Worth the price?
The so-called Jimmy Johnson Trade Value Chart (the “Johnson Chart”) has been around for a while now and is still considered the standard in determining trade consideration, at least by the media. The chart was even posted on Kevin Colbert’s bulletin board in the movie “Draft Day”! There is a belief (definitely overstated) that the Johnson Chart has no objective evidence to support it and is based solely on the experience of Johnson and his associates.
Not everyone regards Johnson Chart as the bible or the most appropriate guide. I read recently that at least some NFL teams, including the 49ers, have developed their own charts to evaluate trades. In addition, there have been any number of attempts by academics and others to produce Trade Value Charts (“Charts”) that are based on sound statistical principles.
Probably the best known of the alternative Charts is the one published by Kevin Meers in the Harvard Sports Analysis Collective (the “Meers Chart”). The Meers Chart relied heavily on the “Career Approximate Value” as published at www.pro-football-reference.com and contains a heavy dose of statistical analysis.
Other than the real or imagined methodologies, the principal difference between the two Charts is that the Meers Chart places a significantly higher value on mid to late round selections in relation to early picks. For example, the 100th pick in the draft is worth about 3% of the 1st pick under the Johnson Chart and about 13% of the 1st pick under the Meers Chart.
Meers view of the Johnson Chart is succinctly stated in that same Harvard article: “The old system massively over value the earliest picks and significantly undervalues mid-to-late round picks.” I tend to agree with Meers on this point but, for reasons that will be discussed below, believe that the difference is somewhat misleading.
In my view, a discussion of the Charts has to recognize that this is no different than any other asset purchase. There are two elements of value in any transaction - - real value and acquisition premium. Acquisition premium is defined as the difference between the actual price paid for something and its real value.
I cannot find anything in published articles that indicates that either the Johnson Chart or the Meer Chart explicitly considers acquisition premium. (I did a google search looking for instances where both acquisition premium and trade value chart were both used in a document and no instances were found.) It is reasonable to assume, though, that the Johnson Chart implicitly (and probably not scientifically) considered acquisition premium. The nature of the Meers calcuations suggests that acquisition premium was not considered, though someone with better analytic abilities than me may be able to find a proxy for acquisition premium in their calculation.
A comparison of the Johnson Chart and the Meers Chart is meaningless if one includes acquisition premium and the other doesn’t.
The last thing the pro football world needs right now is another Chart, but I think my approach is different and transparent enough to be worth presenting. In the remainder of this article I will describe how I would calculate both the real value and the acquisition premium in the context of NFL trades.
I calculated real value by considering and weighting five metrics as follows:
• Five-year starter = 40%
• Pro Bowl at least once = 20%
• All Pro at least once = 20%
• Two-year starter = 10%
• Five-year career = 10%
These weightings are, of course, very subjective and strictly a matter of opinion. I developed a point system based on this analysis. Using these weightings resulted in the following real values for each of the Draft Choice Ranges (with a value of 1000 arbitrarily assigned to selections 1-13):
• Selections 1-13 = 1000
• Selections 14-24 = 839
• Selections 25-46 = 651
• Selections 47-73 = 488
• Selections 74-114 = 296
• Selections 115-187 = 187
• Selections 188+ = 112
It only makes sense that someone who agrees to move down in the draft would be compensated for that move even if two things of equal value are being exchanged. Why else would the potential seller agree to sell his asset?
In developing an acquisition premium methodology I reviewed each of the 57 transactions during the past three seasons that included only current year draft choices (e.g., no players or future year draft choices are involved).
It seems logical that the more spots a team moves down in the draft, the greater the acquisition premium should be. To that end, I calculated the acquisition premium per draft slot for each of the 57 transactions. The acquisition premiums paid in those transactions can be summarized as follows:
• Less than 2% = 13
• Between 2% and 6% = 21
• Between 6% and 10% = 15
• Over 10% = 8
Further, the average weighted premium was 3.8% per draft slot. On an unweighted basis the average was 6.1% per draft slot. Based on this analysis, my methodology would be to include a premium of 5% per draft slot.
Applying the Two-Step Approach
While my approach involves two steps rather than a single Chart, I believe it is much simpler and more transparent than the other alternatives. First, consistent with my approach with Draft Choice Ranges, there would be only seven potential real values to consider. Second, the application of the acquisition premium is a simple mathematical exercise.
A few examples illustrate my proposal:
• A team moves from draft selection #8 to draft selection #1
- There is no difference in the real value between the exchanged slots as both are in the first Draft Choice Range and therefore are of equal value
- An acquisition premium of 35% applies because the team moved seven slots
-1350 (1000 plus 1000 x 35%) points would have to be surrendered for a draft choice worth 1000 points
• A team moves from draft selection #14 to draft selection #13
- The difference in real value is 161 points (1000 – 839)
- The acquisition premium is 5% of 1000, or 50 points
- Total consideration of 1050 points would be surrendered for a draft choice worth 1000 points
• A team moves from draft selections 39 to draft selection 22
- The difference in real value is 188 points (839-651)
- The acquisition premium is 85% (17 slots at 5% per slot) of 839 points, or 713 points
- Total consideration of 1552 (651 + 188 + 713) points would be consideration for a selection worth 839 points
Follow Tony on Twitter @draftmetrics