Writers of Pro Football Prospectus 2008

28 Feb 2011

A New Speed Score Record!

Although I was one of the few FO folks who didn't make it to the Combine this year, I'm happy to report on a bit of news: We have a new record Speed Score.

Auburn running back Mario Fannin ran a 40-yard dash that's been timed at 4.31 seconds; he weighed in at 231 pounds, producing a Speed Score of ... 125.5! That narrowly beats out the previous record score of 123.5, set by Brandon Jacobs during the 2005 Combine. It's the highest score since Chris Johnson put up a 121.9 before the 2008 draft.

As I mention every year when Speed Score comes up, just because Fannin has a high Speed Score does not indicate that Fannin is guaranteed to be an excellent NFL running back. He had fumble issues last year and only touched the ball 330 times in four years at school. 18 of his 61 carries this year came against Louisiana-Monroe and Chattanooga. Chances are that he's an athlete without the ability to produce at the NFL level.

But would I pick Fannin with a sixth- or seventh-round pick? Absolutely.

Wonko from the Bolts from the Blue blog has Speed Scores for every other NFL running back.

Posted by: Bill Barnwell on 28 Feb 2011

31 comments, Last at 03 Mar 2011, 6:39pm by Nick W

Comments

1
by andrew :: Mon, 02/28/2011 - 9:26am

How can Al Davis get back into the first round?

2
by Mattigus :: Mon, 02/28/2011 - 10:07am

Al Davis Joke

3
by Drunkmonkey :: Mon, 02/28/2011 - 11:12am

So when I read last night about how Fannin had this amazing 40 time, and that he was only a 3rd stringer at Auburn, and then that he was a great receiver, is it bad that my immediate thought was "What round is he supposed to go in, so Al can select him 3-4 rounds too high?"

Or is that what everybody else would eventually decide too?

4
by Dean :: Mon, 02/28/2011 - 11:30am

So if he's 231 and can run like that and supposedly had great hands, maybe he can move to TE?

5
by krugerindustria... :: Mon, 02/28/2011 - 11:47am

i'd be awfully suspicious of any guy who put on 30 lbs of lean weight in 6 months (or whenever they play for real again).

UPDATE: just to qualify that statement, not trying to be a smartazz. The few TE's that I checked were at least 255lbs, plus Fanin is listed at only 5'11". So i would stand by the "not a TE" off the cuff evaluation.

7
by Dean :: Mon, 02/28/2011 - 12:57pm

I don't think of 230 being undersized for a pass receiving TE. He wouldn't be a GURT, but as a pass receiving TE, 230 isn't that undersized. On the other hand, yes, I missed the 5'11" part. That would be difficult.

And yes, if he suddenly showed up at camp at 255, it'd be fair to question what sort of vitimins he took to achieve that result.

8
by krugerindustria... :: Mon, 02/28/2011 - 1:07pm

I didn't think 230 was that small either, but even guys like Dallas Clark and Chris Cooley are "listed" at around the 255 mark.

Your point is well taken though, for a guy with hands and a great speed score, what kind of role can he fill (assuming he can hang onto the ball)?

24
by Aaron Brooks' Good Twin (not verified) :: Tue, 03/01/2011 - 10:23am

Hines Ward-type receiver or a Bush-style hybrid? A fast, 230lb fireplug could beat the crap out of a lot of DBs.

16
by ChicagoRaider :: Mon, 02/28/2011 - 5:31pm

Hmm. How is he blocking? Maybe he has the skill set to be like Marcel Reese. He sure would be hell to cover going from fullback to splitting out.

6
by BlueStarDude :: Mon, 02/28/2011 - 12:41pm

Fumble!

No, I wouldn't even use a 7th round pick on him.

9
by Mr Shush :: Mon, 02/28/2011 - 2:09pm

Fumbling problems are at least sometimes significantly correctable - look at late-career Tiki Barber. This guy strikes me as a very reasonable candidate for a late-round gamble. If nothing else, he ought to be a good gunner.

10
by Anon (not verified) :: Mon, 02/28/2011 - 2:28pm

What is a speed score, and what evidence is there that it is a good predictor of NFL performance?

11
by Vince Verhei :: Mon, 02/28/2011 - 3:15pm

See here.

12
by Anon (not verified) :: Mon, 02/28/2011 - 3:46pm

Thank you. Unfortunately the link does not describe exactly how a player's weight is combined with his 40 yard dash time.

More importantly, it gives no reason to believe that a speed score is important. The old speed score record was set by Brandon Jacobs. Now, Brandon Jacobs is an NFL-quality back, but I suspect that he is not in the upper class of NFL running backs. So what exactly is a speed score supposed to tell us?

There have been many prospects who have had impressive combine results but have not had equally impressive NFL careers.

13
by bravehoptoad :: Mon, 02/28/2011 - 3:59pm

Try this: here.

22
by Vince Verhei :: Tue, 03/01/2011 - 12:23am

That's a much better link. Thanks for posting it.

17
by Whatev :: Mon, 02/28/2011 - 6:00pm

Thing is, the reason Brandon Jacobs is not in the upper class of NFL running backs has much to do with his durability issues. That's something that speed score doesn't measure.

21
by Shattenjager :: Mon, 02/28/2011 - 8:35pm

The original article introducing Speed Score, "Five Seconds Can Be a Lifetime" by Bill Barnwell in Pro Football Prospectus 2008, says, "Although Speed Score isn't a quick-and-dirty measure of the viability of NFL running backs, it provides an insight into that viability that surpasses any other metric or measurable on draft day."

It's supposed to be another tool in the bag for predicting which players will have successful NFL careers.

14
by Danish Denver-Fan :: Mon, 02/28/2011 - 4:49pm

To what extend is ballsecurity a skill? I mean a relatively rare occurance such as a fumble or an interception is vulnerable to variance, right.

Also how big of a weakness is it really? I mean the difference between a starting back with good security and one with bad security is, what, two fumbles a year? Three?

Has anyone crunched the numbers?

18
by Danish Denver-Fan :: Mon, 02/28/2011 - 6:17pm

I know it's poor form to answer your own question, but I was able to answer the latter question.

I searched the PFR database for all runningback season with 200 or more attempts going back to 1990 to keep things in the same era. This gave 454 runningback seasons. I calculated each back's fumble percentage, and calculated the average percentage (0.0138), and the standard deviation(0.00813).

A runningback holding on to the ball one standard deviation worse than average, would fumble about 2.2 percent of his carries. A back holding on to the ball one standard deviation better than average would fubmle about .5 percent of his carries.

I prorated those figures to 300 carries (your average starter/featureback) and got a difference of 4.91 fumbles per season.

Now is that enough to totally change you oppinion of a runningback? I don't think so. Remember about half of those is recovered by the offense. So about 2.5 turnovers per season. It should definitely factor into the evaluation, but I don't think it's enough to turn a 2nd round pick into a 6th rounder.

And thats even without discussing to which extend ball security is a real skill or pure variance.

19
by tuluse :: Mon, 02/28/2011 - 7:27pm

Well in this case, the player is a marginal prospect already, so even a minor issue becomes a bigger problem. Look at mendenhall for an elite prospect with poor ball security.

Also, if a rb is getting 200 or more carries he probably already has above average ball security, so your math is incorrect.

Finally, for proof that it's a skill just look at tiki barber before and after caughlin arrived.

23
by Intropy :: Tue, 03/01/2011 - 4:09am

For what it's worth, in the 2010 regular season Mendenhall fumbled twice on 324 attempts for a rate of 0.6%. That's 0.94 standard deviations better than average. Prior to this season I would have agreed with the ball security issues claim, but that may not have been fair either. His rate before 2010 was 1.5%, not too much above average. His career rate is now 0.85%, significantly better than average. Perhaps not the best example of prospect with ball security issues. On the other hand the marked improvement between seasons may lend support to the skill claim. The narrative would be that he took the early ball security complains to heart (and there definitely were enough of those reasonable or otherwise) and made an effort to improve in that area.

26
by tuluse :: Tue, 03/01/2011 - 11:26am

Sorry, I wasn't clear. I meant the draft position of a back with elite talent but poor ball security.

28
by Intropy :: Tue, 03/01/2011 - 4:11pm

Ah yeah, that makes sense. I probably should have thought of that. I didn't pay much attention to that part of football back then.

20
by ChicagoRaider :: Mon, 02/28/2011 - 7:41pm

Consider Darren McFadden. He had real fumbling problems until this year. First year, 113 attempts for 3 fumbles. Second year, 104 attempts for 5 fumbles. Then 223 attempts for 4 fumbles, about half his prior rate. So fumbling can change drastically. And of course, he was a very high speed score, at 120 only a tick behind Chris Johnson.

15
by Dan :: Mon, 02/28/2011 - 5:28pm

Da'Rel Scott, Roy Helu, and Brandon Saine also had impressive speed scores, higher than any RB at last year's combine. They're all projected as late round picks. Scott had 5.6 ypc on 430 carries during his career at Maryland, Helu 5.9 ypc on 578 carries at Nebraska, and Saine 4.7 ypc on 301 carries at Ohio State.

25
by Aaron Brooks' Good Twin (not verified) :: Tue, 03/01/2011 - 10:29am

I think the knock on those guys was who they put theirs yards up against. Those yards weren't coming against the Oklahomas or Virginia Techs or USCs of the world, they were coming against William & Mary or Northern Iowa or Youngstown State.

29
by MC2 :: Tue, 03/01/2011 - 8:35pm

Yes, but as Barnwell points out in the original post, a large portion of Fannin's touches came against the likes of Chattanooga and Louisiana-Monroe.

27
by Drunkmonkey :: Tue, 03/01/2011 - 11:33am

So is anybody else surprised at Rodgers low speed score. I mean, it's the lowest one this year. I thought he was supposed to be a good scat back, but this makes him look like a plodder.

30
by MC2 :: Tue, 03/01/2011 - 8:39pm

Well, when you're as small as he is, it's hard to get a high speed score, no matter how good your 40 time is. Speed score is designed more to reward "freak of nature" types who have a combination of size and speed, like Bo Jackson or Herschel Walker.

31
by Nick W (not verified) :: Thu, 03/03/2011 - 6:39pm

Has anyone calculated speed scores for other positions. I calculated former Rams draft pick, Jon Alston's speed score at 119.