Writers of Pro Football Prospectus 2008

29 Feb 2012

ESPN: Speed Score 2012

Here's a look at Speed Score for the 2012 class of running backs. This year, Speed Score actually falls pretty close in line with conventional wisdom, except for a couple of sleeper power backs with injury questions.

Posted by: Aaron Schatz on 29 Feb 2012

18 comments, Last at 06 Mar 2012, 1:08am by Dan


by Karl Cuba :: Wed, 02/29/2012 - 2:36pm

If you use Nolan Nawrocki's estimates for Richardson's forty time and weight then you get an estimated speed score of 109.7, which would rank second.

Also, RGIII would have a higher speed score than any runner with 117.9 if you evaluated him as a running back, while Steven Hill would score 119.0.

by sanwekwe :: Wed, 02/29/2012 - 3:19pm

Check my math, but workout wonder Poe would clock in at 123.0 based on a weight of 346 and a 40 time of 4.87.

by Karl Cuba :: Wed, 02/29/2012 - 3:28pm

I'd love to see him at tailback.

by Podge (not verified) :: Fri, 03/02/2012 - 4:57am

I got him to about that as well.

I'm curious who would have the best speed score ever regardless of position. I know Mario Fannin holds the record for RBs by running a 4.3 at around 250 and a score of 125 or so, but I wonder if someone else would have it if you extended out to different positions. Raiders OL Bruce Campbell maybe? Didn't he run a really quick 40?

by Aaron Brooks Go... :: Fri, 03/02/2012 - 9:36am

Herschell Walker and Bo Jackson are both higher than that.

by Dan :: Tue, 03/06/2012 - 1:08am

I'm seeing Poe listed at 4.98. Based on these combine numbers, this year's all-position speed score leaders are:

TE James Hanna 124.0 (252, 4.49)
ILB Mychal Kendricks 119.7 (239, 4.47)
DE Bruce Irvin 119.5 (245, 4.50)
OLB Zach Brown 119.0 (244, 4.50)
WR Stephen Hill 119.0 (215, 4.36)
QB Robert Griffin 117.9 (223, 4.41)
DE Nick Perry 116.9 (271, 4.64)
WR Tommy Streeter 116.9 (219, 4.40)

by Misfit74 :: Wed, 02/29/2012 - 5:34pm

More 'Speed Score' stuff, for free, posted yesterday here:


by Aaron Brooks Go... :: Wed, 02/29/2012 - 6:45pm

It's interesting how many DVOA darlings had terrible SpeedScores. Westbrook, Sproles, and McCoy were in the low-90s. Emmitt Smith was a 96.

by tuluse :: Wed, 02/29/2012 - 7:04pm

Westbrook was injured when he ran his 40, so that one isn't exactly fair.

by sjt (not verified) :: Wed, 02/29/2012 - 9:58pm

Sproles's value has come mainly from his receiving, not his running between the tackles. Until this year he didn't have a good rushing DVOA, and he's never even qualified for the large sample size group in that measurement. So it makes sense that speed score wouldn't be as meaningful for him.

by Aaron Brooks Go... :: Fri, 03/02/2012 - 9:39am

Sproles's value has come mainly from his receiving, not his running between the tackles.

Welcome to DVOA!

by Mr Shush :: Thu, 03/01/2012 - 6:51am

Smith is so long ago that I'd be wary of applying a contemporary metric to him. The others are all undersized, elusive players with outstanding lateral agility and excellent receiving skills. I think Speed Score does a much better job with more conventional players, and doesn't speak to their value as receivers at all. It's a way to help identify guys like Ben Tate, not guys like Danny Woodhead.

by jimbohead :: Thu, 03/01/2012 - 3:23pm


Combine 40-yd dash numbers have improved dramatically in the last 10 years due to better prospect preparation. You really can't compare E. Smith's numbers to today's numbers without applying some sort of adjustment to Smith's 40 time.

by Aaron Brooks Go... :: Fri, 03/02/2012 - 9:53am

Yes, except...

Smith's Speed Score was abominable even by comparison to stellar backs of Smith's era. But even the great backs of Smith's era tended to have high peak short length careers that weren't as fully successful as more mundane-speedscoring backs such as Sanders or Smith. But even in Sanders' case, his 110 score is misleading in that his success was not built on straight speed but was built on premier acceleration and lateral agility; often the easiest way to tackle him was to catch from from behind.

I'd argue also that Smith is the near definition of a conventional player, who still excelled despite a terrible SpeedScore.

by Mr Shush :: Fri, 03/02/2012 - 12:05pm

It's not just about the need for straightforward era-adjustment, though. I forget what period was originally used for the regressions that gave us speed score, but it's not clear to me that we should automatically expect the requirements for success in past eras to be the same as in more recent ones.

On the other hand, "no model's perfect nor does this one claim to be, and Smith is just an outlier" is a perfectly plausible line too.

by Rabbit :: Thu, 03/01/2012 - 9:45am

Interesting = unpredictive?

Is 40 yd speed even the most important attribute for a RB? I'd argue that 10 yd and lateral speed (shuttle) are at least as important. Probably more so.

by tuluse :: Thu, 03/01/2012 - 11:54am

FO has run regression tests before, 40 yard dash was the most predictive combine result.

by draftee :: Sun, 03/04/2012 - 2:03am

Seminole LT Walter Jones weighed 301 lbs (http://www.sportsxchange.com/DS97/indy/TOPWT.htm) and ran a 4.65-40 in '97 combine. At weight x 200 / 40-time ^4 = 128.76. Almost wish the Seahawks would have let him run the fumblerooskie a few times.