I was planning to do a follow-up article on DERPP next, but the discussion after that first article led to some very interesting ideas about how else we could evaluate return on investment (ROI) from teams’ draft picks.
So I have chosen to use Approximate Value (AV), a metric developed by pro-football-reference.com, to assign every player’s value to his team every year. Because of this, they have been able to figure out the average career value for players drafted at every spot in the draft.
The method I will use will normalize high draft picks (since you SHOULD get good performance out of them) and low draft picks (since very little is expected from them). I could have used the “Jimmy Johnson” draft slot values for this, but decided against it since they overvalue the top 15 picks, and there is no value to put in the numerator if we use that measure of draft slot value.
So what’s the method?
Drafted player AV divided by player’s expected career AV
Using drafted player AV means that you only count the value generated for the team that drafted the player. Malik Jackson has a career AV of 37, but only 17 of that was generated while he was a Bronco. Jackson was drafted as the 137th pick in 2012. The expected career AV for the 137th pick is 3.3 so Malik’s ROI to the Broncos is 17/3.3 or 5.15.
Let’s look at two other examples of this method. Andrew Luck was taken with the first overall pick in 2012. The expected career AV for the first pick in the draft is 34.6. Luck has a career AV of 59. So the Colts have gotten 59/34.6 or 1.71 ROI out of their selection of Luck.
The Broncos selected Trevor Siemian with the 250th pick in the 2015 draft. The average career AV for the 250th pick is 0.1 (most 7th round picks play very little in NFL games). Siemian has a career AV of 13, so his ROI for the Broncos is 13/0.1 or 130. In other words, the Broncos got “decent” return on a tiny draft investment.
Now, unlike DERPP, this does not look at the quality of his play (although Siemian’s AV of 4 in 2017 was the lowest AV in the league for a QB with 10 or more starts). Se let’s look at the ROI for every draft pick from 2012-2017 for the Broncos.
|Player||AV by draft slot||AV with DEN||ROI|
Keep in mind that this is a now-look, so that players who have only been in the league one or two years (even if they have played quite a bit), probably have not shown large ROIs unless they have torn it up (and frankly, most of our draft picks have not done that recently).
We see that the biggest ROIs the Broncos have, not surprisingly, come from late-round picks, with Siemian, Corey Nelson, Riley Dixon, Matt Paradis, Danny Trevathan and Will Parks as the only players with an ROI > 10.
One of the reasons that I DON’T really like AV as a measure of value to the team is that you can get results like this where Malik Jackson and Max Garcia are shown to have essentially the same ROI for the team. This is decidedly not the case.
So now let’s see how the entire league has done from an ROI standpoint 2012-2017:
|Rank||Team||Total Draft ROI||Total Draft Capital by AV|
You’ll notice that I have also summed the total amount of draft capital that each team has had to spend over these six drafts. Denver has had the third least amount of draft capital (by summed expected Career AV), but has gotten the third most ROI from its draft picks.
Now, before you start screaming about this in the comments, I will fully own that this ROI metric assigns way more value to the play of Trevor Siemian than I would like. You will note that he accounts for 130 of the team’s total of 321. In fact, by this metric Siemian provided more ROI by himself than 11 teams got from ALL of their draft picks combined.
As such, I thought about injecting some kind of scaler into this ROI metric to “soften” the impact of day-three draftees, but I decided against it. Feel free to debate/discuss this with me in the comments.
Did you expect the Broncos ROI for 2012-2017 to be this high?
This poll is closed
Sort of - because of Siemian