We compiled the results of all of the grades of every pick in the past 10 years of drafts from 2000 to 2009. By analyzing the positions, we got a lot of data that can now be examined in aggregate by the 1st-32nd slot. The results are revealing and offer some extremely meaningful conclusions about the NFL Draft.
Without further ado, here is the compiled table.
The first and most staggering observation is that we do not see any real SUBSTANTIVE advantage between the beginning to the end of the round. Yes, by the end of the first round players are averaging a 4.0 and at the beginning of the round they are averaging a 3.5. But in the middle of the round they are averaging ~ a 3.2, so the conclusion is clear:
#1: The draft value chart places way too much premium at the beginning of the draft.
In order to see this more clearly, we will average 5 rounds to smooth out the results and place the draft value chart's point value for that pick.
|Draft Value Chart|
The bottomline here is that reality is flat-lining while perception falls off a cliff. For example, you can look at almost any tranche in the bottom half of the draft and you are looking at a player in the high 3's, yet the draft value chart has these players losing 1/3 of their positional "draft" value.
We are not going to even buy the notion that picking between the 8th and 12th slots in the draft is better(!) than picking in slots 1 to 5. We can assign that to the sample size for each slot (10 players) as not being large enough, thus causing some variance (error) in the results. But we CAN draw this conclusion:
#2: NFL General Managers are not doing a very good job of drafting and NFL Scouts are not doing a very good job of evaluating.
These GMs as a group are terrible. Wonder offers up The Mel Kiper Theory: If you look the part and play the part, you'll continue to get paid.
Let's make a quick comparison between the 4th pick in the draft and the 11th pick of the draft. There is not a single "1" in the entire decade taken with the 4th pick, yet with the 11th pick in the draft, FOUR OF THE TEN YEARS there was a "1" that was grabbed... Dwight Freeney, Ben Roethlisberger, DeMarcus Ware and Patrick Willis. Of course we would expect the data to be a little lumpy, but the evidence is fairly damning- picking ahead is not giving these guys the advantage they think they have.
If you want to come to the defense of GMs and their scouts, then we can offer this conclusion as an alternate:
#2a: General Managers assign too much value to picking higher in the draft and overpay for the right to pick ahead of their peers, given the variability of results from their efforts.
The bottomline is that everyone is dreaming about finding the next Peyton Manning or Lawrence Taylor at/near the top of the draft, and the reality when they wake up is that they have a solid starter and nothing more than that. If you average 316 players taken in the first round of the draft over the ten years, you get a 3.6. And look at the average of the first five slots at the top of the draft (50 players).. it is a 3.5!!! NO DIFFERENCE. That is a large enough statistical sample to make that comparison.
It is not where you pick but who you pick. Stop coveting, sit back and someone special WILL be there at whatever slot you pick. It is up to you as the GM to find him. On balance you are not finding him. On balance it is almost random, because there is barely any real difference in results across this round. If there was discernable skill we would see a steady increase in player rating, but we clearly do not. The first pick in the draft is a 3.8 and the last pick (in the round) is a 4.0. (0.2 difference is well within 1 standard deviation (0.5) for the sample.) But go try trading up to the #1 pick overall and see what you'll have to give away- a king's ransom. All of this clearly and objectively proves Rule #9: Trading down in the Draft is good. It is nice to finally have data that so DRAMATICALLY and UNEQUIVOCALLY demonstrates that conclusion. I think we need to revise the rule- Trading down in the draft is GREAT. This is what happens when GMs as a group collectively cannot shoot straight. You can't either, so you might as well get more picks for the trouble.