A few weeks ago alkatraz76 had a request for the Mile High Report community, and we try to give the people what they want here at MHR. Here's what he asked:
There was a lot of talk about us reaching in the draft. This was because of a few draft "experts" having the players we picked ranked lower than we picked them. So I would like to see a review of these "experts." Take a look at Kiper 's draft board from say 2005, and evaluate how accurate he was. Take a look at the players drafted, and undrafted, from that class and see how their careers have fared. Compare that too WHERE Kiper(and some other experts) ranked them. What players,by position, were ranked higher but have had worse careers than some ranked far lower. Basicly I would like an answer to this question...Are these experts draft boards absolutely correct, are they a close guideline of where a player should be drafted, or are they completely off?
Now this proved much harder than I expected due to the nature of the beast and me just being one man. Sports science journals record this on a regular basis but those are pay-to-read sites and journals so I can't just copy and paste the data from those place so I had to do most of the work myself. Now before I get started I do want state this isn't an accuracy article looking at which mock experts are the most accurate, though you can find that out through the Huddle Report which tracks mocks from most major scouts and sites like Mel Kiper and Walter Football. So that's the place to go if that's what you are looking for, this article will be looking at the variation of where a player was predicted to be drafted, but we'll discuss that after the jump.
When I went into this I wanted to look at two sources for this data, big boards (which list the players purely by talent) and mock drafts (which match players to teams based on a combination of talent, need and value). With this goal in mind I began searching the internet for these two sources. Now I stuck to more reliable and main stream sources for these boards and mocks, but I also looked at some historically accurate, small name scouts and analysts. I ended up with 39 big boards and 32 mock drafts, and while those aren't huge, they do provide a solid sample size.
Now I won't include each big board and mock since that would just be too much and too big. With the big boards and mocks I created an average and from that I could find the difference between the averaged mocked spot as well the standard deviation for the player. So lets take Justin Blackmon, he was 6.8 on big boards on average, which was a difference of 1.8 and with a standard deviation 1.5. Now for Andrew Luck he was 1.1 on our average big board which is only .1 difference from the actual draft and with only a standard deviation .3, much smaller. The larger the standard deviation, the more variation that experts saw in a draft prospect, guys like Dontari Poe and Fletcher Cox fit this bill, either being mocked very high or very low.
To put it simply, standard deviation is the spread of picks. If a player was mocked to go 15, 30, 60 and 90, they'd have a much higher standard deviation than a player mocked to go 19, 21, 23 and 26.
A player who could be considered a good value would be those who were drafted much lower than their projected average and reaches would be considered those drafted higher than the average.
Lets look at an example of these two situations:
- Mike Adams was predicted to go at 28.3 on the average mock draft but he was drafted at 56, 27.8 spots below his projected spot, very good value.
- Bruce Irving was predicted to go at 37.8 on the average mock draft but he was drafted at 15, 22.8 spots above his projected spot, a bit of a reach.
Remember some players may have been reached for because they fit a scheme or the coach had a connection to them and other players may have dropped due to recent events or concerns that arose late in the draft prospect. And this isn't trying to make a statement about one player or another, this is just what the mocks and big boards said, take what you want from it. Now a few miscellaneous notes:
- I only included the top 100 players through the average rankings on the big boards and mocks since my entire list included 342 players, which would be entirely too long.
- Here's who the two tables will be laid out:
- Actual draft position
- Average of all boards/mocks
- Difference between the average and the actual draft position (A negative means a player went before their expected spot [reach], a positive number indicates a player was drafted later than expected [good value])
- Standard deviation between the boards/mocks
- Now for players who went undrafted, I listed their draft position as 254, or one more after the last draft pick, hardly perfect, but it does carry the point.
Let's get started.
|Player||Position ||Actual Draft Position ||Average Board Spot ||Difference ||Standard Deviation
|Player||Position ||Actual Draft Position ||Average Mock Draft Spot ||Difference ||Standard Deviation
You'll notice that some player have a higher standard deviation depending on whether it's the big board or mock drafts. It seemed mock drafts have a lower standard deviation than big boards since teams needs com into play which help determine a pick, whereas a player's talent can vary depending on the scouts on their big boards. Robert Griffin is a player that represents this well, while EVERYONE had him going at #2 overall in their mock drafts, not everyone felt he was the #2 overall player in terms talent.
Now this isn't a multi-year study, nor is it a complete draft study, I just don't get paid enough to do that (considering I don't get paid anything makes me question this whole idea, I should have sold this, shoot) but I hope it was a look at how players can fall or be reached for, and which players fell into those two categories.