The Greatness of Priest Holmes and Economists Teaching People in Sports

As the title indicates, this week’s football column is going to address two unrelated football stories. Before we get to all that, here are the QB Score and RB Score rankings for Week 12:

Week Twelve QB Score Rankings

Overall QB Score Rankings

Overall RB Score Rankings

The Greatness of Priest Holmes

A few days ago Priest Holmes called it a career.  He finished the 2007 season with an RB Score of 1.  Yep, that’s the sum of what he did in his four final games.  So our last impression of Holmes wasn’t too good. 

What story do we tell is we consider his entire career? Across his entire career Holmes posted an RB Score of 4,447 in 106 career games.  This works out to a 42.0 per game mark. 

To put this in perspective, consider the 2007 RB Score rankings after Week 12.  When we look at these rankings, we see that the average mark from Holmes would rank 5th in the NFL this season.  So relative to today’s player, an average Holmes – which we did not see this year – was pretty good.

For even more perspective, let’s see how Holmes compares to a selection of all-time greats.  Holmes only gained 8,172 rushing yards in his career, which is not good enough to break into ESPN’s ranking of the all-time greats in career rushing yards (which only considers the top 25).  Of course, running backs do more than just gain yards rushing. If we consider RB Score – which considers rushing, receiving, and fumbles lost -where Holmes ranks might change.

Before I get to the data, I need to note that the data set I use is from For some reason doesn’t report fumbles lost before 1994 (to see this, here the data from 1993).  So in thinking about the all-time greats, I am going to confine my study to the 16 running backs who played some years after 1994 (so I could at least guest-i-mate fumbles lost before 1994).  Table One reports the ranking of these 16 players in terms of both rushing yards and RB Score per game. 

Table One: Priest Holmes and a Selection of All-Time Greats

Before I discuss the rankings, I want to note that Holmes’ RB Score of 4,447 would rank 11th on this list.  In other words, the career mark of Holmes would best the career RB Score of Edgerrin James, Corey Dillon, Jerome Bettis, Shaun Alexander, and Eddie George.  James and Alexander are still playing, so each can now pass Holmes.  But Holmes did surpass Bettis, a player who currently ranks 5th in career rushing yards.  So if we only look at career totals, Holmes ranks ahead of some fairly big names.

When we analyze sports data, though, we generally don’t focus strictly on totals.  Often we compare players in sports on a per-game or per-play basis.  For running backs I prefer per-game, since I think there is value in putting in more time on the field.

When we look at per-game performance, we see that only four running backs – Barry Sanders, LaDainian Tomlinson, Marshall Faulk, and Tiki Barber – surpass Holmes in RB Score per game. In fact, had Holmes not come back this year, his career mark would have been 43.5.  This would have only been topped by Sanders and Tomlinson.

In sum, although Holmes may never make the Hall-of-Fame (he 0nly gained 1,000 yards in a season four times), he should be remembered as one of the best to play the running back position.

And by the way, a similar story would be told about Tiki Barber and Marshall Faulk.  Both of these backs also rank towards the top in RB Score per game. I sense, though, that at least in the case of Barber, we will see Bettis or Curtis Martin in the Hall of Fame first.  In other words, I think what these players did catching the ball out of the backfield will be discounted by Hall of Fame voters. 

Learning from Economists

And now for a completely different topic (although still about football).

Can people in sports learn from economists?  One of the more outlandish arguments (or perhaps THE outlandish argument) offered in The Wages of Wins is that three economists have something useful to say about sports.

Although it’s possible (albeit extremely – and I mean extremely – unlikely) the team of Berri, Schmidt, and Brook have nothing useful to say, it turns out one economist has shown that what we attempted could indeed be done.  Yes, we have evidence that an economist has actually impacted decision-making in sports.

The following is from Gregg Easterbrook of

“There’s a high school in Arkansas that has made the most significant football innovation we’ve seen since the veer option. This high school is tearing up its state and is on the verge of revolutionizing the way football is played. TMQ suspects that within a few years, the phrase “Pulaski theory” will be as widely known as the phrase “shotgun spread.” In a copycat sport, Pulaski Academy of Little Rock has devised an offensive philosophy that is genuinely new, and it’s winning games left and right.

Pulaski Academy does not punt.

I first heard about Pulaski from Peter Giovannini of Morrilton, Ark., a high school football official who wrote me to report in astonishment that he had just worked a conference championship game in which the winning team never punted, even going for a first down on fourth-and-6 from its own 5-yard line early in the game. “As a devotee of TMQ, I thought you might like to know at least one coach in the vast football universe has experienced the epiphany and refuses to punt the ball away,” Giovannini wrote.

That team was Pulaski — 9-1-1 after having just won its opening-round game in the Arkansas 5A playoffs. Coach Kevin Kelley reports that he stopped punting in 2005 — after reading an academic study on the statistical consequences of going for the first down versus handing possession to the other team…”

The academic study is by David Romer.  In Do Firms Maximize? Evidence from Professional Football (published in The Journal of Political Economy), Romer offers evidence that NFL coaches are too conservative on fourth down. Too often teams will punt, when the expected benefits of going for it exceeds the expected costs.  Although I don’t know if NFL coaches have changed their behavior in response to Romer (and if you do know, please let me know), it’s interesting that a high school coach is using this research and finding success. 

Hat Tip to Steven Levitt

– DJ

For more on QB Score, RB Score and what these metrics mean see

The New QB Score

Consistent Inconsistency in Football

Football Outsiders and QB Score

The Value of Player Statistics in the NFL

Comments are closed.