I'm in the process of writing up this year's final projections. For now, though, here's a look at the accuracy of my projections in previous years. My first attempt in 1998 was done using very little history on which to base my projections and questionable contract information from media reports. As my sources have gotten better and my information more precise, my projections have gotten much more accurate. Aside from the 2007 comp picks, which were more difficult to project because of the huge jump in the salary cap after the 2006 CBA extension, I've typically gotten 25 or 26 exactly correct and missed by one round on about four others. More precise information also has allowed me to project the exact order of picks in many cases. Although my ultimate goal is to project all 32 comp picks correctly and in the precise order of the actual comp picks, there are too many judgment calls every year to make that reasonably possible. I will, however, keep trying.
Here are the result of my previous comp pick projections (with links to all of the projections since 2002) --
1998 — I projected 23 comps; the NFL awarded 19. I had 15 of them going to the correct team, with 10 going to the correct team in the correct round and three more off by one round.
2002 — 22 correct, four others off by one round
2003 — 26 correct, four others off by one round
2004 — 26 correct, two others off by one round
2005 — 22 correct, six others off by one round
2006 — 26 correct, four others off by one round
2007 — 20 correct, five others off by one round
2008 — 25 correct, four others off by one round
2009 — 26 correct, three others off by one round
Note: I do not count "cheap" correct picks. For example, if I project a team to receive a fifth-round pick and a sixth-round pick but they instead receive a sixth and a seventh, I would count both of them as being off by one round. If, however, I am certain that the sixth-round pick was for the player I projected in that slot, I will count that pick as correct. The team's other pick, which I projected as being a fifth but was a seventh instead (and almost certainly for a different player), would simply be incorrect.