February 18, 2009
FOOTBALL: The Greatest Running Backs of All Time
The news that the Jaguars were releasing Fred Taylor started a discussion with some friends about where the oft-injured Taylor would rank among the great running backs, and where he might have ranked if he'd been healthy. As I often do with such discussions, I thought I'd take the broader context to evaluate how the numbers stack up for the all-time great running backs. First I'll offer the data, then a handful of my own thoughts on it.
Now, when you look at baseball statistics, it's critically important to do three things. The first is to understand context. A baseball player's statistics are influenced by many external factors - changes in the game over time, ballpark effects, the influence of teammates, the length of the season. For a variety of reasons, it's much harder to separate football statistics from the context of time and team. The numbers I'm setting out below are not adjusted for the changing offensive contexts these running backs played in, whether they worked in sophisticated passing games (Roger Craig, Marshall Faulk, Edgerrin James), behind great offensive lines (John Riggins) or great blocking backs (Emmitt Smith), or indoors (Barry Sanders) or wartime (Steve Van Buren), or whether they were just given the ball and told to hope the other 10 guys on the field might run into a defender now and then (Sanders, Eric Dickerson). I have, however, averaged their statistics per 16 scheduled games, as a way of evening out the old-timers who played 12- or 14-game seasons or the more recent players who played during strikes (I counted 1987 as a 12-game season except for Craig, who crossed the picket line). Although the job of the running back has changed less than the jobs of other offensive players over the past seven decades, the latter-day backs, reflecting the higher-octane offenses of modern football, are nonetheless overrepresented. I could have run averages per game played, but I preferred to let the numbers reflect the costs to their teams of injuries to guys who missed a lot of games (like Taylor) or whole seasons (Riggins, Garrison Hearst). Durability matters when you are building a football team. The "G/16" column on the chart shows how many games, on average, each of these guys played per scheduled 16 games over his prime seasons, including entire missed seasons (yes, Walter Payton, Jim Brown, Eddie George and Roger Craig never missed a regular season game in their primes).
Second, and relatedly, you have to figure out what portion of a player's career you are evaluating. My own preference, in having these kinds of debates, is neither to zero in on a player's single best season nor to just lump together career totals (since they may include one guy who hung it up in his prime compared to another guy who was just as good and stuck around a bunch more years as a part-timer - a decision that really has nothing to do with how good they were in their primes). So, I'll set out here the per-16-scheduled-games averages not for these running backs' whole careers but for that section of their careers you would identify as their primes, in general the seasons when they were a team's #1 back. In doing so, I've set aside the years after these guys broke down (in most cases, running backs break down pretty quickly and dramatically, around age 29 for the usual back, 31-33 for the longer-lasting ones) and the lengthy second acts of backs like Marcus Allen and Ottis Anderson. The resulting focus rewards the guys who concentrated their best seasons all together. Consistency matters when you are building a football team.
Third, much as I've done for similar baseball columns, I'm breaking the numbers into two charts, one of guys whose primes were longer (7 years or more), one of the guys who were only on top of their games for 4-6 years. It's apples and oranges to compare a per-seaon average of a guy who starred for 5 seasons to one who starred for 10. Longetivity matters when you are building a football team.
More specific to football, once those preliminaries are out of the way, is what numbers to use to rank the running backs. I don't pretend to have a perfect answer to that, but I chose to organize the data around the most basic figure: rushing yards per team game, listed as Yds(R). I could have used combined rushing and receiving yards from scrimmage (YSCM), and you can re-sort the list if you like by that, as in some sense it's a truer picture of a back's total offensive contribution. But while catching the ball well is a useful skill, it's also true that receiving numbers are much more influenced by the team and the era a guy played in, whereas running the football is the purer, man-with-ball-versus-eleven-defenders task that every running back has faced down through the game's history.
The sample here is the 49 running backs who compiled at least 7,000 career rushing yards through the 2008 season, plus five other notables who popped up on various career leader lists (Van Buren, Marion Motley, Larry Johnson, Billy Sims, and Gale Sayers). I'm hoping the charts here will be more an argument-starter than an argument-finisher, as of course I haven't even touched here on playoff performances or other factors beyond the raw, regular season numbers. But we can at least appreciate those numbers for what they can tell us about the yards these men traveled with a football in their hands.
RRTD=Combined Rushing/Receiving TDs. (No stats are included here on kick/punt returning, and no commonly used stats exist on blocking). Fmb=Fumbles.
To save you the time of adding it up, the top 10 ranked by yards from scrimmage rather than yards rushing would be Brown, Tomlinson, Barber, Dickerson, Payton, Faulk, Sanders, Martin, Allen and James. No, I didn't expect Tiki to rank that high, either. Next, the shorter primes:
|8||Steve Van Buren||5||1945-49||25-29||15.16||264||1252||15||4.74||11||129||1381||16||8|
Tomlinson may be getting close, but when you adjust for the length of the schedule, I have to come away convinced that there are three guys who really were the class of the all-time field - Jim Brown, Barry Sanders, and Walter Payton - with the next tier consisting of Tomlinson, Erick Dickerson (who had a shorter career, fumbled a bit much and wasn't much of a receiver) and Emmitt Smith (who is a bit down the chart but had the most seasons as a productive workhorse back). I'd probably rank them thus:
Brown is just the guy you can't get around; he was so good and so far ahead of his contemporaries, he tops the list regardless of whether you include receiving yards, he trails only Tomlinson and Shaun Alexander in TDs, and the only two guys with higher yards per carry figures (Motley and Joe Perry) carried the ball a lot less.
Sanders was always a favorite of mine, and aside from Lawrence Taylor (the one and only LT), he's the most spectacular performer I have ever seen on a football field, a guy who could just leave your jaw hanging open. The knock on him - that he wasn't the best goal-line/short-yards back compared to Emmitt Smith - was a little unfair given how little help he often had on his teams, but it's one small strike, and when you consider the overall package of Payton's fantastic duarbility, value as a receiver, and the fact that he won a championship, I'd give him a slight edge over Sanders.
(As an aside, when you look at the ages at which these guys mainly burned out, my guess is that when both Brown and Sanders hung up their spikes, they probably only had another year, maybe two, before the miles caught up with them - and of course, Sanders had played his whole career under the shadow of a high school knee injury that somehow never recurred despite his doing things on nearly every play that would blow out a normal man's knees).
After those six guys, it gets stickier. Once you get past the very top tier, OJ Simpson quickly comes into the picture. But then, there's a reason why the Juice got to be such a big deal in the first place; the man could play some football, and even he can't take that away from himself. I won't even try to untangle the rest of the top ten - Terrell Davis has to be in there somewhere, but his prime was so short, and we are still too close to a lot of the other guys near the top to have perspective on their careers. I will say that while his numbers even here are depressed by including an injury-plagued season before his last 1,000 yard rushing year, Earl Campbell was just an amazing, amazing guy to watch in terms of pure power and speed; the best I have ever seen at just dragging multiple defenders along for those last few yards. (Sadly, if you have seen Campbell in recent years, he looks and moves as if he aged a year for every one of those yards).
And Fred Taylor, who got us into this discussion? Well, as you can see here, Taylor holds his own pretty well given how much time he missed; he's had a heck of a career and put up some big numbers per game actually played. But there's no substitute for being on the field.
Great post, as always. I understand your point about isolating context being so much more difficult in football than in baseball, but your reference at the end to Terrell Davis forced me to comment. I used to LOVE Terrell Davis - I thought that he was amazing. Then he got hurt and Olandis Gary came in and he was great, too. Then Gary got hurt and Mike Anderson stepped in and HE was amazing. Two years later Clinton Portis came in to the same role and of course he was amazing too.
So, either the Broncos were lucky enough to have a new great RB pretty much every year from 1997 - 2002... or maybe the individual RBs weren't so great after all and it was the system.
Regarding Jim Brown and putting things in context-a lot of his numbers were due to his size and speed relative to the size and speed of his contemporaries. If he played today he would still be a great player, but his numbers would be down. I think if you put Sanders or Smith or a bunch of these players back in the time that Brown played there numbers would be as good if not better than his. He never faced defensive players with the size, strength and athleticism that modern players have. In the late 50s and 60s defensive linemane were like 230-240-you have safeties now that weigh that much and run like the wind. No steriod issues, etc. Its sort of like the Wilt Chamberlain thing-he would still be a great player, but so much of his numbers was due tohim being a relative physical anomaly at the time he played.
It's true that Brown was a physical anomaly but you also shouldn't hold that against him. We don't know how Smith or Sanders would have done, we can only speculate. But we do know what Brown did. It happened.
Fred Taylor reminds me of Barry Larkin (although I think Larkin was a better SS than Taylor was a RB): Great numbers, but the knock on him is that he was injury-prone. But the fact that he was injury-prone actually underscores how truly great his numbers actually are: they were compiled in less games/ABs. Just imagine what the numbers would be if he wasn't injury-prone!
Look, the guy only made the Pro Bowl once. He wasn't one of the greatest RBs of his era, so can't be considered one of the greatest RBs of all time.
I appreciate the comparison that Crank did, but there is too much difference between eras in order for the comparison to be really useful -- no different than baseball. The question, to my mind for all sports, is whether the guy is dominant for his era, rather than how well his stats compare to guys from different eras. That's why Jim Rice is a clear HOFer (6 time top-5 MVP vote) and Jim Edmonds isn't.
I think it is difficult and almost useless to try and apply baseball sabermetrics to football. They are fundamentally different entities. Baseball IS a stats and numbers game. Football just is not, it is more results oriented. There would be very, very few hardcore football fans that could tell you the significant historical numbers of football. Most career touchdown passes for instance. It just happened to the QB for a NY based team (and one of the most famous football players of all time) and I'd bet less than 5% of football fans know what the number is. Name hardcore baseball fans who don't know the significant numbers. Football doesn't play by the numbers the way baseball does and while this is an admirable effort any list that even insinuates Tiki as one of the all-time great backs is a testimony to how stats lie.
I don't think you can safely argue away Brown's physical superiority relative to the competition, any more than you can Ruth's being so much larger, or Gehrig, or Wilt. After all being bigger, faster and stronger is supposedly what he Olympics chases. Add intelligence, and you have a table (instead of a tripod); a four legged need for a great athlete.
It's harder to evaluate a football player than any other sport, since they are so much more dependent upon coaching and teamwork than any other. By a large margin too. Which is why the greatest on any football team may or may not make the team a playoff winner.
That said, I do agree that with Brown, Sanders and Payton as the three greatest running backs. We can always argue about order, but given that anyone else COULD have been as big and strong as Brown relative to the league, but weren't, it's fair to make him number 1. As a running back. As a complete player, I would go with Payton as the greatest PLAYER ever. Maybe the third best pure running back, but the greatest blocking running back ever (Matt Snell and Tony Nathan were good too, but Sir Walter was in a class by himself); and maybe the only running back who was also a backup quarterback, kicker and punter, he would have been the greatest corner or safety ever, and who else would you rather have your daughter bring home?
The Walter Payton mention is a good example as to why football players are hard to categorize. Payton doesn't get any points or stat totals for leveling on-coming blitzers so that Jimmy Mac could throw a TD pass. You had to watch the game and know what was going on to even know it happened. Basketball has the same issues. From a pure stats point of view it makes no sense that guys like Bruce Bowen or Shane Battier are even in the league let alone starting for very good teams. Their stats are horrible but because there are no true defensive stats (blocks, steals, defensive rebounds are it and all of them are as effective at measuring defensive capability as RBIs are at measuring plate productivity) for these guys to be judged by.
Hell, Bill Russell is considered as one of the greatest basketball players of all time. Other than rebounding how do we know? They didn't record blocks or steals back then and his impact on the game as a defensive presence is anecdotal.
jim: John Hollinger made the exact same points about defense in basektball in a podcast with Bill Simmons yesterday. I should note that in addition to rebounds, blocks, and steals, we also have personal fouls to measure defensive players (although that stat is a negative one).
jim, there is a story Aurbach told about Russell. He and the Celtics were touring Europe in the 50s, trying to tout these pros as better than the famous collegians then. So when he played against the Czech team, the Czech coach bragged how they beat the better American College players, so these Celtics had no chance. And Auerbach told Russell not the let their start center, whomever he was, not to score a single point. The punchline is that of course the guy didn't.
Do I know if it's true? No, but jim is right. Stats are all that are left sometimes. Yes yes, Michael Jordan is the greatest, and I assume he is kind of getting into the Hall of Fame this years. But when all you have are numbers, you do get Russell more of an also ran than the greatest ever. And that Jerry West could not have been as good (I still say better) than Oscar Robertson. Same with Payton. If I had to choose who would start for me on offense and defense, I would pick Sir Walter on offense, John Elway (and I saw Unitas) second; on defense, Deacon Jones first, LT second.
As an aside, and here is a sign of the times. Brad van Pelt died today of a heart attack, and I wish nothing but condolences for his family. But my first thought on this 57 year old super athlete dying and I wondered if he took steroids. Damn Balco and everyone.
That's sad news about Van Pelt. My grammar school used to have an annual "Sports Night" where a local athlete would come speak to us and do autographs. We had Phil Simms, Ken O'Brien, Lou Piniella, Joe Morris, Bobby Meacham (the quality varied from year to year)...Van Pelt was the first one I remember.
Don't even get me started on the "assist" stat in basketball. At least a rebound is a rebound and a steal a steal. What the hell constitutes an assist. Look at freaking hockey and the assist. Pass it to a guy (even if you're the goalie), he goes the length of the ice and scores. Assist to you. Baseball works nicely for stats (sometimes too nicely and people get overly caught up in tying stats to the overall quality or comparative values of individual players) and you can get a pretty good assessment of a player from what the numbers say. Football? Hard to do? Basketball? Hard to do. I watched Curtis Martin for years with the Pats. Great back. Really. When it mattered? More A-Rod then Manny. Not one of the great backs in history other than for sheer durability and decent year in, year out production.