All Americans and the NFL Draft

Submitted by bronxblue on

With the NFL draft rapidly approaching and college football existing in a tantalizing “spring game and then random non-mandatory-but-if-you-plan-on-playing-you-better-show-up practices” realm until the fall, my mind began to wander
about football. Perhaps spurred on my Mel Kiper’s impeccable coif, I started to think about the NFL draft and the immense differences in how college and professional players are rated depending on who is providing the analysis.

With college players, the focus tends to be on the player's "fit" within a system, while in the NFL the focus seems to be
far more on a player's raw "numbers" and physical attributes. For example, the proliferation of spread-style offenses in college has made 5'-10" WR/quarks far more valuable than they are to a professional football team, while some CFB teams have moved away from the massive-but-speed-impaired OLs that still find homes on most NFL rosters. While there remains a significant number of skills that translate well no matter what person is doing the analysis, it does seem that being a star in college does not necessarily oneself of that success following you into the pros.

Now, that probably is not a revelation to most people, but it got me thinking about CFB All-Americans and how they
are evaluated by NFL scouts, the theory being that the highest-rated players tend to be drafted early. Every year, it always seemed like two or three All-Americans (usually QBs, but other positions as well) either were drafted in
the lower rounds, or not at all. It always struck me as odd how these guys can dominate the college game, oftentimes
against players who are later drafted before them, but still evaluated as borderline pros by NFL front offices.

With that in mind, I set out to determine how success in college translated to the NFL, broken down by position.

First, the set-up:
Set-up and caveats:

  • Draft years: 2001-2008
  • I used only the players from the first-team AP All-American list. My reasons are two-fold for selecting this particular cross-section of college football's elite.
    1. I chose the AP list because it is the one most commonly cited when discussing a player's collegiate success, and it seemed to have far less WTF selections compared to ones put out by the Coaches or TSN.
    2. I limited by analysis to the first team because those players were regarded as the "best" that year, and enough guys moved up and down the 1st, 2nd, and honorable mentions throughout their careers that it made
      my head hurt trying to keep them all properly slotted.
  • So you didn't get drafted... - Believe it or not, some All-Americans were not drafted by an NFL club (sorry, Mr. Shazor). In those few instances where a player went undrafted, I assigned them a round of 8 and an overall draft position of 256, which are 1 more than the respective limits of rounds and positions in the NFL draft.
  • APs and FBs - In college, a position on the All-American team is reserved for an uber-athlete who, in most instances, is a kick/punt returner. In the NFL Draft, though, these players are listed at their "preferred" position.
    Similarly, players are drafted as FBs even though the AA team does not feature such a position. So both those
    categories are out. Note to those worried about these omissions corrupting my numbers - most of the APs were WRs
    or RBs drafted around the same round and pick number as the rest of their position, while FBs never factored into
    most AA teams and were drafted so low that they would have artificially depressed the numbers for whatever group -
    RB or TE - I tried to shoehorn them into.
  • The Lines - People may notice that I lumped defensive and offensive line players into two groups - DL
    and OL - without regard for their position along said line. I know, this undoubtedly screwed with my numbers. The first problem is that the AA teams do not list separate positions on either line except at C. Sure, I could have gone based on the position they were drafted/listed at, but that might not be the position at which they attained AA status
    while in college. So yeah, good LTs are going to be gone far sooner than good RGs, as will good DEs compared to good NTs, but I'm willing to accept that variance here. Furthermore, guys in college move around all the time, sometimes lining up as DE on one play, a hybrid linebacker the next, and on the inside as a tackle on another, all depending on the matchups. Similarly, while a great LT might not be moved around much, injuries and even particular formations may lead to guys bouncing around from G to T throughout the year, making a single position difficult to ascertain. So I bunched everyone into line play, and
    I accept all criticisms that come with that decision.
  • DBs - See above for my logic with bunching guys together. While the AA team does have a separate listing for safeties and cornerbacks, their variability of position (CB, S, or LB) at the draft made it difficult to determine where many of them fell. For example, Marlin Jackson was a CB while at UM, but has been more of a safety while in the pros. That seemed to happen more with DBs than any other position I followed, so I figured I might as well
    lump them together and accept that the numbers would be a little skewed.
  • Sample size - I know, I know. With samples of 7-8 players, of course one or two outliers are going to
    knock everything out of whack. For that, I
    apologize, but I am gainfully employed, recently married, and only have a finite number of hours a day to spend
    surfing the Internet for All American teams
    and yearly drafts. Take all of the numbers with a massive grain of salt; that said, the trends you'll see in the
    numbers, at least to me, keep in line with
    my expectations going into this project and match, I hope, with the conventional wisdom shared by others.
  • Math: I'm a computer engineer from UM who was, at one point, decent with statistical analysis methods.
    Over the years, though, my knowledge has retreated farther and farther in the recess of my mind, replaced with Family Guy quotes and the rules of eminent domain. As a result, I limited my analysis to average draft position for the All Americans, the average draft position for every player at that position (with the All Americans removed from the pool
    so as to not skew the numbers), and standard deviations for both. Since my sample sizes were relatively small, the standard deviations are all over the place, and are practically useless beyond a "hey, that's interesting" viewpoint. I know there are other models and methods that might make more sense of this data, so look below for a link to part of my data (I can upload the full file if anyone really wants it).

My expectations - i.e. my uneducated beliefs about football:

Before jumping into the data, I'll quickly recount my expectations going into this little analysis.

  1. I've been Weinke'd - Though this was based mostly on my recollections of such college studs-turned-pro-duds as Chris Weinke (the greatest travesty in Heisman history), Jason White, Eric Crouch (not in sample), and Tim Tebow (jury is still out, but just saying...), I figured the QB position would show the greatest divergence between All American status and actual draft position. In college, where specialized systems are rampant and guys like Graham
    Harrell, Colt Brennan, and Chris Leak can dominate despite clear deficiencies, it would make sense that they would
    no fit snuggly into most pro systems and, as a result, drop in the draft.
  2. Fast little guys - I have always heard from the talking heads on ESPN, Fox Sports, etc. that the two
    positions where the transition from college to the pros (outside of special teamers like Ks and Ps) is easiest is at RB and DB, especially for college corners. That makes sense to an extent, as those positions rely most heavily on pure athletic ability. So I expected to see the the greatest deviation in draft position at these two positions, with AAs
    being drafted far higher than the "average" player at that position.

So on to the chart? Yeah, chart:

Position Count Average Round Average Draft Pick Std Dev Round Std Dev Draft Pick
C - CFB 8 4.5 135.38 2.78 101.15
C - NFL 51 4.88 152 1.84 66.72
 
DB - CFB 31 2.35 60.71 1.98 68.64
DB - NFL 356 4.23 127.82 1.93 69.35
 
DL - CFB 28 1.96 44.96 1.67 58.13
DL - NFL 317 4.24 127.64 2.06 73.73
 
K - CFB 6 5.33 163 2.5 90.92
K - NFL 15 5.87 183.27 1.19 50.53
 
LB - CFB 18 3 79.44 2.35 83.62
LB - NFL 234 4.32 130.36 1.81 67.14
 
OL - CFB 27 2.44 58.33 1.8 62.86
OL - NFL 260 4.67 143.92 1.92 69.99
 
P - CFB 7 5.29 165.71 1.98 70.33
P - NFL 12 5.33 163.17 1.15 45.01
 
QB - CFB 7 3 81.71 2.77 100.75
QB - NFL 101 4.2 126.04 2.14 77.71
 
RB - CFB 15 2.87 77 2.33 80.38
RB - NFL 138 4.29 131.53 2.08 76.91
 
TE - CFB 8 2.5 66.38 2.51 84.96
TE - NFL 112 4.57 142.23 1.91 69.89
 
WR - CFB 14 2.36 53.29 2.1 73.82
WR - NFL 245 4.44 136.22 2.03 74.51
 
AP - CFB 7 3.57 107 2.23 80.01
FB - NFL 31 5.16 160.94 1.39 53.33
 

For individual draft positions of AAs and the NFL draft in general, click here.

So that was interesting. Some observations:

  • Being an AA clearly helps your chances of being drafted. At every position, guys who were AAs were drafted
    before the "average" player at that position. If one throws out Ks and Ps, in fact, most players AAs were drafted between 1 and 2 rounds before the average player, which amounted to millions of dollars in compensation and a far greater odds of making it on an NFL roster. So as a PSA - kids, try to be All Americans in college except...
  • If you kick for a living and/or are used to having a guy's hands between your legs. Ps, Ks, and Cs received
    comparatively small bumps in their draft stock for being AAs, though all three positions were drafted far later on average than other positions on the football field. While the special teamers really did not surprise me, one always hears how Cs are the smartest guys on the field and, as such, you would think such a commodity would be at a premium
    come draft day. I will leave the explanation as to why Cs are drafted so much lower to those who know more about football than me.
  • QBs on both side of the line struggle - You always hear about LBs being treated as the "QBs of the
    defense," and at least on the AA team that seems to be true - both positions were consistently drafted lower than others. As I said above, sample size and what-not certainly had something to do with this theme, but the QB position in college is almost a different species compared to the NFL, so the divergence in draft status versus college success
    doesn't really surprise me. LB was a bit more of a shock, but it does seem that LBs (and DEs) benefit the most
    from the various systems run at the collegiate level as well as the relative strength of the line in front of them. If the DL can hold up the blockers from reaching that second level, it makes sense that free-flowing LBs are going to rack up huge tackle numbers that, inevitably, raise their national prominence. Plus, there does seem to be a
    movement in the pros to draft smallish DEs in college to play LB in the pros, so maybe the LBs in college are getting squeezed down the line because of this phenomena as well. Again, smarter people than me can probably explain this better.
  • DLs are rolling in the money (money!) - The old maxim is that you can't teach size and speed, and clearly NFL scouts have taken this to heart when evaluating DLs from college. The average AA DL is drafted before the end of the second round, which means the average DL is assured of millions before even stepping onto the football field. Furthermore, the standard deviation for the position was the lowest of all positions, meaning that most top college DLs
    were gone on the first day of the draft. Even though some of these DLs are undoubtedly projects, it is clear that solid DLs in college are at a premium in the pros, and teams are willing to take fliers out on these physical freaks.
  • OLs are not doing too badly either (um, slightly less money!) - Again, a premium on speed and size on one end would beget a premium on the guys on the other. Pancake factories and hulking bulls are evident in college and, it seems, are quickly snapped up by the pros as well. While OLs were drafted a little later on average than their DL counterparts, both sides have clearly benefited from the increased exposure and emphasis teams have placed on the
    trenches.
  • DBs and RBs, plus WRs - I proferred the theory that DBs and RBs tend to possess skills that translate well between college and the pros, and as a result top players in college would be drafted early on in the pros because "you know what you are getting with them." Well, it looks like I was partly correct. Top DBs tend to get drafted early on, as the speed and size maxim held up. If you can run really fast backwards and then be able to jump really high to knock down a pass, you will likely dominate at every level of football. While there certainly are systems in both college and pro that can mask some deficiencies in speed and size, DB seems to be one of the positions where great athletes are easily identifiable and measurable. That aptitude shows up
    pretty quickly on the college field and, apparently, in the eyes of pro scouts.


    A similar story can be found for the men who most frequently match up against DBs: WRs. If you can run a 4.4 40 and stand 6-5 or more, you will certainly dominate college and, at least initially, be looked upon favorably by pro scouts. While there have certainly been a number of high-profile flubs from this group (looking at your, inmate #4587...I mean, Charles Rogers), there have also been some on-the-spot hits (Larry Fitzgerald, Calvin Johnson, Braylon - if
    he remembers how to catch the football again). Plus, I think part of reputation surrounding WRs in the draft is
    directly related to the incompetence of the Lions' front office, with an abnormally large number of turds floating in their punch bowl.


    As for RBs, I think the reason they are drafted over such a dramatic range (check out the Std Dev) has to do with the fact that NFL GMs see them largely as replaceable parts, pieces that break down quickly and, thus, should only be highly paid if they are exceptional. The oft-quoted statistic is that starting RBs last about 3 years in the NFL, so most teams are loathe to spend a first- or second-round pick on that position unless they believe he will have a long, successful career. Another factor that may play a role in the draft deviation is that the running back position is being deemphasized on a number teams,
    with more teams adopting a back-by-committee approach. With less of a focus on a single dominant workhorse, top backs are being drafted farther down the line. Finally, and I guess I'll call this the Ron Dayne-Javon Ringer scenario, some college RBs dominate because they run the ball an ungodly number of times, resulting in huge numbers at the expense of shortened pro careers. Pro scouts have likely noticed that some of these guys have well over a thousand
    college carries on their legs before taking a snap in the pros, and again don't want to pay for players likely to
    break down.

So that's about it from my end. I would love any comments or criticisms, so go crazy in the comments section

below. Furthermore, if you note some glaring flaw with my data and/or analysis, please point it out (but in a way that doesn't sound TOO condescending) as well.

Comments

Chrisgocomment

April 19th, 2009 at 6:16 PM ^

This is really cool, nice job. Interesting topic and some painstaking analysis, thank you for all the work you did here.

FYI...the last time I was anywhere near a statistics book was in 1997 so I'm no help there, but I do know that the WR you mention is "Andre Johnson" not "Charles Johnson". Seems you had Rogers on the brain!

DesHow21

April 19th, 2009 at 7:35 PM ^

What in the devil does draft round for CFB mean? I know Tressel and Carroll pay their guys, but do they have a draft going now?

Explain what average round means for:

C-CFB
C-NFL

the_white_tiger

April 19th, 2009 at 9:17 PM ^

It is true that Jason White did nt transition into the pros, but I remember his OU days fondly and without 2 season ending knee surgeries he might've benn a decent NFL QB. Good luck to him.

Six Zero

April 21st, 2009 at 9:20 AM ^

If for nothing else, this much work deserves a grade. You just earned 3 credit hours at THE Brian Cook University.

Seriously, so many posts often seem to display what I can only describe as negative thought and/or analysis, so it's nice to see such an organized, well-prepared and convincing argument.

Or as they say in Columbus, Your brain thinks good.

Seth

April 21st, 2009 at 5:28 PM ^

Very impressive work, and a reminder of why I still browse the new diaries first.

If there's anything I think we can take away from your quite thorough analysis, I would say it seems that the difference is how much of the NFL skill set for each position are necessary in the college game.

Meaning, if you're really big, but slow, you can set career records at tailback, but the speed will hinder you in the NFL, where you can't bowl as many players over.

This seems most prevalent at Linebacker and Quarterback. A QB with legs and an O-line can take home a big bronze statue in college, but the NFL still drafts for arm strength and accuracy. A linebacker with a lot of speed (e.g. Boss Bailey) can dominate the college game, but doesn't have the body to succeed at LB in the NFL. Or vice versa.

In other words, for certain positions, if you can do one thing extraordinarily well, you can be an All-American. In the NFL draft, a weakness at one position skill will kill you (unless you're a slow-ass receiver and the Lions' pick is coming up)