returns and three more on kickoff returns, he just missed showing up on these lists, and 2) His jersey is hanging on my wall, and I wanted to see where he stacked up.

Before we get to the final numbers -- found using this website -- we have to make one more adjustment: We're taking Hester's missed field goal return away. Call it fair, call it unfair, but since none of these other guys ever had an opportunity to return a missed field goal (as far as I know), we're removing it to even the playing field for everyone.

After crunching the numbers, I found that no player could match Hester's touchdown rate, and dubbed him the best kick returner of all time. The issue, however, was not decided .


New Research

The original study I published is already out of date, as three players -- Hester, Hall, and McGee -- are still active, and Hester and McGee have each scored since that study was posted. I've also added Allen Rossum to the list; his kickoff return touchdown for the Steelers this year brings his ten-year career total to seven kickoffs returned for scores.

With that, the updated list as of Week 6 of the 2007 season:

PLAYER KICK
RETURNS
KR TDS PUNT
RETURNS
PR TDS COMBINED
RETURNS
COMBINED
TDS
TD
RATE
Devin Hester 35 3 67 5 102 8 12.8
Gale Sayers 91 6 27 2 118 8 14.8
Travis Williams 102 6 13 1 115 7 16.4
Jack Christiansen 59 0 85 8 144 8 18.0
Bobby Mitchell 102 5 68 3 170 8 21.3
Ollie Matson 143 6 65 3 208 9 23.1
Darrick Vaughn 103 4 0 0 103 4 25.8
George McAfee 18 2 112 2 130 4 32.5
Cecil Turner 108 4 27 0 135 4 33.8
Jon Vaughn 103 3 0 0 103 3 34.3
Terrence McGee 174 5 0 0 174 5 34.8
PLAYER KICK
RETURNS
KR TDS PUNT
RETURNS
PR TDS COMBINED
RETURNS
COMBINED
TDS
TD
RATE
Tony Horne 143 4 7 0 150 4 37.5
Deion Sanders 155 3 212 6 367 9 40.8
Claude Gibson 16 0 110 3 126 3 42.0
Timmy Brown 184 5 71 1 255 6 42.5
Rick Upchurch 95 0 248 8 343 8 42.9
Abe Woodson 193 5 123 2 316 7 45.1
Andre Coleman 193 4 43 1 236 5 47.2
Dante Hall 380 6 200 6 580 12 47.8
Ron Brown 199 4 0 0 199 4 49.8
Billy Johnson 123 2 282 6 405 8 50.6
Emlen Tunnell 46 1 258 5 304 6 50.7
PLAYER KICK
RETURNS
KR TDS PUNT
RETURNS
PR TDS COMBINED
RETURNS
COMBINED
TDS
TD
RATE
Eric Metcalf 280 2 351 10 631 12 52.6
Lynn Chandnois 92 3 66 0 158 3 52.7
Tamarick Vanover 226 4 197 4 423 8 52.9
Darrien Gordon 5 0 314 6 319 6 53.2
Eddie Drummond 213 2 108 4 321 6 53.5
Mel Gray 421 6 252 3 673 9 74.8
David Meggett 252 1 349 7 601 8 75.1
Desmond Howard 359 0 244 8 603 8 75.4
Michael Bates 373 5 10 0 383 5 76.6
Brian Mitchell 607 4 463 9 1070 13 82.3
Allen Rossum 429 4 257 3 686 7 98.0
Glyn Milburn 407 2 304 1 711 3 237.0


With just one score on a handful of returns, McGee passed Timmy Brown, Claude Gibson, Deion Sanders and Tony Horne. The lesson here is that kick return stats are very volatile. By their nature, they're derived from small sample sizes, and a play or two here or there can always have significant impact. In fact, had Hester not scored against Minnesota, he would have been perilously close to slipping behind Sayers into second place.

(We've also learned that if I'm not supposed to use the term "touchdown rate," then I don't know what term I should use. Technically, it should be "average kick returns per touchdown," but that's awfully unwieldy, wouldn't you agree?)

With that out of the way, let's answer some questions that were asked by readers of the original article.

How did Dante Hall's 2003 season, when he was discussed as an MVP candidate, compare?

In 2003, Hall returned 57 kickoffs for two touchdowns, and also returned 29 punts for two touchdowns. That's a touchdown rate of 21.5 -- that's great, but not quite as good as the entire careers of the men at the very top of the list.

But that doesn't quite capture Hall's glory that year, because those four touchdowns came one at a time over four straight games. The Chiefs went 4-0 in those games, including twice by seven points or less, capped off by a 93-yard punt return touchdown in the fourth quarter that put the Chiefs ahead of the Broncos 24-23, the game's final margin. Over those four games, Hall returned a total of 21 kicks, for a touchdown rate of 5.3. I don't know, but I suspect that's a record.

Is it better for a kick returner to consistently rack up good returns? Or to get the occasional long touchdown that offsets dozens of other short returns?

I don't know this for certain, but I assume that it's better to have a boom-and-bust guy. This may seem counter-intuitive, and it goes somewhat against the nature of what Football Outsiders is built on; DVOA is designed to punish boom-and-bust runners and passers, while rewarding those players who pick up chunks of yardage at a reliable rate.

There is a difference, though, between kick returns and plays from scrimmage. A one-yard run counts as a bad play in DVOA because it hurts the team; they are now less likely to pick up another first down, and more likely to punt. Punt returns, however, are always positive plays, so long as they gain at least one yard and the ball isn't fumbled away. After any punt return, the offense is going to be facing first-and-10 (unless it's a really great return that sets up first-and-goal or outright scores). A one-yard punt return may be a below-average play, compared to what other punt returners might have done, but the bottom line is that player has helped his team by putting them one yard closer to the goal line than they were when he caught the ball.

Another factor to consider is the fluidity of field position. The chart run every year in the "Statistical Toolbox" portion of Pro Football Prospectus shows that yardage gained at either end of the field has more value than yardage gained towards the middle. When a kick returner catches a kickoff at his goal line, that first yard he gains is the most valuable yard on the field (well, it's tied with the last yard on the other end of the field). The second yard he gains is a little less valuable, the third yard less still. This diminishing process continues all the way out to the 50-yard line, when it reverses, and suddenly each yard becomes MORE valuable than the yard that came before it.

How does this information answer our question? Suppose Returner A and Returner B each catch two kicks at their goal line. Returner A gets his first kick out to the 40, but his second kick only out to the 20. Returner B returns each of his kicks out to the 30-yard line. Each has returned two kicks for a total of 60 yards. Who has helped his team more? The answer is Returner B, because the yardage between the 20 and 30 is more valuable than the yardage between the 30 and 40. If there was a Returner C who returned one kick to the 10 and one to the 50, he would be the least valuable yet.

This would indicate that a steady kick returner is better than an inconsistent, explosive guy, but the answer is not that simple. Eventually, that explosive guy is going to rip off a 100-yard touchdown that the reliable returner never will. And after he crosses the 50, each yard he gains will have increasing value. How many 20-yard returns will that make up for when his counterpart is picking up 30 every single time? We'll have to hit that in future research.

Do these statistics account for fumbles?

No they don't, and Hester certainly fumbles a lot -- eight times in 2006, and four times already in 2007. Unfortunately, I don't have historical data for fumbles, so I can't compare Hester to the other players on this list. It seems reasonable, though, to expect that very, very few of them fumbled as often as Hester has.

Is it fair to combine Hester at his peak to other men over their entire careers?

Probably not. As one poster pointed out, it's very doubtful that Hester is going to get better than he is right now. He's going to decline, like all players do. If we wanted to compare Hester to other players right now, the best way to do it would be to look at each player's peak value.

In 2002-03, Dante Hall returned seven kicks for scores. Two other men had returned six kicks for scores in back-to-back years: Gale Sayers in 1966-67 and Jack Christiansen in 1951-52.

And now, with eight return touchdowns in two years, Hester has bested them all -- and he still has ten games to go.

Of those four timeframes, Sayers has the best touchdown rate (8.0), followed by Christiansen (10.0), then Hester (12.8) and then Hall (24.6).

Which leads us to our next question...

Is it better to have a high peak value over a short career, or a lower peak value over a long career?

I don't think either one is necessarily better than the other. The best way to measure this would be to measure each player's value each season, in terms of yards and touchdowns, and compare them to what an average or replacement level returner that year would have done with the same number of opportunities. We'd then have a better picture of how each man fared compared to his peers, in the environment he played in. This would somewhat stifle the guys with short, brilliant careers like Sayers and Christiansen, and would boost guys like Hall and Eric Metcalf.

Is combining total kickoff return numbers and total punt return numbers into one figure the best way to do this?

In hindsight, probably not. The best way would be to use the process above to measure each guy's value as a kickoff returner, then do it separately for each player as a punt returner, then combine the two value totals into one sum, or a Bill James power/speed type number if you wanted to find players who excelled at both roles. The original study was spawned by one sentence Gregg Easterbrook wrote in Tuesday Morning Quarterback, and a desire to put that into historical perspective.

Which is more important: The return man, or his blockers?

This, of course, is the general issue with all football statistics: They're a measure of all 11 men on the field, not just the one with the ball in his hand. In baseball, we know a pitcher who walks a lot of guys is failing on his own. In basketball, we know a player who hits a high percentage of free throws is succeeding by himself. In football, every single individual statistic is reliant on the performance of teammates. Even field goal kickers are reliant on their snappers and holders.

With that being said, I think it's obvious that returners are far more important than their blockers. The Bears returned just one kick for a score in 2005. In 2006, with Hester, they had five. Which is more likely, that Hester was five times the returner that Bobby Wade was, or that the Bears' kick return units improved 500 percent?

It looks like answering these questions has opened the door to study new ones, which we'll have to do later this season or in the off-season.

(Ed. note: I'm guessing that if we were to look at yardage, rather than simply scores, the blockers would turn out to be more important than the return men except in the extreme cases like Hester. Also, I will get to the whole "how does squibbing to Hester affect the Chicago special teams DVOA?" question sometime soon, I promise. Really. -- Aaron)

Posted by: Vince Verhei on 19 Oct 2007

46 comments, Last at 22 Oct 2007, 7:10pm by JonL

Comments

1
by DGL (not verified) :: Fri, 10/19/2007 - 12:42pm

A general rule of scientific inquiry is that extreme cases open up new lines of investigation. Seems to me that Hester is one of those extreme cases that has opened up a whole new set of questions about special teams. I look forward to seeing what The Outsiders come up with.

It would be interesting to see how Chicago's average starting field position after kickoffs (with Hester on KO returns) compares to the rest of the league (or any Chicago KO returns without Hester back). That would seem to incorporate both the impact of Hester's returns and of kicking away from him.

2
by Will Allen (not verified) :: Fri, 10/19/2007 - 12:43pm

Yeah, fumbles really need to be accounted for. My evaluation of running backs has changed greatly now that I've considered the wide disparity in fumble rates. Also, it was much harder to avoid fumbles in decades past, once slow whistles and instant replay reversals are factored, so if Hester fumbling a lot more than somebody from the '60s or '70s, it really is quite significant.

3
by Jimmy (not verified) :: Fri, 10/19/2007 - 12:46pm

Is there a reason why Hester's KO return in the Superbowl is not included?

It was not in the regular season, but was still an NFL game. Are there any variables that playing in a Superbowl creates that discount the data?

With the available data on Hester being very small (compared to a guy like Brian Mitchell) why throw any usable data out?

4
by TGT (not verified) :: Fri, 10/19/2007 - 12:58pm

Going from 1 to 5 touchdowns is a 400% improvement, not 500% improvement

5
by dryheat (not verified) :: Fri, 10/19/2007 - 1:16pm

I'm not arguing this one way or the other, but I'm not sure the Bear's return units would have needed to improve 500 (or 400) percent. In any punt return, generally speaking, once the returner avoids the first, or at most, the first two enemies, the rest is easy. Especially since the rules have changed now where only the two outside guys can take off before the ball is kicked on punt returns (another variable that must be taken into account when comparing Hester with players from previous eras). If the Bears were able to get STers better able to hold up the gunners, or another deep man that can effectively neutralize the first man downfield, that would have gone a long way towards the assumed 400% improvement.

6
by Starshatterer (not verified) :: Fri, 10/19/2007 - 1:30pm

Is there a reason why Hester’s KO return in the Superbowl is not included?
At a guess, it's because the NFL keeps only stats for regular-season games for things like career records. (They obviously keep the post-season stats, but don't count them for records.)

7
by Rich Conley (not verified) :: Fri, 10/19/2007 - 1:54pm

"With that being said, I think it’s obvious that returners are far more important than their blockers. The Bears returned just one kick for a score in 2005. In 2006, with Hester, they had five. Which is more likely, that Hester was five times the returner that Bobby Wade was, or that the Bears’ kick return units improved 500 percent?"

They didn't improve 500%. They improved from -3.9% to +3.4% on kick returns.

They also improved in every single other special teams category, from coverage to kicking, so I'd say its MUCH more likely that the coverage team got better than that Hester is the difference.

8
by Tom (not verified) :: Fri, 10/19/2007 - 2:09pm

"so I’d say its MUCH more likely that the coverage team got better than that Hester is the difference."

You must have never seen Bobby Wade return a punt for the Bears.

9
by Rich Conley (not verified) :: Fri, 10/19/2007 - 2:13pm

8.

I do think Hester is A difference, but they've improved everywhere, not just on returns.

Didn't they change ST coordinators before the 2005 season? They've been constantly improving since then.

10
by dgc (not verified) :: Fri, 10/19/2007 - 2:37pm

I wonder if coaching mentality has changed as well. Was Brian Mitchell the best return guy for the Redskins, or was he the guy that Joe Gibbs and Norv trusted the most with ball security? I remember that when Gibbs really needed a return for a TD, he put Darrell Green in as the returner.

11
by SuperBears (not verified) :: Fri, 10/19/2007 - 2:55pm

One idea I would like to see expanded on squibbing is to take account the Bears "offense." When given the choice of Hester touching the ball (and remember he has had a number of long returns, not just touchdowns) or the Bears offense at the 40, kicking out of bounds sounds like a great idea. Something tells me (its probably on the premium database) but the Bears offense is probably really poor compared to the league average at scoring from their 35-40. Would teams have to change their strategy if he played for a high scoring offense, could you imagine giving Brady/Manning the ball on the 40 every drive?

12
by Jimmy (not verified) :: Fri, 10/19/2007 - 3:14pm

#7

The Bears kick return DVOA only improved from -3.9 to 3.4, but the punt returns improved from -2.5 to 11.7.

Hester didn't start to return kicks last year until the game against St Louis (the 12th game). That probably (at least partly) explains why the kick returns show less improvement.

One of the scariest things about Hester is that he has scored all those TDs despite only having returned both kick and punts in the last ten regular season games (out of 22) he has played in, he could have returned twice as many kicks during the period. That would pro-rate at his current rate of production to 6 KRTDs and 5PRTDs or 11TDs in 22 games, which is even freakier than his already outlandish numbers.

13
by Jimmy (not verified) :: Fri, 10/19/2007 - 3:35pm

Sorry to double post and it may be a little of topic, but as you will see there is a clear link.

I was wondering if anyone knew whether the Hester who plays for South Florida is a sibling of Devin?

14
by CA (not verified) :: Fri, 10/19/2007 - 4:04pm

Re: 3 Is there a reason why Hester’s KO return in the Superbowl is not included?
It was not in the regular season, but was still an NFL game. Are there any variables that playing in a Superbowl creates that discount the data?

I think there's a good chance that last year's Super Bowl was completely ruined due to weather and field conditions and that we may as well throw out every piece of data from that game as unreliable, the output of a contaminated experiment. Having issued that caveat, I do think you have a good point in general. Compiling post-season statistics, however, probably would make a study like this significantly more time-consuming.

On a related note, I wonder if the terrible field conditions that plague Soldier Field during bad weather have anything to do with Hester's success on returns. I'd be interested to know if a high proportion of Hester's TDs have come when the grass was wet and muddy (as it was last week).

15
by Eddo (not verified) :: Fri, 10/19/2007 - 4:53pm

14:
2006

PR @Green Bay in September
PR @Arizona in October
PR vs. Minnesota in November
KR @St. Louis in December
KR @St. Louis in December

2007

PR vs. Kansas City in September
KR @Detroit in September
PR vs. Minnesota in October

So of the 8 returns, only 3 came at home (and the Chiefs game was beautiful weather). Three came in domes, one in Arizona, and one in Green Bay in a sunny week one game of the 2006 season. So at first glance, it doesn't seem like field conditions really come into play.

16
by Vince Verhei :: Fri, 10/19/2007 - 5:15pm

#3: In almost any study done on this or any other website, postseason data will not be included. It's partly an issue of time, and even more an issue of selection bias. Most players in a given year don't make the playoffs, and so don't have a chance to add to their totals. If you look at any of the stats listed on this site -- DVOA, DPAR, ALY, Sack Rate -- they'll all be regular season only.

17
by Brian (not verified) :: Fri, 10/19/2007 - 5:31pm

Of those four timeframes, Sayers has the best touchdown rate (8.0), followed by Christiansen (10.0), then Hester (12.8) and then Hall (24.6).

This was the most interesting of everything in the article and is against Hester being the GOAT.

18
by Skeptic (not verified) :: Fri, 10/19/2007 - 5:54pm

17 Brian

Comparing Hester now with others' at their peak is not a good approach. If Hester can only be judged on his first two seasons and the other players get to cherry pick two (consecutive) seasons, they have an extra advantage which biases the comparison.
The better comparison would be the first two "serious" years for each returner. If that happens to be the best two, then my argument is moot, but otherwise using the first two years for everyone is more reassonable.

19
by Jimmy (not verified) :: Fri, 10/19/2007 - 6:01pm

#16

I get that most of the stats are regular season based, but with the stat in question being measured per return and Hester having such a small sample size I just wondered whether in this particular case it might help to include Hester's postseason plays.

20
by seamus (not verified) :: Fri, 10/19/2007 - 6:08pm

I'd have to agree that the returner is somewhat more meaningful than the blocking. You just don't see great return blocking "systems" on franchises the way you see, say, the Broncos consistently great run blocking system.

The other thing to note is that what separates a great return TD career from, say, Bea Arthur's career, is 8 TDs. That's a really tiny number, and it gives a clue why so few players besides Sayers, Hall, Deion, and now Hester have developed reps as game-changing returners. (Deion would have been one of the few men able to keep a roster spot as a pure returner even if he didn't play a little CB.)

21
by Peter Libero (not verified) :: Fri, 10/19/2007 - 7:08pm

13: I was curious about that too. Apparently the answer is no, however. Lenorris Hester was his older brother, who got signed to FIU, showed a lot of potential, then broke his pelvis and may have never played. I can't really find any answers on what happened to him, aside from the fact that he now lives with Devin.

Link to a 2003 player profile about him in my name.

22
by fyo (not verified) :: Fri, 10/19/2007 - 7:27pm

Seeing as I was the one "who was content to point out that [you] were misusing the term 'touchdown rate'", I'll provide an answer as well:

Use the phrase touchdown rate.

And then inverse all your results!

Then everything makes sense. A higher touchdown rate is suddenly good, as it's supposed to be.

If you don't feel comfortable using decimal numbers, feel free to use percentages. That would make Hester's touchdown rate 7.84% in the above.

Simple solution. Easier to understand... and CORRECT.

23
by Tim (not verified) :: Fri, 10/19/2007 - 7:29pm

You should put some measure of standard errors around your numbers. A simple way would be to convert your TD Rate to the fraction of times the player scores, then you could report the binomial confidence interval. It would shed some light on whether or not a particular player has a problem with the sample size.

24
by Jason Cain (not verified) :: Fri, 10/19/2007 - 8:33pm

How does Hester (and the others) compare to Pacman Jones' punt returns?

25
by Pat (not verified) :: Fri, 10/19/2007 - 8:36pm

#22: Simple solution. Easier to understand… and CORRECT.

What's the big deal with using inverse rate? It's just as meaningful ("average returns per touchdown") since it's the same stat anyway, and the numbers involved are more "tractable".

Intuitively, you just don't have a feel for what a "7.8% touchdown rate" is, because when you watch a return, you don't think "OK, he has a 7.8% chance of breaking one here!" One touchdown every ~12-13 returns, though, you have a good feel for, especially since you know there are usually ~7-8 returns a game. So once ever 12-13 returns means a TD every other game on average.

Lot like interception percentage. "3.0% interception percentage" just doesn't have the same intuitive feel as "1 interception every 33 passes".

26
by Jimmy (not verified) :: Fri, 10/19/2007 - 10:30pm

#21

I was watching NFL network yesterday when I saw the Hester to which I am referring to score a receiving TD for FIU (I think). All I know is he is also called Hester and is currently playing in college.

27
by Jerry (not verified) :: Sat, 10/20/2007 - 12:10am

Doug Drinen has a blog post about returners' TDs that's linked on my name.

28
by fyo (not verified) :: Sat, 10/20/2007 - 5:42am

: Intuitively, you just don’t have a feel for what a “7.8% touchdown rate� is

I disagree, quite strongly.

Touchdown rate sounds like something that should be a high number. The higher, the better.

That's the way all the other "rates" in football (and other sports I can think of) work.

You could make the EXACT same argument about completion rate. "Well, you don't think that he has a 65% chance to complete the pass, you think he misses one every 2.9 passes."

Bull.

29
by BillWallace (not verified) :: Sat, 10/20/2007 - 8:05am

One of the things I loved about Brian Mitchell was that he never gave up available return yards. Especially on punts, he pretty much caught the ball and ran within 20 degrees of straight ahead. He was bigger and slower than almost all return guys, and had a uniquely tough style, as far as what I've ever seen.

I don't know the true value of consistency vs boom & bust in the return game, and it would take really intense statistical analysis to even begin to compare Mitchell to a solid boom/bust guy with a higher TD rate. But I know it annoys me to see a guy catch a punt with no defender with 15 yards and end up with a net -3 return.

I'd never put Mitchell ahead of a great returner like Hester or Hall, or in Mitchell's era Deion, but if I could have 10 years of 8 yard punt returns and what I believe was a low fumble rate, I'd take that over gambling on a speed boom/bust guy.

30
by Athelas (not verified) :: Sat, 10/20/2007 - 8:54am

#26-
The college Hester says some people in his family think his grandfather and Devin's may be related, but there is not any known link.

31
by Spielman (not verified) :: Sat, 10/20/2007 - 9:52am

"Intuitively, you just don’t have a feel for what a “7.8% touchdown rate� is, because when you watch a return, you don’t think “OK, he has a 7.8% chance of breaking one here!�"

I would say just the opposite.

32
by Trogdor (not verified) :: Sat, 10/20/2007 - 10:12am

The comparable one I would think of is home run rate, which is most often expressed in AB/HR. I have absolutely no problem with it being returns/TD. Nor would I have a problem with it being TD/return - for a few of the guys on the list. Maybe I have a mental block at 5%, but once you get smaller than that, I'd much rather have it in the inverse. So that leaves four guys for whom it doesn't matter, and everyone else in history for which I'd rather have it as it is.

33
by JoshuaPerry (not verified) :: Sat, 10/20/2007 - 11:38am

Did anyone else feel ripped off abouthis hyped 100 speed rating on Madden, when it was just 99 highlighted? I was.

34
by Josh (not verified) :: Sat, 10/20/2007 - 12:23pm

Call it "touchdown pace."

35
by Bruce (not verified) :: Sat, 10/20/2007 - 1:53pm

How about "TD ratio"? Kind of like HR ratio per time at bat in baseball. The word "ratio" lets most people know its a comparison.

36
by CA (not verified) :: Sat, 10/20/2007 - 2:16pm

Re: Vince: We’ve also learned that if I’m not supposed to use the term “touchdown rate,� then I don’t know what term I should use. Technically, it should be “average kick returns per touchdown,� but that’s awfully unwieldy, wouldn’t you agree?

"Returns per touchdown" would suffice, as would "inverse touchdown rate." Or, as fyo suggested, you could simply take the inverse of the values and the field name would be accurate. Don't dumb down the study out of a stubborn refusal to use the correct terminology; this is FO!

37
by SomeDude (not verified) :: Sat, 10/20/2007 - 2:52pm

I just want to throw in my vote behind #22's suggestion. I don't recall another stat expressed as "X tries per result" in football.

I also find it a very counterintuitive metric. “Average kick returns per touchdown," is as unwieldy a concept as it is a phrase.

38
by Tom (not verified) :: Sat, 10/20/2007 - 3:56pm

Re 37:

Talking heads love to tell us interceptions or touchdowns per pass.

The reason most stats aren't listed that way, is that most are over 1. So it would be like saying .25 rushes per yard. I think people deal with numbers greater than 1 more intuitively, so stats are represented in that matter.

39
by Sammy (not verified) :: Sat, 10/20/2007 - 4:57pm

Where are Ted Ginn Jr. (and his family)?

40
by Blackmallow (not verified) :: Sun, 10/21/2007 - 10:29am

Isn't there an "average expected scoring for every drive started from a specific numbered line on a first down" statistic available?

Like let's say you get the ball first and 10 on your own 20, what is the average expected score of the drive starting at that position.

As a numbered example, 15 occurences happen during one weekend and 3 touchdowns and 4 field goals are eventually scored on drives that start on your own 20. Easy calculation would give ((3 x 6) + (4 x 3)) / 15 = 30 / 15 = 2.00 expected scoring from your own 20.

Now let's say a kick returner took the ball from his own goal line (same as starting from your own 20) and returned it to the 30, where the average expected points is at 2.15 (just throwing a number out there), he would have gained 0.15 points on his return.

Add up every return expected scoring and divide by number of returns (without the no gains from kneeling in the end zone and the fair catches) and you get a rough value of the returner's contribution.

41
by Jeff (not verified) :: Sun, 10/21/2007 - 10:40am

I don't know if anyone has already suggested this, but how about saying "touchdown frequency," or "average touchdown frequency," instead of "touchdown rate." It solves the accuracy problem, and is short.

42
by Jeff (not verified) :: Sun, 10/21/2007 - 12:29pm

Actually, I would like to take back my last post. It is not a touchdown frequency either. Sorry, I shouldn't have posted without coffee.

43
by azibuck (not verified) :: Mon, 10/22/2007 - 10:31am

The South Florida Hester is Jesse Hester Jr. I don't know if he's related to Devin, but he's (boliviously) Jesse's son, and I've never heard Jesse linked to Devin.

44
by C (not verified) :: Mon, 10/22/2007 - 4:55pm

A nice start, but you have to factor in fumbles, return yards and also starting field position on kickoffs, which would take into consideration "kicking away" from Hester or intentionally kicking out of bounds. If you start to consider squib kicks, etc., is gets murkier.

45
by Rich Conley (not verified) :: Mon, 10/22/2007 - 6:42pm

"I’d have to agree that the returner is somewhat more meaningful than the blocking. You just don’t see great return blocking “systems� on franchises the way you see, say, the Broncos consistently great run blocking system. "

There aren't? There seem to be plenty of them. The Ravens have done it the last couple years... Sams goes down... Figurs plugs right in. The patriots are doing it, with everyone from Maroney to Hobbs to Willie Andrews. The jets have done it with Leon Washington and a bunch of other nobodies.

46
by JonL (not verified) :: Mon, 10/22/2007 - 7:10pm

I know it's pretty simple to do myself, but if this article goes through future iterations, it might be useful to see who had the best touchdown rate (or, if you prefer, frequency) in each category as well as combined. It might be interesting to see who was really good at kick returns but terrible at punt returns, for example.