Draft or free agency? A look at the economics of roster construction

Building an NFL roster is no easy feat.

That statement probably goes without saying, but there are layers upon layers of moving pieces at every stage of roster construction. It’s something like trying to put together a jigsaw puzzle with the pieces all face down, only a limited number of pieces the general manager can use, and 32 general managers all competing for pieces out of a common pool. All while the puzzle box itself tries to push the general managers to assemble the same mediocre picture.

At times it seems like a minor miracle that any roster gets assembled at all, let alone multiple competitive rosters.

There has long been a sense of economics in roster construction, as teams all need to make the best use of the various pieces available to them. Jason Fitzgerald of OverTheCap has been looking into the practical economics of roster construction, attempting to look at real world outcomes to free agency, the draft, and how the two interact with each other.

Fitzgerald framed this piece of his work as such, “In general there are two prime movers of talent in the NFL- the draft and free agency. Free agency requires salary cap space and a big budget while draft picks require the draft capital to have the chance to draft top players. But I think too often we lose sight of the fact that free agency should impact the draft and the draft should also impact free agency.”

Fitzgerald compared the value of drafting players of each position compared with the value of signing them in free agency. The product of this comparison looks like this:

The X axis is the difference between the average annual salary of the 16th draft pick and the 10th highest paid player at the position. This illustrates how much we gain in our salary cap bucket if we hit in the draft, money which in turn can go toward free agent acquisitions. The Y axis shows the percentage of players who currently rank in the top 20 by salary who were initially acquired in free agency (for example a player like Kirk Cousins would still be considered a free agent acquisition even though he is now on an extension).

Speaking generally, Fitzgerald found that there is better value in signing positions in the upper left quadrant in free agency, while positions in the lower right quadrant are best acquired in the draft.

The quarterback position stands out as an extreme outlier in value. And taking a step back, that’s a fact we all understand. It’s well established that having a starting quarterback on a rookie contract is an incredible advantage for teams. Fitzgerald conceptualizes it by stating that if a team is able to draft a QB who is capable of winning games for a team (not just one with whom the team is able to win with), is worth signing a cornerback and a linebacker in free agency.

In general, Fitzgerald posits, it’s better to draft quarterbacks, wide receivers, offensive tackles*, and EDGE defenders highly. Meanwhile, tight ends, linebackers, safeties, running backs, and interior offensive linemen don’t give the same kind of salary cap value as the “premier” positions and are better acquired in free agency or later on in the draft. Basically, there isn’t the same gulf between rookie contracts and veteran contracts at those positions, so there’s less motivation to prioritize those positions in the draft.

Raptor’s thoughts

What Fitzgerald showed is something we have known — at least intuitively — for a while now. The concept of “cornerstone” or “premium” positions is a fairly common one when talking about team building strategies.

However, by putting discrete values on the various positions and how much hits on those positions are worth could help an economically (and likely analytically minded) team make decisions in prioritizing positions on their big board. After all, rookies (now) have set contract values, and we no longer have situations where a rookie QB can re-set the market for veterans before he even takes a snap.

There are two interesting caveats that I think we need to take a close look at.

The first is the right tackle position. Fitzgerald lists right tackles as a poor choice to target highly in the draft and likely better left to free agency. Personally, I think this is a conclusion that will change pretty rapidly in the coming years and is influenced by the momentum of some older notions. It used to be that the right tackle position was subordinate to the left tackle, based on the logic that defenses would line their best pass rushers up on the defensive right to attack a (right handed) quarterback’s blindside.

In practice, particularly with the spread of “multiple” defenses, we’ve seen more and more teams come to the realization that both offensive tackles need elite traits and skills. After all, if teams are sending waves of athletic edge rushers and blitzers at both sides of the quarterback — and data has shown that there’s basically no difference in outcome between rushes from the left vs. right sides — why wouldn’t a team want to invest in both tackles? I stopped differentiating between left and right tackles a couple years ago, and I think the differences in contracts and draft positions between the two positions will shrink in short order. There might not be a difference between left and right tackles by the time Jedrick Wills and Tristan Wirfs hit free agency.

The other position is cornerback. Fitzgerald has this to say about the position, “Cornerback is one of the most interesting positions. It is a pretty big cap benefit but also a position that has a high level of turnover in free agency.”

Cornerback is somewhat complicated by the use of sub-packages and the slot corner position. Slot corners are important and virtual starters in just about every NFL defense. It likely makes good economic sense to draft starting outside corners highly, as they generally command massive contracts on the free agent market. Slot corners, however, can generally be had more cheaply in free agency. Likewise, teams need a lot of cornerbacks on their roster. Not just because of the natural attrition and injury risk at the position, but also because they are useful as special teams players. That back-end roster churn likely creates a lot of the turnover Fitzgerald saw when looking at the position.

There is, of course, still more work to be done parsing out the relative value and impact for every position. However, the work of Fitzgerald (and likely economists, as well as analytics offices, in front offices around the NFL) could help teams as they approach free agency. Every year Ed Valentine puts forth his “Rules For Draft Success,” and work like this could help provide something similar for teams around the league — at least the ones into logically maximizing every advantage along the way as they build their rosters.

As Fitzgerald points out at the beginning of his piece, the Draft and free agency are interconnected and play off each other. The calculus of maximizing a team’s resources in each phase should be at the front of their minds throughout the offseason.