Much like my last post, this one is going to be technical. I will use math (and statistics). If this bothers you, stop reading now. Having a great deal of mass and being able to accelerate quickly (quicker than the offensive linemen) is beneficial for a defensive end (or 3-4 OLB) in football. In my last post we looked at some of the guys in the the 2013 draft in terms of how quickly they could accelerate and how much momentum they could generate.
The M10 statistic from combine (and pro-day workouts) is how much momentum (velocity times mass) these guys have at 10 yards in their 40 yard dash. This is measured by their 10-yd split on the 40. One of the questions that came up in the discussion of the last post was: how does this translate to NFL success for these guys? Is there a correlation between M10 and NFL success for pass rushers? Read on to find out...
Of course no single stat is going to tell the whole story of any player, but some are more predictive of NFL success than others. Draftmetrics.com has done a great deal of work on this, but they have not done this study (mainly because they don't use either stat that I am using and they judge success by years as a starter).
So first off we have to define "success" or at least figure out how to measure it for a DE (or 3-4 OLB). I chose not to focus on sacks, which are the gold standard for some, but not me. Pro-football-reference.com has developed a metric called Approximate Value (AV) that allows us to compare the relative value of every player in the NFL on any given season. For example, Wolfe had an AV of 8 last season. Von Miller had an AV of 17. PFM had an AV of 15. You can argue over relative values if you want, but AV provides a good way at looking at all of the edge defender contributions to winning - more than just sacks.
So what I have done is go back and calculate the M10 value for every DE (and some 3-4 OLBs) that were drafted from 2008 to 2012. Some players did not run the 40 so they were excluded since their M10s can't be calculated. I then looked at each players AV/yr in the NFL and tested to see if their was a correlation between M10 and AV/yr. In other words, does the ability of an edge-rusher to generate momentum quickly have a correlation to NFL success? I'll show you the data for each year and then the actual correlation value. A (theoretical) correlation value of +1.00 would mean that AV/yr tracks perfectly with M10, meaning a higher M10 always tracked with a higher AV/yr. A negative correlation factor would mean that that number went up AV/yr went down. As always, correlation is not causation.
Players were selected by listing as DE by pro-football-reference.com and/or NFLdraftscout.com along with a few of the larger ~260 OLBs from NFLdraftscout if the player is not listed as a DE by PFR.com
Listed by AV/yr
Correlation factor = 0.16
So for this year, it does not appear that M10 correlates well to NFL success. Johnston had the best M10 value is struggled to make an NFL roster. Golston had a good M10 and has been a historically bad bust. Conversely, Groves had the worst M10 of the DEs and yet has been a productive NFL player, albeit not a star.
FWIW, no combine stat showed any significant correlation with AV/yr for the 2008 class.
These guys are not in draft order (I had to add some of the guys at the end who were OLBs in the draft by some reckonings)
Correlation factor = 0.499
So for this year M10 was a much better predictor of NFL success for edge defenders. The methodology fails when these guys are big enough to play DT in the NFL even though they were listed as DEs in the draft. Bigger guys are always going to have higher M10 since the weight range (232-306 lbs) is much greater than the velocity range (5.1 to 6.1 m/s) for all of these players. I struggled with whether or not to put a weight cap on the study (Hood, Magee and Jackson are all ~300 lbs). Slade Norris only weighed in at 232 at the combine.
Two other combine stats that showed a decent correlations to AV/yr for this year were broad jump (0.37) and vertical (0.35). Nothing else showed any significant correlation.
These guys are listed by draft order
Correlation factor = 0.466
We get roughly the same correlation that we got from the 2009 DEs. JPP had a fairly low M10, but he has been the best edge rusher in this draft. Ej Wilson and Clifton Geathers, are two guys with big M10s that have not done anything in the NFL.
There were significantly fewer DEs (or at least guys listed as DEs) drafted in 2011. Here they are listed in AV/yr order
Correlation factor = 0.33
This year had mediocre correlation. Part of that is due to the smaller sample set, but the other part of that is that the guys who have gone on to be successful from this year generally get their mass moving quickly, but have less mass. The combine stat that had the highest correlation to AV/yr for this class was broad jump (0.60) with explosion number (0.59) a close second.
There were 31 "edge rushers" drafted in 2012. Here they are listed in draft order. The caveat they only have one year of NFL production so these guys have potential to get better where as the guys from 08 highly unlikely to get significantly better at this point in their NFL careers.
Correlation factor = 0.13
So we have the lowest correlation factor of the 5 years last year. Perry, Hicks, Bilukidi and Tyson all have amazingly fast 10-yd splits for men of their mass, yet none of these guys contributed much as a rookie. The combine metric that had the best correlation to AV for the 2012 rookie edge defenders was vertical, but it was a negative correlation (-0.31), meaning that the guys with the best verticals did not do as well. Upshaw, Winn, Irvin and Coples all had below average vertical leaps.
So what does this all mean?
It looks like M10 might have slightly more predictive power than other combine stats, but that it is certainly not something to look at in a vacuum.
Here are the M10 values for the 2013 edge defenders (along with 10-yd split and CoD values)
Were I working in the FO for the Broncos, I would use this data to look for hidden gems in the draft (Taylor? Edwards?) or to look at guys who succeeded in the college game without great measurables (Moore? Werner?). Of course I would only use this as one leg of a tri-pod. The three legs being: combine numbers, in-game production, and "eye-test" film watching.
Thanks for reading and Go Broncos!