First is worst in NFL Draft

Reading Time: 5 minutes

 

Starting from the Top

The objective in the NFL is to win Super Bowls.  In order to win Super Bowls, you have to win significantly more games than you lose, in most seasons.  Winning games requires scoring more points than you surrender to the opponent.  Outscoring your opponent requires some combination of better personnel, strategy, and execution (which I think of as the interaction between personnel and strategy).  With the annual NFL Draft approaching, this piece focuses on the personnel part of the equation.  

“If they want you to cook the dinner, at least they ought to let you shop for some of the groceries.” – Bill Parcells

Assembling a team, in a very simplistic way, is no different than many business decisions.  It’s all capital allocation at the end of the day.  A GM has a finite set of resources (salary cap) to allocate as efficiently as possible to build a sustainably competitive team.  Maximize talent per dollar spent.  Adding talent to your roster comes from two distinct markets: free agency and the draft.

The draft has become a spectacle.  In recent years the league has switched from a 2 day, weekend-long marathon to a 3 day, prime-time television event.  Presumably to increase revenue.  Aside from the pomp and circumstance, the objective for personnel departments remains the same.  Gain a competitive edge with better players.

Can GMs Forecast?

In order to execute on this, GMs must accurately forecast the performance of the players entering the draft and avoid overpaying for that talent.  Collectively, NFL GMs do a good job of selecting talent at the appropriate point in the draft.  Looking at the average value of a player vs. the round they were drafted in, there’s a clear decline in experienced value as rounds pass (source: fivethirtyeight, pro-football-reference.com).  The draft, in aggregate, is an efficient system.

However, as you can see from the abundance of dots both below and above the red line, there are lots of times when a given player outperforms his draft location and underperforms.  This begs the question, is any team better than their peers at drafting players that outperform the value set by draft position? Based on this article from fiverthirtyeight.com, the answer is “not really”.  There is virtually no correlation between draft performance from one year to the next.  Over 3 year periods there is a modest positive correlation (ie: success is mildly persistent).

“The implication is that much of what each team gets from its draft picks is determined by pure chance” – Neil Paine, fivethirtyeight.com

This should come as no surprise.  As mentioned in this previous post, experts are not accurate forecasters.

An Alternative Route

So, the conclusion seems to be that individual teams and GMs cannot forecast.  At least not reliably enough to claim it as a consistent and repeatable competitive edge.  If you accept this as fact, then the only way to reliably reduce the number of times you overpay for a given player is to have a strict policy of conservative contracts.  Similar to the value investing philosophy popularized by Warren Buffet.  While the players a given team selects do not receive competing bids from other teams, overpayment comes in the form of drafting a player too early.  This violates the conservatism constraint.

So what’s a GM to do? If you can’t reliably forecast player value/ability and have to pay them according to where they were drafted then how do you go about building your roster intelligently?  One strategy is to let the draft come to you and bias your selections toward taking the best player available when it’s your team’s turn to make a selection.  But there’s an additional step one could take.

In 2013, the most recent economics Nobel laureate Richard Thaler published a paper with Cade Massey studying this exact topic.  Their goal was to study the predictions of market participants with large sums of money at stake from a behavioral economics perspective.  Without getting deep into the weeds on their methodology, their conclusions suggest that GMs could improve their drafting strategy in a counterintuitive way.

Instead of focusing as heavily as they do on “who” to draft they should try to better position themselves in terms of “where” they draft.  In other words, certain rounds offer more talent per dollar spent than others, on average.  This is an outgrowth of the differences in the talent curve shown above (which is technically a slightly different calculation) and the corresponding compensation curve draftees are paid according to the CBA.  Looking at the ratio of player value to salary across all picks in the draft produces the following yellow curve.

Based on this analysis, Thaler and Massey suggest the best value in the draft is located in the late first round and through the second round.  This is the sweet spot of the draft from a valuation perspective.  Potentially shocking to some, the most counterintuitive conclusion they draw; the first pick in the NFL Draft is the worst position in the first round.

“Our findings are strikingly strong.  Rather than a treasure, the right to pick first appears to be a curse.”  Thaler & Massey (2013)

Look at any list of top NFL draft busts, there is one common theme.  They were predominantly drafted in the top half of the first round.  They were players that created huge opportunity costs for their teams by soaking up disproportionate sums of the salary cap without commensurate performance.  The top half of the first round has an assymetric risk to the downside given the dollars involved and expectations set for the players drafted.

So what’s the play?

You’re John Dorsey GM for the Cleveland Browns.  You have the 1st and 4th picks in the 2018 NFL Draft.  Your team is perpetually finishing in the bottom of the league in win percentage.  The trajectory of the Browns franchise rests on your decisions.  What do you do?

Do you accept that you and your staff’s ability to forecast player ability is limited at best?  Do you buy the fact that the first 15 picks have more downside than upside?  If you do, then you try to trade down and accumulate more picks in that sweet spot in the draft.  This has 2 benefits.  A) you avoid the downside risk associated with picking in the top 15 and B) you accumulate more chances of adding talent with future value to your roster.  If I’m John Dorsey, I’m trying to figure out who the most desperate GM in the league is and selling him at least one of my top picks.

This is easier said than done.  Trading away the perceived value of top draft picks puts you directly in the crosshairs of public and media scrutiny.  In the investment business making the unpopular decision, where you stand out from the consensus opinion, is commonly referred to as career risk.  It’s no different in this scenario.  People are biased toward sticking with the crowd.  This herding behavior is what causes market bubbles in stocks and draft picks.

I understand the top pick provides you the opportunity to choose anyone available.  There’s no worrying about missing out on that player you’ve fallen in love with.  But if you are taking this point of view, you haven’t accepted the fact that you can’t accurately project who the best player will be. There are multiple behavioral biases that make this hard to accept, one being the illusion of control.

We shall see…

Ultimately, this will all play out in a couple weeks when the first round of the NFL Draft kicks off on April 27th.  The Browns kept their top pick last year, but that was under a different general manager (Sashi Brown, a lawyer by way a background).  The most intriguing common element between last year and this year is a man named Paul DePodesta, the Browns Cheif Strategy Officer.

DePodesta is one of the central figures in the movie Moneyball, played by Jonah Hill.  A driving force behind the changes in professional baseball roster management.  He took what was largely an expert-driven system reliant on intuition and turned it into one of empiricism and probabilities derived from first principles thinking.

I’m interested to see if it translates to football.