clock menu more-arrow no yes

Filed under:

What draft position tells us about finding Pro Bowlers and regular contributors

New, comments

A snapshot from 2017 NFL rosters reveals some pretty interesting findings about how the quantity and quality of players are sometimes clustered together by draft position.

2018 NFL Draft Photo by Tom Pennington/Getty Images

The Panthers rookie mini-camp is now in the rear-view mirror and nobody knows how good – or bad – any of Carolina’s new selections will turn out to be. While we all know that on average the higher players are picked the better their careers tend to be, I got curious to see what research exists showing which players actually get regular playing time based on draft position.

Depth, after all, is incredibly important in a league with a motto of “next man up.” Unfortunately, I couldn’t find any good data so I spent several hours pulling it together by myself. As an inquisitive NFL data nerd, I actually kind of like doing stuff like this. I figured since I’ve already done the work I might as well share it with my CSR family. It’s a helpful analysis, though not perfect by any means, and here’s what I did:

First, I found all “regular contributors” who played in at least 40-percent of their teams’ offensive or defensive snaps in 2017 using data from Pro Football Focus. With as cut-throat as the NFL is, any player who is on the field for at least 40-percent of his team’s offensive/defensive snaps should at minimum be considered “good enough to play.”

Second, I segmented each of these regular contributors into groups consisting of 10 consecutive picks such as Nos. 1-10, 11-20, 21-30, etc. This segmentation is helpful because the same number of players are drafted in each group every season – 10 draftees per group – so the sample size entering the NFL is consistent year over year. Based on the results (data set here), this is where every 2017 offensive and defensive regular contributor was drafted:

Here are a few takeaways:

  • Top 20 picks dominate the distribution of 2017 regular contributors, but there is virtually no difference between the number of picks from No. 1-10 (73 players) and No. 11-20 (72 players).
  • Picks 21-40 representation is noticeably less than Top-20 picks, but substantially higher than picks 41 or later.
  • Picks 41-100 have a fairly even distribution by draft group despite spanning 60 picks.
  • Picks 101-150 have a fairly even distribution by draft group as well despite spanning 50 picks.
  • Picks 151-180 have a fairly even distribution while picks after 180 represented a very small percentage of regular contributors in 2017.

Now, the big limitation to this data is it only tells us which players are “good enough to play” but not exactly how good they are. To account at least somewhat for player performance, let’s also look at where the 2017 Pro Bowl selections were drafted. Since we are looking at draft position of players who played at least 40-percent of offensive/defensive snaps, I’ve excluded special teams, fullbacks, and undrafted 2017 Pro Bowlers.

In all there were 641 players who met this criteria in 2017 and 101 of them were selected to the Pro Bowl, per Football Reference. Here is the draft position of the “elite” 15.7-percent (101 of 641) of the NFL who made the 2017 Pro Bowl:

When looking at the two charts together (snap counts and Pro Bowls), here is what I’m left to conclude when it comes to finding quality and quantity via the draft along with the implications for draft day trade strategies:

  • Top 20 picks – The most realistic way to improve roster quality with Pro Bowlers is to nail your Top 20 picks.
  • Picks 21-40 – There is virtually no difference in the quantity or quality of players who were drafted between 21-30 (51 regular contributors, 10 Pro Bowlers) and those drafted between 31-40 (52 regular contributors, nine Pro Bowlers). Therefore, teams have sacrificed very little by trading down from the late first round into the early second round.
  • Picks 41-100 – Picks within the six different groups making up Nos. 41-100 (e.g. 41-50, 51-60, etc.) have almost equal representation among regular contributors, but there are slightly more Pro Bowl selections among earlier picks. Therefore, teams have sacrificed very little to trade back from the middle of the second round into the third round.
  • Picks 101-150 – Pro Bowlers are virtually non-existent at this point so the focus is simply finding players who can be on the field for at least 40-percent of offensive/defensive snaps. Once again, there is very little difference in the number of 2017’s regular contributors drafted between these five different groups (e.g. 101-110, 111-120, etc.). Therefore, teams have sacrificed very little by trading back from the early fourth round into the fifth round to acquire additional picks.
  • Picks 151 and later – Not a lot of quality or quantity here. Teams might as well be playing roulette.

Those of you who have read my stuff before know I’m a huge advocate for trading down in the draft. When I wrote the Constitution for Trade Down Island it was driven largely by my observations that drafting players is unpredictable, so more draft picks means more chances at landing starters and regular contributors.

Now after analyzing data from the 2017 season, I’m even more convinced that there is far more upside than risk in trading down a few spots to pick up additional picks. For example, there is little difference in the quantity and quality of players drafted between No. 21-30 and No. 31-40. Trading down from the late first round into the early second would therefore make sense. Same goes for No. 41-60 and No. 61-80 and a few other clusters.

However, based on this data, in the future I’m going to be much more supportive of holding on to Top 20 draft picks because that’s clearly where the Pro Bowlers are found.

Now, there’s nothing revolutionary in this data and you can draw your own conclusions. However, the main “a-ha!” moment for me was to see how there were clusters of similar groups rather than just a gradual, smooth curve from high draft picks to low. In particular, I was surprised to see the clustering of similar results between picks No. 41-100 followed by the clear drop-off starting at No. 101.

This data sheds some additional light in finding roster depth (quantity) instead of just Pro Bowlers (quality). Let’s now hope the Panthers 2018 picks provide a great deal of both quality and quantity for years to come.


What data did you find most interesting?

This poll is closed

  • 9%
    The quantity of Pro Bowlers drafted in the Top 20
    (25 votes)
  • 29%
    There is virtually no difference between picks 21-30 and 31-40
    (79 votes)
  • 33%
    How little difference there is between each of the 10-pick groups spanning from picks 41-100
    (89 votes)
  • 13%
    The vast drop-off between picks 91-100 and 101-plus
    (36 votes)
  • 14%
    Nothing really interesting here - I intuitively knew all of this
    (38 votes)
267 votes total Vote Now