We all know the NFL draft can be an absolute mess. General managers and coaches make and break their careers based on the players they select. Can you avoid picking Jamarcus Russell first overall? Can you find the Day 3 gem in Tom Brady or Shannon Sharpe? Player evaluations are not my strong suit. People far more knowledgeable than me watch thousands of hours of tape on these prospects, and yet still about half of all first-round picks do not live up to expectations.
Inspired by the 49ers trading up from the 12th-overall pick to the 3rd-overall in this years' draft to address their quarterback problem, I wanted to see the relative value of these draft picks to see whether trading up would be an overpay or not. Using the past history of draft-day trades, I would be able to do this, but I realized that most teams use draft value charts. The most popular of which is Jimmy Johnson's.
Rather something more interesting to look at is whether this draft value chart matches up with the value of the players selected at those picks. In other words, does the value of a player selected with a higher pick make up for the value of getting a greater number of lesser quality players?
I took Pro-Football-Reference's Approximate Value (AV) as a starting point one-number metric to evaluate the players picked at each position from the 2000 to 2015 NFL drafts. Obviously, there are a decent amount of problems or areas for improvement within the original metric itself, but generally the players are valued appropriately, so for my purposes, it will be perfectly fine.
Below is a logarthmic regression of the AV's of the first 224 picks along with the Jimmy Johnson values. The y-axis is scaled down such that the first pick is given an arbitrary value of 100:
The first glaringly obvious realization is that the later picks are undervalued by Jimmy Johnson's system. Or instead, the first few picks are heavily overvalued. As the draft value chart represents what a team is willing to pay to move up into those spots. Teams are often overconfident in their own evaluations, overpaying to "get their guy" high up in the draft.
This initial conclusion doesn't bode well for the 49ers: trading down seems to be the optimal strategy rather than trading up. However, this effect may not be as pronounced as the graph shows, due to upside value. If you have general managers in the league the option of getting two 4/10 players or one 7/10 player and one 1/10 player, every single one will prefer taking the 7/10 player. My regression model treats these two scenarios as equal.
In the NFL, teams don't have to undergo the downside of picking a bad player. If the player is no good, they won't get playing time and teams will be able to fill the role with a replacement-level player either from a later pick in the draft or through free agency. The only "cost" here is the opportunity cost of a better player taken by another team with a later pick.
Undoubtedly as the draft buzz heats up over the next week until the Jaguars are on the clock, we will see far more trades down the board. After the draft, I'll release another article documenting the trades and seeing the winners and losers of each exchange.
Comments
Post a Comment