clock menu more-arrow no yes

Filed under:

What’s the real value of the first overall NFL Draft pick?

New, comments

How Many Players is a #1 Overall Pick Worth?

Detroit Lions vs. Los Angeles Rams Photo by George Rose/Getty Images

Jared Goff wasn’t the first Cal QB to be taken 1st Overall. That honor fell on the man in the picture above: Steve Bartkowski, selected 1st Overall by the Atlanta Falcons in the 1975 NFL Draft. Every year there is a debate regarding who will be selected 1st Overall. Due to my training as an Economist from Cal (all hail Evans Hall), I always looked at it in a relative value. That is, how many other players in the NFL Draft would it take to match the value of a #1 Pick.


The Cleveland Browns most recently began trading away their top draft picks for more later in the Draft. In the last two drafts I wrote about value of draft-picks in a “Pick for Pick” manner. In this article I will explore the ex-post value of the #1 pick. The question I ask here is :

In terms of AV, How Many Players From the Same Draft Would it Take to Match the #1 Overall Pick’s AV?

Assumptions and Considerations

It assumes that the AV values stack in a linear manner (ie. Two players with AVs of 5 each = One player with AV of 10). This simplifying assumption allows us to compare data easily. Another model assumption: I am only comparing player’s AVs that they accrued with the team that drafted them, this allows us to restrict the analysis to the realm of “value for a #1 Pick” and not “value of the Player”. For example, Alex Smith’s AV used in this comparison is only from his time as a 49er and not as a Chief.

Another caveat: Pro-Football-Reference assigns 0 and NAs differently, 0s are assigned to players who have played but contributed 0 AVs. NAs are given out to players who have not played. I will show two visualizations, one with and without NAs.

Example #1: With NAs

Click on the images to see them in full size!

This shows with NAs, that is players who were drafted but did not play.

How to analyze this chart?

Let’s look at Peyton Manning and the value next to his name: 97.925%. This means that one would need the AV of nearly 98% of the players in the to match his contributions to the Colts. With this chart at hand one would have to offer the Colts, for the #1 Overall Pick, 98% of the picks in the draft to match Manning’s production.

Across the spectrum we can see that even JaMarcus Russell, the biggest bust of busts himself, still played better for the Raider’s than nearly 48% of the players drafted in that year would’ve combined.

If you look closely, and chances are it was the first name you looked at, Jared Goff is doing the worst of them all. Even with NAs, why? Because his poor performance in 2016 netted him a -2 AV. He is by this measure a worse than having no-one play. This number will change in the future. But this illustrates the fact that Jared did so poorly that nearly any player from the 2016 Draft would’ve been a plus to the Rams (holding all else constant).

Example #2: Without NAs

This shows without NAs, that is players who were drafted but did not play.

Now we exclude all players that didn’t step unto the field. Thus we are comparing players that actually played. Often times the players with NAs as AVs are players someone took a flyer on in the 6th-7th, borderline UDFAs and thus it would be unfair to comp them to players who weren’t truly expected to contribute. Thus we will look at the percentage of productive players it took to match the AV of a #1 pick.

What we see here is a drop off for everyone, including Peyton Manning, although at varying magnitudes. The shaded grad area depicts the difference between Visualization #1 and #2. This shoes that most of the low performing players had most of their seemingly superior performance bolstered by the NAs. Biggest beneficiaries were JaMarcus Russell, Ki-Jana Carter, and Dan Wilkinson.

We can see here that despite the adjustment 6 of the 22 #1 overall picks graded at 90% of their total class with the number growing to 9 when we lower the threshold to 80%.


From a bird’s eye view of the data-set we can see that even with the biggest draft bust of the modern era (sans Goff due to his 1 year abbreviated stint) still has a much production as a linear aggregate of 47%-35% (depending on the counting method) of all the draft picks of their respective drafts.

This contradicts my initial thesis about the lower than expected value of a high pick (see previous draft related posts), but it does show that more often than not a team will get a player that for the team will contribute more than half of the draft combined.

Go Bears!