Eight weeks ago, the Vikings were undefeated and ranked No. 1 in DVOA. Now they're 6-6 and probably done. What went wrong?
19 Oct 2007
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:
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 .
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:
(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.
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.
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.
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.
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...
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.
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.
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)
46 comments, Last at 22 Oct 2007, 7:10pm by JonL