by Tony Villiotti
March 07, 02013
The NFL has had seven rounds in its draft since 1994. The teams with the worst record in the preceding year draftsfirst and the Super Bowl champion has the last pick in each round. It has been ingrained in our thinking that the value of the draft choices is perfectly logical with the first choice being worth more than the second choice,the second being more valuable than the third, all the way down to the end of the draft.
DRAFTMETRICS, though, has consistently held the belief that the draft can be divided into several groupings of selections within which all the draft selections have an equal chance of success. (These groups have been called Value Groups in past DRAFTMETRICS writings and will be referred to as such in this article.) As discussed below, this means a player selected with the 13th pick has the same chance of success as the player selected first.
In its analysis of drafts from 1993 through 2012, DRAFTMETRICS has concluded that there are indeed seven “rounds” of the draft, just not the same seven as the NFL. The seven Value Groups designated by DRAFTMETERICS are:
1. Selections 1-13
2. Selections 14-40
3. Selections 41-66
4. Selections 67-86
5. Selections 87-149
6. Selections 150-189
7. Selections 190 and later
These Value Groups were determined based on a review of a number of objective factors but is, by necessity, somewhat subjective in establishing the cut-offs. The following table, which considered all drafts from 1993-2012, shows some of the factors considered and summarizes the results by Value Group.
There are two points of clarification about this table. First, regarding Pro Bowl appearances, a player is only given credit for a Pro Bowl selection if he was an original selection. No credit is given for being selected as an alternate or injury replacement. A player receives credit whether he actually participates in the game or not. Second, it should be noted that the probabilities cited in the table are somewhat understated because of the inclusion of the 2011 and 2012 drafts. For example, 90% is actually a perfect score for the probability of playing three seasons or more because players drafted in 2011 or 2012 can’t possibly have reached that milestone.
In a future article or articles, DRAFTMETRICS will present success probabilities based on a shorter time period, with all players considered in the analysis having the opportunity to achieve the selected milestones. Using the full 20 years works fine for this analysis, however, as it is intended to be viewed more in relative than absolute term and treats all Value Groups the same. DRAFTMETRICS will also address differences by playing position in future articles.
There are some interesting value implications that will be addressed in a future article. In the 2012 draft Andrew Luck, the first player selected, received a contract with a guaranteed value of $22 million, while Michael Floyd (the 13th pick) signed a contract with a guaranteed value of $10million. This example seems a little silly give the respective rookie seasons of the two, but historically the two players have the same probability of success, yet one is paid twice the amount of the other. This has some interesting strategy implications, but that’s a story for another day
A final point that should be emphasized is the “bumpy” nature of the results. There are significant variances in results within each of the selection ranges. As an example, here is the range of results in the first 13 selections and then for selections 14-40.
This bumpiness indicates that there are outliers in each of the Value Groups and does not invalidate DRAFTMETRICS Value Group approach.
Tony is the founder of DRAFTMETRICS.COM can be e-mailed at firstname.lastname@example.org and followed on Twitter @draftmetrics