Grading how a team did in a draft is stupid. Sure, maybe we can look if a team addressed their needs or if they really reached for a player or not. But, at the end of the day, grading a team’s draft selections is going to be a completely subjective process based on how a certain talent evaluator feels about the players that a team drafted.
As a recently minted NFL Draft Insider (I’m not), what am I to do? I feel immense pressure from the powers that be (I don’t) to give my insight on how teams did in the draft. So, as opposed to just giving you my canned takes on the players and how they may or may not fit onto the team they were picked by, I have decided to do something different. Something you may not know about me is that I have a Masters Degree in Predictive Analytics from Northwestern University:
And instead of just looking at my degree like I usually do as it sits on my desk at home while I pursue a career that has absolutely nothing to do with predictive analytics, I decided to actually put it to use, by bringing some objectivity to the draft grading process.
The Model:
I made a fancy excel sheet with every first round pick, where they were selected, their “average big board position” (I pulled from 7 different draft boards), what number of pick they were for their position and their average positional ranking.
From there, I created a neat little formula that was some combination of the difference between draft position/big board ranking with a multiplier for if there was either a positive or 0 difference between positional ranking and where they were selected among players at the same position.
To give you an example: Saquon inputs looked something like this
[{Draft position (2) – Avg Big Board ranking (1)} +1] +([{RB Taken (1) -RB Avg ranking (1)}+1] x 2)
Saquon’s “value” is a +4.
So, if you were picked exactly where ‘you were supposed to’ then your value would be +3.
However, something important must be noted. I do not really care how large the positive value is, only that the number is positive. The reason for this is that players selected early in the draft cannot have very high values. Because Saquon was selected 2nd overall, the ‘ceiling’ on his value number is immensely smaller than someone who is picked in the back half of the first round.
Things I did not include in my model:
- Team needs
- System fits
- Anything other than 1st round picks
- Positional value (i.e. leniency for QB’s OT’s)
Categories:
- Clear net positive
- Near zero on either side
- Clear net negative
The Results:
Clear Net Positive (in order drafted):
#2 Overall Pick (NY Giants) Saquon Barkley, RB, PSU
Saquon has the rare distinction of being the only player in the first round with no decimals in his overall net value number. He was the #1 overall player and #1 RB on every draft board I pulled from, giving him a nice clean +4 ranking. Saquon is thought of as a generational talent at the running back position and should give the Giants an instant surge of offense as both a runner and receiver.
#5 Overall Pick (Denver Broncos) Bradley Chubb, DE, NC State
‘
Chubb is another guy that had a clear plus value in my formula. His average big board position was just under 3 and he was the #1 DE on every board I used. As most agree, the Broncos continue to draft well and got a day 1 starter with an incredibly high floor at the #5 overall pick.
#6 Overall Pick (Indianapolis Colts) Quenton Nelson, G, Notre Dame
(this play WILL NEVER get old)
Here I am, writing about Quenton Nelson again, we just can’t get away from each other. Nelson had a nearly identical value metric to Chubb. His average big board ranking was also right around 3, was only one pick behind Chubb and similarly had an average positional ranking of 1 and was selected 1st out of his position group. Good work Indy, ya got a good one.
#8 Overall Pick (Chicago Bears) Roquon Smith, LB, Georgia
Although his average positional value was slightly lower than where he was taken (~1.3 vs 1), his average big board position (just under 6) was higher than where he was selected (8th). Most people are lauding the Bears for drafting someone who has a chance to be one of the better players in this draft and my super awesome model agrees.
#10 Overall Pick (Arizona Cardinals) Josh Rosen, QB, UCLA
It was very unsurprising to see Barkley, Chubb, and Nelson receiving net positive values. They had the three highest average big board rankings and all were drafted lower than where they were ranked. However, seeing someone like Rosen, who was a very polarizing prospect, in this category, remains a bit of surprise. His average big board ranking was basically identical to his draft slot (~11 vs. 10) and he had a slightly better positional ranking (2.6) than where he was taken among QB’s (3). He may be ‘too smart’, but value-wise, it looks like Cardinals drafted well.
#11 Overall Pick (Miami Dolphins) Minkah Fitzpatrick, S, Alabama
We can file this one under the Nelson/Chubb/Barkley no surprise category. Minkah was actually one of the more polarizing players in the draft, coming in as high as #3 on some big boards and as low as double digits. However, his average ranking (~6) was well below where he was picked (11th), making him a surefire net-positive player.
#16 Overall Pick (Buffalo Bills) Tremaine Edmunds, LB, Va Tech
Tremaine Edmunds had one of the highest overall value scores, which given how high he was ranked by some experts, is no surprise. Physically, there isn’t much more you would want out a linebacker prospect than what you get with Edmunds. He is 6’5” 250 lbs and runs a 4.55. But, he is only 19 years old and at times is very undisciplined. Because of this, he likely fell a few more spots than expected and gives the Bills great value at #16.
#17 Overall Pick (LA Chargers) Derwin James, S, FSU
Personally, I could not believe James fell all the way to #17, amd my model agrees, granting James the highest overall value metric. His big board ranking of ~8 was nearly 10 spots above where he was drafted (17th) and was ranked as the #1 safety by two of the boards I pulled from. James looks like he belongs on the field in the late 1990’s in a FSU vs Miami game, and I say that in the best way possible. James seems like a great addition to the Chargers quietly awesome defense.
#22 Overall Pick (Tennessee Titans) Rashaan Evans, LB, Alabama
Another year, another ‘Bama LB going in the first round. Evans had a positional value slightly higher than where he was selected (~3 vs. 4th) and had a big board ranking just above where he was selected as well. All signs point to a nice, value-add by the Titans.
#26 Overall Pick (Atlanta Falcons) Calvin Ridley, WR, Alabama
Calvin Ridley had the second highest overall score in my whole model, with a massive differential in avg. big board position vs. actual draft spot. Although DJ Moore became a hot name late, Ridley was still thought of as the top wide-out in the draft, with his elite speed and ability to create after the catch. Concerns about his slight frame and iffy hands likely gave Moore the edge.
#29 Overall Pick (Jacksonville Jaguars) Taven Bryan, DT, Florida
Bryan quietly had one of the highest value metrics in my model. Although nearly all experts had him below Vea and Payne, Bryan was pretty firmly thought of as the 3rd DT in this draft. On top of that, he also enjoyed an average ranking of ~25, a solid 4 spots above his draft position. Sacksonville got another piece to add to their already frightening DL.
#30 Overall Pick (Minnesota Vikings) Mike Hughes, CB, UCF
If my model included mock draft positions, Hughes may not have made the net positive cut. On most big boards, he showed up in the mid 20’s but experts understood he would likely fall a few slots because of off the field issues. Well, my model is not perfect, and according to it, Hughes is a net positive selection. I do think this may turn out to be an awesome pick, Hughes is insanely talented and going into a good situation like Minnesota will be perfect for him.
#32 Overall Pick (Baltimore Ravens) Lamar Jackson, QB, Louisville
Lamar had by far the highest rating for any quarterback. This is not necessarily surprising because his positional ranking was nearly identical (s/o to the one expert who had him 4th ahead of Josh Allen, I agree) to where he was picked among QB’s and his average big board ranking was in the mid 20’s versus a selection spot of 32. If I included mocks, his number would likely even been higher as many experts had him going in the 15-20 range. Love this pick for Baltimore.
Near Zero Value
This next grouping of players had a net value between -2 and 2. Basically meaning they were picked in the exact right spot.
An example of how this works: Sam Darnold
[{Draft position (3) – Avg. Big Board ranking (6.33333)}+1] + ([{QB Taken(2) – QB Avg ranking (1.5)}+1]x2) = .66667
#3 Overall Pick (New York Jets) Sam Darnold, QB, USC
#7 Overall Pick (Buffalo Bills) Josh Allen, QB, Wyoming
#13 Overall Pick (Tampa Bay Bucs) Vita Vea, DT, Washington
#23 Overall Pick (New England Patriots) Isaiah Wynn, G, Georgia (I entered him as a guard as that’s what most boards had him as)
Net Negative Value
Everyone Else
