In previous iterations of this exercise, I highlighted how mock drafts almost universally use an overly subjective method to assign players. We can't really know how teams view themselves and so typically mock drafts simply go off pure need.
I wanted to find a data-driven method to do a mock draft. It is no more accurate perhaps to what will happen on draft day (trades for one are impossible to gauge) but one interesting outcome of this exercise is that it applies a level of predictability to picks since it combines both known needs and best player rankings from reputable, neutral sources.
In this case, I'm using Mike Mayock's Top 100 Prospects board. Later tonight he will conduct his own mock draft, the only one he does all year.
To find the team needs I went to another NFL.com source - their draft team's identification of team needs.
On some limited level, teams boards are similar. For instance, the top receiver on everyone's board is probably Sammy Watkins. The top pass rusher on everyone's board is probably Jadeveon Clowney, and so on.
However, once players start flying, the pressure to fill a need rises - and that's when GM's get into trouble as they abandon boards. The patient teams are comfortable with chaos. They might simply take the best player that shows up at their pick, or they may move back so they find a better one.
The other interesting outcome of this exercise is that it helps identify trade opportunities because of the value differential that shows up in some picks. For example, does Houston really want to take Mack at #1? Doubtful. They are desperately trying to trade back so they can get a QB and acquire picks, similar to the RG3 trade two years ago. St. Louis is doing the same thing because they would prefer to move back a few slots and extract a fortune from someone. Atlanta however is probably wanting to trade up - they would love Mack or Clowney, but only if its feasible.
On the other hand, when a team is on the clock and the BPA is a guy they don't necessarily want (let's say a Running Back in round one) then trading DOWN becomes imperative. A team knows its bad value to take an RB maybe at 29 just because he's the 25th best prospect - it might be better off trading back, get picks, and pick someone else, rather than reach for another need. Green Bay seemed to have done this last year, moving no less than twice back and still landing Eddie Lacy, a back with a first round grade from most.
When the BPA draft shows a huge value differential, that is a trade opportunity.
We also know that players fall on Draft Day. Aaron Rodgers, Randy Moss, Shariff Floyd: all these players fell tremendously when the moment of truth came. So unusual reaches and falls are not uncommon.
Anyhow, I took the top 100 and mocked it as far as Baltimore's 3rd round pick by simply selecting the best guy available at one of the four positions of need. By the third round, the scenario starts to break down a little because it gets hard to fill team needs within the BPA model that late so i would say its value drops a bit after pick 50 but thought it would be interesting to see who could be available at 79.
Without further ado:
- Mayock's ranking of Khalil Mack at #1 is stunning. It is hard to know what teams think this way but this is a really amazing development. I have heard this a few times but didn't really give it much legitimacy.
- I can't see a scenario in which Clowney actually falls to 5. This is what I meant above by trade opportunity. If Clowney somehow didn't go at #1, someone will be on the phone with St. Louis, and then Jacksonville and so on. But in this scenario Clowney does fall so expect to see some movement because of it. If its true Houston prefers Mack (and they've actually indicated that they do, though this could be smoke too of course), then we are in for some really interesting action tomorrow.
- Quarterbacks: Mayock has Manziel and Bortles very high but Bridgewater way down at 42. Both his top QBs go early to Tampa and Minnesota
- Kyle Fuller is the best CB on his board. This is a big time adjustment from most boards out there but not necessarily too surprising. Much of the buzz around Fuller is extremely positive as a sleeper for best CB.
- Nothing surprising happens until Baltimore's pick. The tackles were long gone by this time, as were Clinton-Dix and Ebron. That leaves a choice between CB Dennard and WR Brandin Cooks. They were rated #19 and #22 repsectively. Since we'll stick to BPA, the choice is Darqueze Dennard.
- Arizona takes the third QB on the board, Derek Carr. Now, while Arizona needs a QB, other teams need them more. I think this is another major trade window. Teams like Cleveland, Houston, etc who failed to get a QB early, will have to start thinking seriously about moving back into the 1st round for a QB, just like Baltimore did in 2003.
- Mayock has Cyrus Kouandijo at #24 which is interesting. That's probably earlier than the negative hype would have us believe but we'll see.
- Cleveland's pick at 26, a reach per this board for Joel Bitonio is again a trade opportunity. That pick wouldn't make sense if that was Cleveland's actual board. They need a QB still but they pick up Mettenberger later at 35. This is one of the most complicated parts of the draft to predict - we just don't know how many teams are going to have a guy in mind they get desperate for. However, most of these QBs now are not first round values and so per this board, they don't go until the second round.
- Baltimore's second round pick just feels perfect in so many ways. A highly touted free safety, Jimmie Ward would make a lot of people happy. While it might seem counterintuitive to go double defense when the offense was the problem last year, this is also what the best player mantra suggests. Given that, maybe Baltimore does look for a trade down so it can go offense. Still there's no doubt that these two secondary picks fill major needs and at reasonably good value.
- Jacksonville is the final QB needy team to take one, selecting AJ McCarron in the third round
- Finally, Baltimore, lacking any particular need sees a ridiculous value with Tre Mason still on the board. With a value discrepancy of 23, this is the type of move Baltimore would make in a heartbeat.