Here at PFF, we’ve been abundantly clear about our thoughts surrounding NFL draft trade values. Between Timo Riske’s work creating a draft trade chart using PFF Wins Above Replacement and a guest column from upshot student Zach Drapkin and Wharton professor Cade Massey, one of the first researchers ever to broach the subject of draft pick values, we’ve exhaustively discussed the inefficiencies associated with the Jimmy Johnson draft value chart.
Over the three decades since former Dallas Cowboys executive Mike McCoy created the chart — which he later described as a “rinky-dink concept” — NFL teams have agreed to do their best to make all draft pick trades as close to “even” as possible according to the chart. And while some front offices have rejected the use of this particular method, the results speak for themselves.
The biggest issue with the Johnson chart from a salary cap analyst’s perspective is that the implementation of the rookie wage scale in 2011 drastically changed the value of each pick. Yet, the chart’s prevalence has remained relatively unchanged.
No. 1 overall pick Sam Bradford’s rookie contract, signed in 2010, was a five-year, $78 million deal ($12 million per year) with $50 million in total guarantees. The following year's No. 1 overall pick, Cam Newton, signed a four-year, $22 million rookie deal ($5.5 million per year) with the full $22 million guaranteed — less than half of Bradford’s guarantee from a year earlier.
In what world does it make sense to value 2010 draft picks the same as picks from 2011 and on?
From 2011-2020, the 278 pick-for-pick trade packages fell within 26.4 points on the Jimmy Johnson chart on average. These 26.4 points represent the No. 162 pick, which falls in the middle of the fifth round. If you remove just five of the largest trades over the decade (big trade-ups for quarterbacks Robert Griffin III, Jared Goff, Carson Wentz, Sam Darnold and Josh Allen), the average difference in value between the trade-up team package and trade-down team package was 19.4 points. These 19.4 points represent the No. 179 pick, which falls at the end of the fifth round).
Here is a quick example to illustrate what we mean:
|Team A||Team B||Value Differential per Johnson Chart|
10Johnson value: 1,300
|Receives Pick 12, 84
Johnson value: 1,370
Team B = +70
70 on Johnson chart = Pick 112
Effectively, one of the teams in the trade is paying a premium of a mid-to-late fifth-round pick to move up or down on average. For 165 of the 278 trades, the differential was 11 Johnson points or less, meaning that teams paid a premium equivalent to Pick No. 200 at most.
Let's look at the same information from a percentage standpoint. If we were to break down the trade packages by comparing the total values of each, the average* surplus value paid by the “losing” team was 8.97%. Using the same example as above: 1,370/1,300 = 5.38%. Team A paid an additional 5.38% in total compensation to move up to pick No. 10.
*11 outlier trades of the 278 total had to be removed wherein the percentages were wildly different because of the Johnson chart’s sharp decline in the 200s.
With the Johnson chart still generally controlling the marketplace — though there is evidence that the Rich Hill chart, a similarly constructed chart, has become even more accurate in recent years — it becomes a challenge to test whether NFL teams have started to approach things differently.
When proposed trades have to be close to “even” or zero-sum per the Johnson chart, there won’t be much variation. The way intelligent teams find edges in the draft trade marketplace is by coming up with trade packages that are “even” per Jimmy Johnson but favor themselves according to charts based on actual data.
Two draft value charts that have now been publicly available since the 2020 NFL Draft are the Harvard Sports Analysis Collective chart and the Fitzgerald-Spielberger chart from Over The Cap. According to Benjamin Allbright of KOA Colorado, at least one team uses all three charts in an effort to gain an edge. As illustrated below, the Harvard SAC chart and Fitzgerald-Spielberger chart are very similar.
Below is a table illustrating the steepness of draft value charts based on the value of later picks as a percent of the No. 1 pick:
|Pick No.||Jimmy Johnson||Harvard SAC||Fitzgerald-Spielberger|
The Johnson chart values the final pick of the first round as equivalent to 20% of the No. 1 overall pick. The HSAC and Fitzgerald-Spielberger charts suggest that it is closer to 40%.
If we run the same analysis as above from 2011-2020 but instead use the Fitzgerald-Spielberger chart, we see the 278 pick-for-pick trade packages fell within 435 points on the Fitzgerald-Spielberger chart on average. Those 435 points represent the No. 158 pick, a mid-fifth round selection. If you remove the aforementioned five largest trades over the decade (Griffin III, Goff, Wentz, Darnold and Allen), the average difference in value between the trade-up team package and trade-down team package was 403 points, or the No. 168 pick (middle-to-end of the fifth round).
While the two charts have similar averages over the entire draft, they arrive there from very different starting spots (Pick No. 260 below means the trades were “even,” or rather the surplus price paid was less than any pick on the Fitzgerald-Spielberger chart).