≡ Menu

In the 2016 NFL Draft, 32 cornerbacks and 31 wide receivers were selected, making those the two most commonly-drafted positions this year. That’s not too surprising, of course, as cornerbacks and wide receivers litter the field on Sundays. But the graph below shows the number of plays drafted at each position:

players position 2016 draft

A little more interesting would be the Draft Value used on each player: after all, spending a high pick on a player means a lot more than spending a low pick on one. Here, we see that cornerback stands out: teams are more likely to use high picks on cornerbacks and late-round picks on wide receivers, at least in 2016:

value position 2016 draft

Here’s the same information in table form:


But I think the most useful way to look at this information is on a per snap basis. Based on PFR data, the average offensive snap last year featured 1 quarterback, 0.98 RBs, 0.20 FBs, 2.47 WRs, 1.30 TEs, 2.09 offensive tackles, 1.91 guards, and 1.05 centers. On defense, the breakdown was 1.79 defensive ends, 1.49 defensive tackles, 3.01 linebackers, 2.55 cornerbacks, and 2.16 safeties. Unfortunately, there is no OLB/ILB breakdown, so I will make one up: my hunch is it is in the ballpark of 1.2 ILBs and 1.8 OLBs per snap. Also, to account for special teams, I am going to pretend that there are 0.3 special team players per snap: that’s a fiction, obviously, but I needed to use some number for the following purpose.

The next graph then shows the Draft Value used per position divided by the number of players on the field per snap. This, I think, is the best way to isolate how teams valued positions in the draft. Take a look:

val per pos

Now quarterback vaults to the top of the list, but only barely. And while cornerback was number one before, it’s been surpassed by both defensive ends and defensive tackles. There were talks before the draft that this was a very good one for defensive tackles, and the numbers here bear that out. Only a quarterback-needy market that drove the top prospects to become the first two picks prevented defensive tackles topping the charts in draft value spent per snap.

What do you think?

  • Carl Yedor

    I found it interesting that the value spent per position (per snap) was so high. I imagine that Elliott going 4th had a lot to do with that. What would the graph have looked like if Elliott went in the 10-15 range and everyone below him slid up? Obviously this is imperfect because the assumption that every player between 5 and 10ish would automatically slide up is not a good one, but I’m curious to see how big of a difference it would have made.

    • Agreed.

    • Dave Archibald

      The other factor with RB is that even though there is only one on the field, teams still carry three to five on the roster. The attrition and injury rate is high, careers are short, and the position is grueling enough that even the top players don’t play more than 75-80% of snaps. Most teams rotate situationally and so some top backs only play half the snaps. And the position is specialized enough that teams can’t just plug in a WR or TE to substitute for a RB. Contrast to OL, where careers are longer, starters almost never come off the field, and one player can back up multiple positions. Even though a team will have five times as many OL as RB on the field most plays, it will generally only have twice as many OL on the roster. We shouldn’t be surprised teams continually draft RBs. The same effect is present at DT, another position that is subbed out frequently.

      • While true, that should really only impact the # of players drafted, not the draft value. I agree with what you are saying, but given that there is so much rotating going on at RB, that would seem to encourage teams to use lots of 4th, 5th, and 6th round picks on RBs…. not 1st and 2nds.

        • Dave Archibald

          I could see it affecting both – if teams expect a 3rd RB to play more than a 7th OL, they might value that higher and draft it higher. I guess you can argue that the 1st RB isn’t going to play as much as the starting C and should be valued less. Right now, there seems to be a significant spread in how teams value RBs. Other than Elliott, Derrick Henry was the only RB drafted in the first 72 picks this year.

  • Do you have this last chart/data for prior drafts as well?

    • No, but putting that together is on my to-do list.

  • mrh

    How can there be 1.05 centers per snap? Are teams putting 2 centers on the field in case one gets banished from the face-off circle?

    • Good question! In 2014, the Giants had Weston Richburg and J.D. Walton play a ton, and both are listed as Centers for their primary positions. Obviously Richburg played G, though. Similarly, for the 2012 Patriots, Dan Connelley and Ryan Wendell were both listed as Centers. The 2014 Dolphins had this too (Satele, Pouncey) and a few other teams.

      So it’s sort of a database error, but sort of not. I mean, if teams are drafting centers because they can also play guard, then I think it makes sense to consider 1.05 as the appropriate number in this context. Practically, of course, 1.00 and 1.05 are the same for these purposes.

    • JeremyDeShetler

      Extra center on empty net situations.

  • Richie

    ” Based on PFR data, the average offensive snap last year”

    Is snap data publicly available on PFR? I can’t find it.

    • I don’t think it is.

    • Dave Archibald

      It’s in game box scores and on team pages, but not on league pages or individual player pages.

      • Richie

        Nice, now I see it on the box score. But I still don’t see individual snap totals on team pages. Can you link to an example?

        • It is there, actually:

          For players, it’s under fantasy (but it works for offensive lineman, too):


          On the team page, you need to go to More+ (after the Injury Report) and then you should see it:


          • Richie


            I could have sworn I looked under the “More” tab, but didn’t see anything. I can definitely see it now.

            It’s really one of the crazy things about football stats. In baseball, we spend a lot of time looking at plate appearances, innings pitched, batters faced, etc. But I have no idea what a typical number of “snaps” is for, say, the leading receivers. Understanding opportunities that players get is important, but it’s rarely discussed.

            • Richie

              JJ Watt (21 AV) and Aaron Donald (20 AV) were the top 2 players in AV last year. Watt had 1078 snaps. Donald had 991.

              So Donald was slightly more efficient than Watt. Donald gained a point of AV for every 49.5 snaps, while Watt needed 51.3 snaps.

              • I don’t think that’s an appropriate way to measure things, though, because I don’t think AV looks at snaps. I think the only way it measures efficiency is in terms of games played. But AV would have no clue if Watt or Donald played a higher percentage of their team’s snaps, once you get past the filter that both played enough to receive significant postseason recognition.

                • Richie

                  I understand. My point was just that Donald accumulated almost as much AV as Watt did, but he did it with 8% less time on the playing field.

                  • I know we are on the same page, but I do think it’s important to provide clarity for other people.

                    Donald had nearly the same AV as Watt, but that’s based on (I think) 4 things:

                    Both being 1APs
                    No other player on either defense being a Pro Bowler or All-Pro
                    HOU/STL allowing almost identical points per drive numbers
                    Watt having more sacks

                    So, what you’re really saying is it’s impressive that Donald was a 1AP despite spending less 8% time on the field than Watt. But that’s not really a thing. I mean, Watt got 50 votes, so there wasn’t more he could do 🙂 Anyway, I apologize for being overly pedantic here, but I know people are always curious about AV.

                    • Richie

                      Ah, I see what you’re saying. Yeah, the 1st Team All Pro bonus is so huge for defensive players that it is a huge factor in the AV.

                      I did a quick calculation, and it looks like the 1AP was worth 11 of Watt’s 21 AV. (If Watt didn’t get any All Pro honors at all, and everything else stayed the same, his AV would have been 10.)

                      When I was first comparing AV to Snaps, I thought there might be some circular entanglement issues.