Probabilities of the NFL Draft

In the recent article “Breaking Down the NFL Draft” various metrics were used to separate draft selections into “Draft Ranges”. Each of these Draft Ranges contained selections that had similar results from the draft (e.g., selections five through 14 had similar career outcomes, etc.). This article takes that analysis a step further and discusses what history says about the probability of a player achieving specified milestones. Future articles will compare the data from this article to data by playing position and other perspectives. The major difference between the two articles is that “Breaking Down the NFL Draft” is based on drafts for the most recent 20 years (1995-2014). This is useful for the purpose it served, but does measure probabilities very well. A player selected in the 2014 draft, for example, is not going to be a five-year starter after his first season. The same can be said for players selected in the 2010-2013 drafts. In this article, the same starting point (1995) is used to measure an achievement but the ending point is tailored to that milestone. In order to tailor the measurement period to the time a player has been in the league, the measurement period used is the minimum time required for a player to achieve a milestone plus one season (thereby providing a cushion). This results, for example, in the 1995-2009 drafts being used to determine whether a player achieved the five-year starter metric. The only exceptions to this are milestones that are achievable in a player’s rookie season. The milestone “Rookie Starter” is the most obvious example of such an exception. As a reminder, the following definitions are used in all of my articles:
  • A starting season is defined as one when a player starts at least eight games. This definition is used for the starter metrics.
  • A player receives credit for a Pro Bowl selection only when they are an original selection, whether they played in the game or not. Alternates do not receive credit for a Pro Bowl selection.
  • A player receives credit for an All-Pro selection if he is selected All-Pro by either the Associated Press or the Pro Football Writers of America.
The following table shows the probability of a player in each Data Range achieving the nine designated milestones. Where the “<” sign is used, it means that a player did not quite achieve the percentage shown. For example, <100% indicates that a player has more than a 99% probability of achieving a milestone but less than 100% (e.g., 99.5%). The “Last Year” column is the ending draft year included in the analysis (with 1995 always being the first year). All other columns are self-explanatory. There are a number of observations that can be made from the above table:
  • Virtually every player drafted will play at least one year in the NFL
    • Even the latest draft choices have a 75% probability of making a roster
    • This may not be for the team that drafted them or in the year they were drafted
  • There is practically no difference among first round selections in a player lasting five years in the league
    • The probability declines rapidly beginning late in the third round
  • The principal difference between the first and second Draft Ranges is the probability of drafting a Pro Bowl or All Pro player
  • The chances of drafting a five-year starter decline rapidly after the first 24 selections
    • Chances are better than 50-50 in the first 24 picks
  • The chances of drafting a player who wins post-season honors is remote after the first 73 selections
    • The drop-off is even more drastic when it comes to multiple All-Pro selections
Follow Tony on Twitter @draftmetrics

Upcoming Games