How accurate are mock drafts and when do they start becoming more accurate? As one way to test this, I looked at Mel Kiper’s initial mock drafts from the 2013 compared to his final version, and the the actual results in the draft.
All the disclaimers should go here about this being one draft class (though digging through and finding all of these was hard enough, in a world full of over-saturation.) Still, it is sort of informative about when mock drafts start having value, and what that value is.
Mel Kiper did five official mock drafts last season: one in January, after the bowl season ended; another in early February, before the combine but after the all-star games; another in early March, after the NFL Combine; the fourth in early April, pretty much around the time pro days were concluding; the final one on the day of the first round of the draft.
Here is a summary of the “average error” between Kiper’s selection in a mock and the position the player was eventually drafted, with categories for picks that were exactly correct, within 3 picks either way, within 10 picks either way, and those that were more than 20 picks away from what actually happened.
It should be noted that Philly.com looked at Kiper and four others in early February–so this would be for the Mock #2 above–and Kiper was the most accurate by average error of the group. He got none absolutely correct, and was within 3 draft slots of where a player was eventually taken for only 3 players. So this isn’t an indictment of Kiper, as much as showing that most mock drafts provided virtually no one placed correctly before the Combine.
We see the first big leap in accuracy post-Combine. People like to complain that these offseason workouts and activities are overvalued, but it is either that greater clarity is provided, or more cynically, that greater intelligence in terms of what teams are actually thinking is gleaned at these organized activities.
The error rate was more than cut in half by the March Mock Draft. This was mainly the result of eliminating most of the picks that were way off the mark, and getting far more that were in the ballpark.
Not much changed between the March and early April versions in terms of accuracy. In fact, the number of correct placements decreased.
The next big jump was draft day. Yes, Mel Kiper got 8 correct in the first round (which is actually above average), but that was twice as many as he got in the four previous Mock Drafts combined. There can be little doubt that was a result of “intelligence” and information on teams at the last minute. Free agency was largely done by the time the early April one was done, so team needs were defined.
The above is a summary of the average actual draft position of the players identified in the top 10, picks 11 to 20, and picks 21 to 32 of the first round, in each Mock Draft.
We can see that the biggest jump in the final mock draft was in having the players who would actually be in the Top 10 mostly correctly placed, where the original Top 10 had players who were on average drafted at position 20.4.
When do mock drafts become valuable? It depends on what you value. If you want to know the ballpark of picks that would be in consideration for your team, then it would appear that checking mock drafts after the Combine at least gives you close to a 50/50 shot of seeing players within a small range of where they will be picked. If you are wanting to know actual selections, then it is not until the leaks of draft day that we begin to see some clarity.