How to Draft a Pass Rusher

From the FanPosts. Love this research. -Joel

The Chiefs aren't in terribly bad need of pass rushers going into this year's draft. Of course more never hurts, but I'm happy with what we could field today. That said, I suppose this isn't the most germane post as far as looking ahead to this draft goes, but pass rush specialists have a special place in my heart. "Why?" you might ask: well the answer is simple. Ten years ago a pass rushing linebacker shot a gap, sacked me, and in doing so split my patella in half. Thus ended my HS football career(my Al Bundy "coulda gone pro, but..." story). Since then I've taken pass rushers pretty seriously. Take a stroll with me through some wonky data as I explore what makes a nightmare pass rusher and where they come from.


I gathered together all players who landed in the top 10 of sack totals 2005-2011 at least two separate years. This is what constitutes what I'm calling my core sample. I thought this would be a fairly good "cream of the crop" grouping. There are a total of 15 players in this sample group. Being an elite group also means that it is a small group, so I also wanted an expanded pool to broaden the statistical base. I did this by including players who were top ten at least once in the same time period. I also wanted to look for some players that passed the eyeball test, but no one really obvious seemed to be omitted. I did have to factor out some players due to insufficient data. Still others were omitted due to being drafted too long ago or having erratic careers which I felt polluted the data. Among such players are Mike Vrabel and Patrick Kerney. The combination of this peripheral sample and the core sample is called the total sample.

For all such players I compile data for height, weight, conference of play, college sacks, and pro sacks. For height and weight data I use the modal datum. If the numbers are all over the place, as was sometimes the case with weights, I'll use a median number or the one I felt was most reasonable. The only note on conference of play is nomenclature. I used current nomenclature regardless of what the conference may have been called when the player was drafted. The college sack data was hard to find, so it is missing for some players on the table, but not many. Yr2 is the main column to pay attention to; it's their final year of play in college. Yr1 is their penultimate year of play but really wasn't terribly significant. If all I had was a combined sack total for two years in college, I gave half to each year(eg. John Abraham). The pro sack number is adjusted sacks per season (career sacks/games played) x 16. This is basically how many sacks they've had for every 16 games they've played.

Finally here, I want to note just what I'm doing. This is a hot-spotting technique. I'm taking a definable pool of success -- or stated otherwise -- a known, desired outcome, and seeing if I can project it backwards to find indicators of future success among collegiate players. This method will ferret out indicators that DO lead to future success. DO is an important because it isn't a WILL statement. This is to say that following this track will probably get you a good player, but could overlook a good one who may not be as conventional. Yes their is survivor's bias in here, but that's really what I'm trying to harness.


Here is the date table I've constructed. Core Sample guys are bold while rookies are in red. There is no statistical adjustment made in account for rookies, I just denote so you can toss it if you feel they don't count. The only real impact they had was to make the Big 12 look better. Now, as I was filling this out, some things became pretty apparent early on. Height was the first one. 75in, or 6'3, was clearly the modal height with little deviation. Weight deviated more liberally, but still kept a solid, and frankly intuitive range. I'll talk more about what the deviations mean later. No BCS conference really stood out atop the others in producing pass rushers but I do think a point about where to look is made. Sacks as a final year player in college seemed controlling. Penultimate seasons were all over the board ranging from better than the last year to useless. Adjusted sacks/season shouldn't surprise anyone given that a high number was basically the criteria for entering the pool. If you care to know, just about all of these players were good for 0.6 sacks/game.

(Note: The SEC total at right is incorrect. It should read "2")

The following table is a scatter plot of heights and weights for each player. What we get is something of a mix between a slope and a bell curve. The slope is an intuitive function. As people get taller they get heavier. The bulge in the middle is what I find interesting. You get a few more players as Ht/Wt increase until you hit 6'3, where it just explodes. You get a gentle tail off at 6'4 but still a good number of players. After 6'4 you find a steep dive. I'd also like to note that the top right quadrant of the chart seem as though it can be sectioned off as an altogether different species of player. Notice that all players who are 6'4 are within the 255-265 range, making them more similar to the 6'3 and under players.


Now, not all of these players come into the league playing the same position. They all rush the passer primarily, so they're tasked similarly enough for collective analysis, or so I think. Just to make sure, I explored the difference between the DEs and OLBs in the list. Averages for both positions are overlay-ed onto the scatter plot for comparative purposes. The lines for OLBs are in blue, the DEs in green, the gross average is in red. Again we see a fairly intuitive conclusion. Ends are bigger than OLBs, on the average. The Ends have a higher weight ceiling than the OLBs but are only marginally taller. There is still generous enough overlap that I wasn't too put off with assessing them all together. If you look at the area sectioned off in the middle by the intersection of all medians I think you will see what could be an optimum range.



In looking at the data I think I've across some indicators of future performance. So, now that your attention span is already exhausted, I'll get to my rules for drafting a pass rusher. I include explanations and idicators.

Rule 1. Look for producers in college.

Your first task is to find guys who were good in college at doing what you're going to ask them to do in the pros. On average college players were worth about 80% of their college sack production in the NFL. The real number to look at is their final year. The "one year wonder" effect didn't seem to bear out in these numbers. As long as their final campaign was good, they were worth drafting. The impact of previous college seasons, if any, was reassurance if sack totals went down the following year. A player is still worth drafting if their sack total went down their last season, even markedly, so long as it was still respectable. I'll call respectable 8+ sacks. Double digit sacks in college over two seasons, even in the midst of a drop off, should not cause hesitation.

Rule 2. Look for players as close to 6'3, 260 as possible

This just really seems to be the wheelhouse for great pass rushers. The real effect of this in your decision making is to determine if a player who meets rule 1 can keep up their performance in the pros. Their is something very important about the height number that I suspect is at play. Though I did not have numbers to analyze it appears to me that 6'3 is the height at which players are tall enough to have an arm length that allows them to defeat NFL Tackles at the edge. I think height is only a correlating factor to what is really at play here which is the arm length.

Rule 3. Wingspan 32

Wes Bunting has a good article on offensive tackles and how arm length plays in. The same is true for edge rushers who have to likewise need long arms to not get leveraged out and keep clean in their attack. I've adapted his formulation for tackles to suit edge rushers. Basically its a down ward adjustment rank for rank since OTs regularly exceed 6'7.

OK - 32"

Good -33"

Great- 34"

Basically 32 inches is a passable arm length. Really you want to see it get up over 33 though to have high confidence. If you should find a player with over 34 inch arms, and they have reasonable tape, you should jump on them. 35 1/2 inch arms is what I feel allowed Aldon Smith to post such a gaudy sack number for a rookie. The average arm length for a top ten OT is a shade under 35" so if you're pushing or exceeding that, you stand a great chance of winning at the edge. This is an area I'd love to have more statistical support for, so if anyone knows where to get back arm length data, let me know and I'll do an addendum.

Rule 4. Know how to deviate

If you're going to deviate from the 6'3, 260 formula, there is a way to go about it. On height, deviate up, not down. If you like a shorter player you should only draft them if they pass the wingspan test and pass it comfortably. This could actually get you a value pick later in the draft since the undersized guys tend to slip. Robert Mathis is a good example of this. He's undersized but has 34 1/2 inch arms. On weight one should deviate in as small a percentage as they can, either way, from the 260 mark. If the player is in the 250s or 260s you're fine. If you crack 270, now you need to check their 40 time. Basically if they're going to have the weight, it can't be dead weight. They need to prove they can move it well. The big boys, who are really more 4-3 players, had very good relative 40 times: J. Peppers - 4.68; M Williams - 4.70; J P-Paul - 4.71. If they're nearing of clearing 4.7 at 280lbs, they can be considered.

Rule 5. Know where to look

In a surprise to me, no conference really stood out as the go to place. I really expected the SEC to have a bunch of players among the top rushers but it didn't bear out. The SEC only had 2 players make it, which was a weak showing. Troy university alone produced that many, which accounts for all of the Sun Belt players. The ACC showed up as the best conference, which was again, a surprise to me. Big East, Big 10, and Pac 12 all had a respectable 4 players make the chart. The basic takeaway is that you can get a run pass rusher in the mainline BCS conferences. There is no go to well to find these guys. The other big thing has to do with small school guys. This is where Rule 1 and 5 dovetail. If a player at a small school has a ridiculous sack total, they're draftable. When they're over-performing don't hold the level of competition against them, so long as they measure up well otherwise. The best example here is Jared Allen. This is again another way to snag a starting caliber player late. Speaking of draft stock, the obvious story was the one told. If you want a sack artist the 1st round is typically the place they reside. If you hope to find a guy late, there are two reasons good players fall. First is size and second is weaker level of competition. Agian, size should be a red flag, but if they have the arms and the sack stats you should be okay picking them. Competition level is okay to overlook if the player was dominant.

And finally, a hit, a miss, and a projection

These rules would have hit on Justin Houston. He fits the bill in every department. He's 6'3, 258 with 34 1/2! inch arms. He had 10 sacs in a mainline BCS conference is final year of play. He's basically the prototype. I expect a lot form him.

These rules would have missed HUGE on Tamba Hali. His colleg sacks and height were the only measures he passed on. Coming out he was too heavy and 275 with a bad 40 at 4.8. He's now lighter I assume, probably more like 265. His 3 best years were as a lighter 34 LB though. The big death nail for him on my draft board would have been his arms . They're barely over 30 inches. This emphasizes what I said earlier about a DO not WILL statement on how these things project.

So who projects the best, according to my rule book, out of this year's draft?

Whitney Mercilus. He's 6'4, 261 with nearly 34 inch arms. He led the nation in sacks his final year in a mainline BCS conference. This guy looks like a Demarcus Ware clone. He looks way better to me than guys who will go ahead of him like Upshaw and Ingram.

That's it. Thank you for reading.

This is a FanPost and does not necessarily reflect the views of Arrowhead Pride's writers or editors. It does reflect the views of this particular fan though, which is as important as the views of Arrowhead Pride writers or editors.

Log In Sign Up

Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Arrowhead Pride

You must be a member of Arrowhead Pride to participate.

We have our own Community Guidelines at Arrowhead Pride. You should read them.

Join Arrowhead Pride

You must be a member of Arrowhead Pride to participate.

We have our own Community Guidelines at Arrowhead Pride. You should read them.




Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.