Athleticism and the Defensive Tackle

Since I kind of enjoyed comparing the athleticism of offensive linemen to their CarAV scores (Career Approximate Value), I thought I would do something similar for defensive linemen.  For this post, I will only be looking at players who could be defensive tackles or 3-4 defensive ends, so I'll just use the 280# mark as the cutoff.  As much as possible, I want to be comparing apples to apples, so having a simple cutoff point determined by weight should be reasonably sufficient. 

I'm going to lay this out the same way that I did for the post about Athleticism and the Offensive Line part 2.  For each draft class, from the year 2005 through 2010, I will list the three players with the highest CarAV score (a basic measure of their productivity and how active a starting role they have) who were taken in the 3rd round or later.  I will also include the one player from each draft class, who was selected in either of the first 2 rounds, who had the lowest CarAV score.  What we are looking for is some sign as to whether the mid-to-late round "successes" are demonstrating above average athletic ability, and whether the more highly drafted disappointments showed below average athletic ability.

To keep things relatively simple, their athletic ability will be measured based on the Kangaroo Score and Agility Score.  Both scores are shown in the form of how many standard deviations above or below average that a players is, relative to the average player in their position group (in this case defensive tackles).  I'm also going to toss in their times from the 40 yard dash, even though this is of questionable value for defensive linemen.  I have the numbers, so I might as well throw it in there, in case anybody is interested.

As I have said before, I have some reservations about using CarAV to evaluate players.  Still, as long as you are comparing players from the same draft class, at similar positions, CarAV is generally adequate.  Including a player amongst the "successes" from a given year, doesn't mean that I endorse them.  It merely means that they have gotten on the field more often, and probably contributed more, than most of their peers from that draft class.  I'm mainly using CarAV as the measure of success, since it seems more objectively acceptable than just throwing out my opinion that "this guy sucks".  No current statistical measure can quite capture the "suck factor", at least not in a way that satisfies me.  So, without further ado, here are the numbers.

Player                         Year      Pos.         Pick#   40-yd          Kangaroo        Agility     CarAV

Geno Atkins	2010	DT	120	4.75	0.793	1.056	33
Corey Peters	2010	DT	83	4.98	-0.344	-0.153	18
Greg Hardy	2010	DE	175	4.82	0.517	0.631	14
Torell Troup	2010	NT	41	5.12	-0.481	-0.158	1
Terrance Knighton	2009	DT/NT	72	5.21	0.293	-0.758	21
Roy Miller	2009	DT	81	4.95	0.317	-0.837	16
Vaughn Martin	2009	DT	113	5.00	1.179	1.442	12
Ron Brace	2009	NT	40	5.48	0.157	-1.040	6
Kendall Langford	2008	DT/DE	66	4.95	-0.007	-0.237	29
Ahtyba Rubin	2008	DE/NT	190	5.20	0.125	-1.560	26
Red Bryant	2008	DE/DT	121	5.00	0.660	-0.159	16
Kentwan Balmer	2008	DE/DT	29	5.29	-0.245	-0.451	7
Brandon Mebane	2007	DT	85	5.15	-0.762	-0.495	39
Paul Soliai	2007	NT	108	5.10	1.336	-0.139	27
Antonio Johnson	2007	NT	152	5.15	0.236	0.611	18
Justin Harrell	2007	DT	16	5.04	0.041	-0.343	2
Barry Cofield	2006	DT	124	4.95	0.529	1.095	40
Domata Peko	2006	DT	123	5.27	-0.365	0.174	39
Kedric Golston	2006	DE/DT	196	4.88	0.659	0.549	23
John McCargo	2006	DT	26	5.15	0.283	0.137	3
Jay Ratliff	2005	NT	224	4.85	0.821	1.530	51
Sione Pouha	2005	NT	88	5.02	1.381	0.072	36
Jovan Haye	2005	DT	189	4.74	0.003	1.069	26
Shaun Cody	2005	DT	37	5.05	-1.168	0.163	27
Avg. Successes			128.3	4.99	0.409	0.216
Avg. Busts			31.5	5.18	-0.235	-0.282

There are two notes I should mention about this list.  First of all, I used John McCargo to represent the highly drafted "bust" for 2006, rather than Claude Wroten.  I did this because the data for McCargo was more complete.  If you are interested, Claude Wroten has a CarAV of 2 (1 point lower than McCargo), and a Kangaroo Score of -0.338 (worse than McCargo's 0.283).  It was due to Wroten's lack of data regarding his agility scores that led me to leave him out.  If I had included him it would have lowered to average Kangaroo Score for the busts to -0.339 (as opposed to -0.235), so I feel I am being somewhat generous by excluding him.  Secondly, since the Kangaroo Score combines the players performance in the vertical jump and the broad jump, this created a problem in the case of Vaughn Martin, who only participated in the vertical jump.  Since that was the only data available for him, that was what I used.  It shouldn't create much of a problem though, since the vertical jump tends to be more reliable than the broad jump as a predictor of success.

With that out of the way, we can get to some of the simple observations.  Overall, players who were drafted in the 3rd round or later, who achieved some degree of success, were 0.644 standard deviations better than there more highly drafted peers, when it came to the Kangaroo Score.  When it came to the Agility Score, they were 0.498 standard deviations better than their more disappointing and highly drafted counterparts.  The players taken in the later rounds, who had some arguable success, also ran the 40 yard dash 0.19 seconds faster, on average, than the more disappointing players from the first couple of rounds.

While there are numerous players in the "success" group that I don't particularly care for, this isn't my concern.  Success, in many of these cases is somewhat relative.  My biggest concern with assembling a list like this, is that I really don't think all defensive linemen should be lumped together in one group.  I feel that as you move out from the nose tackle position towards the defensive end position, the importance of the Kangaroo Score (raw explosive power), begins to diminish somewhat, and the value of the Agility Score goes up.  At the same time, I don't think the Agility Score matters too much for nose tackle (though it is still nice to have a good score).  None of this really matters though, for what I am doing here.

For this extremely simple examination, I'm just curious as to whether late round players, whom teams expect less of, and probably give fewer opportunities to, will rise to the top, to get their team's attention.  On the flipside, it is interesting to note whether the highly drafted players, who became disappointments, were obviously lacking something when it came to their physical measurables.  On both sides of the equation, things seem to work out pretty much as you would expect.  While the highly drafted disappointments might not have abysmal scores, they do tend to fall into the mediocre range, with none of them scoring better in either category than a 0.283 (John McCargo's Kangaroo Score).  Compared to some of the more noteworthy names from the later rounds, where 38% of the 18 "successes" had at least one score of 1.000 standard deviations above average, and 61% had at least one score that was 0.500 standard deviations above average, the results of the highly drafted disappointments becomes a bit glaring.

If people want to dismiss a player's physical measurables as being inconsequential, that is their right.  I wouldn't do it, but I'm just an bozo who rambles on the internet.  I have no problem with the idea that a player can succeed despite poor test scores. I just think that if I were making a pick in the first couple rounds, I would probably have fewer concerns about a player who was physically superior.