SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com » Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisReggie Bush and Maurice Jones-Drew, Part I
[Note: This can also be seen over at the Footballguys.com blog:]
Reggie Bush and Maurice Jones-Drew are two of the most exciting and talented young players in the league. Both have the necessary skills to earn annual trips to Honolulu. Bush, of course, was a Heisman Trophy winner and top three pick in the NFL draft … just like O.J. Simpson, Tony Dorsett, Earl Campbell, Bo Jackson and Barry Sanders. Jones-Drew averaged over 5.5 YPC and rushed for 12+ TDs last year, joining an elite club consisting of only Drew, Jim Brown, O.J Simpson, Eric Dickerson and Clinton Portis. And only Portis did that as a rookie.
It is not controversial to state that Bush was viewed as the better prospect and that MJD had a better rookie season. The key question now, is which one trumps the other? Forgetting the names for a minute, which RB would you expect to have the better career: the highly touted prospect or the rookie stud? Reuben Mayes rushed for 1353 yards at 4.7 yards per carry as a rookie, but had just 2,131 more rushing yards the rest of his career. Additionally, John Stephens, Karim Abdul-Jabbar, Greg Bell and Terry Miller all had 1,000 yard rookie seasons, but fizzled out quickly. On the other hand, we know that high draft position doesn’t mean everything, either. Ki-Jana Carter, Brent Fullwood, Blair Thomas, Alonzo Highsmith, and, well, Terry Miller, were top five draft picks that recorded fewer than 2,500 rushing yards in their careers.
I looked at all RBs drafted from 1978 to 1997. That gives us 20 years worth of drafts, with little worry about active players. There are a few guys still remaining, but Tiki Barber and Corey Dillon just retired from that ‘97 draft class, and I doubt we’ll be seeing significant changes to career totals to cause concern. Here’s a pretty intuitive chart:
Draft Pick #RBs Career Rush Yards
1 - 3 13 7132
4 - 10 15 4663
11 - 20 29 3642
21 - 40 54 2930
41 - 70 57 2135
71 -100 52 1629
101-150 70 1268
150+ 154 1032
There were 13 RBs drafted among the top three over this 20 year era, and they each averaged 7,132 career rushing yards. There’s a very strong correlation between draft pick and career rushing yards, which isn’t terribly surprising. Based off this, you might have projected last year that Reggie Bush (#2 overall) would outrush Jones-Drew (#60 overall) for his career by about 5,000 yards.
But now it’s this year, and we have some more information at our disposal. Jones-Drew rushed for 941 yards last year, while Reggie Bush rushed for only 565 yards. What does history tell us to think about those data? The table below breaks down all rookie RBs into tiers based on their number of rushing yards during their rookie season. The last column shows the average rest of career rushing yards (AROCRY) for those players:
Rushing Yards #RBs AROCRY
1400+ 8 9702
1100-1399 11 5053
900-1099 13 4161
600-899 31 3109
400-599 45 2527
200-399 80 1590
01-199 206 960
0 50 716
Like the draft pick chart, there are no surprises here. The more yards you rush for as a rookie, the more yards you’d expect a player to rush for during the remainder of his career. Based off this table, you might project that MJD will rush for about 1600 more yards than Bush for the rest of their respective careers.
But we DO know something else about Bush and Jones-Drew, namely their draft position. We’re getting closer to answering the question of which RB should be expected to rush for more yards: the highly drafted player or the better rookie. We can use multiple regression analysis to tell us how these two variables play off each other. One of the problems with using regression analysis, however, is that it treats the difference between the 1st pick and the 20th pick the same as it would the difference between the 201st and 220th pick. The solution? The NFL pick value chart. So now the 1st pick is worth 3000 points, the 2nd 2600 points, the 10th 1300 points, the 116th pick 62 points, etc. So if we use rookie rushing yards and NFL draft value as our two variables to solve for remaining career rushing yards, here’s the formula we get:
Remaining Career rushing yards =
439 + 0.88 * (Pick Value) + 4.49 * (rookie rushing yards)
So let’s say Adrian Peterson (the 7th pick, pick value 1500) rushes for 1,000 yards this year. This formula would project him to rush for 6,249 rushing yards for the rest of his career. If he rushed for 1,500 yards, we’d project him at 8,494 remaining career rushing yards. If the first pick in the draft rushed for 1,206 yards as a rookie, we’d project him for the same 8,494 remaining career rushing yards. An undrafted rookie would need to rush for 1794 yards to be projected for the same remaining career rushing yards. Those numbers “feel” about right to me, so I think our formula will work.
How do Reggie Bush and Maurice Jones-Drew fare? Bush was the 2nd pick (2600 points) and rushed for 565 yards, so that projects him out at 5,264 rushing yards for the rest of his career. Jones-Drew was the 60th pick (300 points) and rushed for 941 yards, projecting him at 4,928 yards. So at least for now, it looks like Reggie Bush holds a slight edge.
In part II of this series, we’ll compare the two players using different variables than rushing yards. In part III, we’ll take a step back and reign in our optimism just a bit.
This entry was posted on Monday, August 13th, 2007 at 8:00 am and is filed under Fantasy, History, NFL Draft, Statgeekery. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
Pretty cool idea. Not that it really matters, but what is the r-squared and what are the significance values for pick value and rushing yards? I'm curious what matters more: pick value or rookie performance? Did you try yds/carry instead of total season rushing yards?
Hey Brian,
I don't have the data handy anymore (I'm on vacation), but the significant values were extremely high. That's not surprising, though, because of the extraordinarily large sample size. When you have between 500-1000 RBs, you're going to have lots of results that are significant in a statistical sense. But I don't have the numbers handy, sorry.
Try not to steal the thunder on tomorrow's post.
But yes, YPC will be included.
You aren't taking into account that MJD is an umpa-lumpa. Also, RB's are all different. Total yards from scrimmage might be a better gauge. Reggie will probably be catching more balls in the flats than anyone else of his generation. Just a guess.
That puts Joseph Addai at 5838 career yards beating both of them. Sounds about right.
I know you have to draw the line somewhere, but the problem with extrapolating from the pick value chart is that the running back position is not valued the same now as it was in the 1980's. There were 28 RB's drafted in the top 15 for 1978-1987, 16 for 1988-1997, and 14 for 1998-2007.
If Shaun Alexander, Deuce McAllister, Larry Johnson, or Steven Jackson got in a time machine and went back to 1980, they would have been drafted at a higher relative position than they were. This does not have as much affect on projecting outside the first round, because the value chart increases are much smaller.
I looked at what this regression would have projected for first rounders taken 1998-2004. Of the 21, 10 have already exceeded their projection, 5 will not, and 6 are to be determined. Of the TBD, James is a virtual lock (less than 200 yards away), Lewis will get there barring a catastrophic injury this year, Steven Jackson should pass it this season if he stays healthy, Chris Perry's projection is so low he would reach it if he could just play as a 2nd rb for two healthy good seasons, Michael Bennett will reach it only if he gets to start one more season somewhere, and Kevin Jones is unlikely to get 4700 more yards. I'd put the over/under at 3.5 exceeding the number out of those 6.
In looking at these guys, I think there are a couple of things that might lead this regression to underpredict the majority of current first rounders. First, advances in medical treatment--guys like James and Lewis had ACL injuries early in their careers and returned to have several productive seasons. Second, there seems to be more of a tendency for teams to have drafted running backs a year before they needed them, and work them into a starting role in years two and three--Alexander, McAllister, LJ, Steven Jackson, McGahee are all examples of this recently. This could lead to underprojection of guys like Maroney or Bush in this system.
Also, did you consider adding age as a factor. Joe B mentions Addai, but he was 23 last year, while Bush, Maroney and Jones-Drew were 21. Ronnie Brown, the other 2nd overall pick at RB recently, was 24. I would be more comfortable taking Bush over Addai in career yardage for rest of career.
Great post, JKL. Lots of stuff to think about there. Unfortunately, I don't see how you can look at drafts any later than 1997.
Fascinating!
I have a random tech nerd question though. What do you use to do multiple regressions? I hope it doesn't start with an M and end with an A...
If, like JKL says, RB's were taken earlier from '78-'87, and you seemingly have a very large sample size, why don't you run the data just using the 10 year range of '88-'97? A similar amount of backs were chosen in the first 15 picks during that range as in the last 10 years, so it might be more applicable.
This doesn't have to be complicated. How about instead of trying to justify a player doing better in his career over another, we totally ignore his draft position and look at performance? That would seem, uh... Logical.
As of this post, Jones-Drew has rushed for 400 more yards than Reggie Bush on eight LESS touches. That's insane how efficient Drew is compared to Bush, at least so far. Eight less touches, 400 more yards. Factor in receiving and everything else, and Jones-Drew has nearly 100 more total yards from scrimmage than bush does on 82*** less touches. Again, that's insane efficiency by comparison.
I understand trying to give people an edge for their draft position, since generally they have more upside... but let's be real here. The only thing that matters in this case is performance, and Drew is everything everyone wanted Bush to become and then some. Drew is all of that, and one of the most physical RBs playing today. You wouldn't believe it by hearing his measurements, but he is.