How much difference is there between drafts?
This is the time of year the experts tell us how good or how bad this year’s draft class is, which playing positions have a lot of depth, and so forth. By the time the outcome of the draft class can truly be measured, though, the class has lost its identity and any discussion centers around anecdotal situations. (e.g., the 2004 QB class).
In this article DRAFTMETRICS sets out to measure the results of past draft classes in order to determine how results vary from year to year. The 1993 through 2006 draft classes were reviewed for this purpose. The analysis ends in 2006 because that class has pretty much taken form by that time and players have had a reasonable amount of time to achieve the five-year milestones that DRAFTMETRICS often employs. Classes after 2006 are still evolving and it may be at least somewhat premature to conduct a full review of their results, at least in regards to the five-year measures. The following table summarizes the draft results by class.
This table shows that some years are indeed better than others. It is somewhat interesting that the gap between best and for five-year starters is lower, on a percentage basis, than any other measure except for three-year careers.
The fluctuations are more pronounced when viewed by playing position. The following table shows the number of five-year starters by playing position from each draft class. Every position shows fluctuations where the best year has at least double the number of five-year starters as the worst year. For example, the 1994 draft class produced only six five-year starters at defensive back while the 1998 draft class ended up with 16 five-year starters.
Another way to compare the draft classes is by number of games started. This permits DRAFTMETRICS to track results on an intermediate basis and even the most recent years can be compared to other years. The following table shows the cumulative number of games started for each draft class after each season (limited to 10 years for presentation purposes). The table indicates, for example,that players from the 2003 draft classstarted a total of 703 games in their rookie season, a total of 1857 games for their first two seasons combined and so on down the line.
This analysis supports the contention that there are good and bad years. The table also allows the reader to see which years are shaping up as good years before the final tally is in. 2006, for example, looks to be a very strong year.