≡ Menu

Are some positions safer than others in the draft? Conventional wisdom tells us that quarterbacks are risky, while offensive linemen are safe. But is that true? Jason Lisk wrote a great article on bust rates three years ago at the old PFR Blog, and I’ve decided to update that article based on more recent data.

I looked at the first rounds of all drafts from 1990 to 2009. Over that time period, 46 quarterbacks were selected in the first round, and those quarterbacks were selected, on average, with the 9th or 10th pick. The table below breaks down each position, the number of players selected in the twenty-year period, and their average draft slot:

PosCountAvg Pk

After quarterbacks, defensive ends, outside linebackers, offensive tackles, and defensive tackles have the highest average draft slot (if you include the data from 2010, 2011, and 2012, defensive tackle actually slides into the #2 spot). This reflects the Planet Theory in the NFL: there are only so many people that big with that much athleticism walking the planet, so when NFL scouts find them, they draft them. Early.

We can also see the average AV produced by each position in each of their first four seasons, along with their total AV through four seasons. Inside linebackers and running backs are the most productive players early on in their career, which makes sense if you consider that those positions tend to age very poorly and youth rules the day:

PosCountAvg Pk1234Total

But looking at just average AV doesn’t tell us a lot about bust rates. So here’s what I did. I went through and measured how each draft pick performed (in terms of AV) through four years, relative to expectation (i.e., his draft slot). There were just over 600 players in the study: I labeled the top 100 overachievers (relative to expectations) as “Great” and the next 100 as “Good” while I tagged the 100 biggest underachievers as “busts” and the next 100 as “Bad.” Everyone else was labeled average.

Here’s how you can read the table below, based on the defensive end line. There were 74 defensive ends selected in the first rounds of drafts from 1990 to 2009, with the average first round end being selected with the 14th pick. Only 5% of them were huge overachievers, while 18% were good, 36% were average, 20% were bad, and 20% were labeled as busts.

PosCountAvg PkGreat (%)Good (%)Avg (%)Bad (%)Bust (%)

Mingo against Jake Matthews, expected to go in the first round of the 2014 draft.

Mingo against Jake Matthews, expected to go in the first round of the 2014 draft.

Quarterbacks do have high bust rates. This method identified Andre Ware, Dan McGwire, Todd Marinovich, Tommy Maddox, David Klingler, Heath Shuler, Jim Druckenmiller, Ryan Leaf, Akili Smith, Cade McNown, Alex Smith, Matt Leinart, JaMarcus Russell, and Brady Quinn as busts (and left Patrick Ramsey, Kyle Boller, Rex Grossman, and Jeff George in the “bad” category), and it’s hard to argue with any of those labels.

On the other hand, defensive ends and outside linebackers are just as likely to fall into one of the bottom two buckets as quarterbacks. Pass rushers may be in high demand but they’re not any safer than quarterbacks. You shouldn’t be surprised if two out of Dion Jordan, Ziggy Ansah, Barkevious Mingo, Jarvis Jones, and Tank Carradine turn into this generation’s version of Aaron Maybin, Vernon Gholston, Jarvis Moss, Larry English, Derrick Harvey, or Robert Ayers. Just like with quarterbacks, I think teams find themselves all too willing to overlook certain flaws when it comes to scouting pass rushers.

I’m a little surprised to admit it, but offensive lineman do appear to be relatively safe here, coming in as neither particularly high in the great or bust category. In general, though, I’m skeptical that we can infer much from this data — the sample sizes aren’t large, and I wouldn’t be surprised if some of these trends reverse themselves over the next few years.

What are your thoughts?

  • Great stuff. It’s interesting to group your Bad and Bust percentages together. That changes the dynamic a lot. For example, QBs have far and away the highest pure Bust%, 10 points higher than DE and more than double that of a G or CB. But when it’s Bad&Bust%, QBs finish just behind DE and only a few ticks away from G and CB:

    Pos Bad&Bust%
    DE 40
    QB 39
    RB 37
    OLB 36
    DT 36
    CB 36
    G 32
    T 30
    S 28
    ILB 27
    WR 26
    TE 21
    C 0

  • James

    My thoughts are I’m impressed and jealous of your database and programming/querying skills. To be able to find the the marginal AV of the first four seasons for all the cornerbacks drafted in the first round from 1990-2009 is amazing.

    • Chase Stuart

      Well how else am I going to get people to keep coming to the site!

      To be clear, I used AV and not marginal AV for this post.

  • draftrobot

    I love any analysis on draft theory, but I think measuring bust rates by position as a means to tell you you should or shouldn’t pick certain positions at a certain slot is a bit of a red herring because it doesn’t take into account risk/reward. Teams know which positions help them win games and as we all know teams overdraft players who can pass the ball, protect the passer, catch the ball, pressure the passer, and cover pass catchers. Many times analysts can look at a draft class and tell you some guard, tight end, or inside linebacker is the safest pick, but GM’s know that if their team stunk the year before, a new guard, tight end, or inside linebacker in the first round probably isn’t going to change their fortunes. Like it or not, sometimes you have to hit when you draw a 14. Yes, you may bust, but you know you’re not going to win if you stand on 14.

  • draftrobot

    I think San Francisco is a really interesting case of value positions versus non-value positions because right now they have one of the most complete rosters we’ve seen the last few years and two of their best players are top 10-15 picks at non-value positions, inside linebacker and tight end. However, it took them a while to realize the fruits of those picks. For the first five years Vernon Davis was on the team, they went 33-47 and for the first fours years Patrick Willis was on the team, they went 26-37. Two coaches and one general manager got fired before a team with those two would be any good.

  • Dr__P

    One small quibble and it is basically unavoidable – you have to watch how the AV is calculated for each position.

  • 76mustang

    Just found this and like the overall approach, but it’s interesting that in your QB bust list you have Alex Smith – he may have belonged then (debatable) but he certainly can be moved to a different category.

  • Pingback: “Getting down in a [three-point] stance isn’t natural.” | Get The Picture()