Every year it seems there are new and better ways to sift through the sea of information on all the NFL rookies. And over the years RotoViz has been a huge part of this, providing the fantasy football world with all sorts of fun, meaningful metrics that help us all make better decisions in our player analysis.
Recently, our very own Blair Andrews wrote a piece on the most predictive production measurements for NFL success. That (among other curiosities) got me thinking about how all these great measures of production could be combined into one score to perhaps make better visual sense of any incoming rookie class. So, I decided to dive in and see what I could build.
I call it the Adjusted Production Index.
“What’s an Adjusted Production Index?” you might ask.
Simply put, it’s a combination of Peak Adjusted Dominator Rating (ADR), Peak Adjusted Yards Per Team Pass Attempt (AYPTA), and Peak Adjusted Touchdowns Per Team Pass Attempt (ATDPTA). I used z-scores (essentially standard deviations away from the mean) for each variable to properly scale them and then combined them into one score. If you’re already familiar with those three metrics, feel free to skim this next bit. If you aren’t familiar, that’s okay! I got you:
Adjusted Dominator Rating accounts for the number of games that a wide receiver plays and calculates the percentage of team yards and touchdowns that player accounted for within their offense. The goal is to avoid just raw yardage or touchdowns and look at just how impactful that wide receiver was within their team’s offense, regardless of overall team offensive volume.
Adjusted Yards Per Team Pass Attempt and Adjusted Touchdowns Per Team Pass Attempt are similar in that they adjust for the total games a wide receiver played. However, beyond that, they’re exactly what they sound like. How many yards or touchdowns did a player record per team pass attempt?
Why those three? Well, for starters they’re some of the stickiest (most predictive) production metrics in existence outside of draft capital and breakout age. But beyond that they speak to similar production in different ways. Does the wide receiver account for a large percentage of their team’s receiving offense overall? Do they stand out when we look at efficiency in yards gained (whether via air yards or yards after catch). Are they unstoppable scoring machines regardless of their offensive scheme or scale?
The better question is perhaps what all this means for a player’s NFL success. To help shape that answer I looked at 290 wide receivers that entered the league in between 2005 and 2016, their career points per reception (PPR), and career annual average PPR and split the results into quintiles. Looking at peak college production numbers should give us some insight into a player’s NFL upside.
|API Quintiles||Career PPR||PPR Yearly AVG|
Right away it’s easy to see that those scoring in the top 20% (or even 40%) of Adjusted Production Index tend to produce at a much higher level for a much longer time than anyone else. But let’s break it down one step further and look at it in deciles (tiers of 10% each).
|API Deciles||Career PPR||PPR Yearly AVG|
This view tells a little different story. It looks like the top 30% destroy the competition when it comes to prolonged high-end production (on average). But how about we look at some names near the top?
The top Adjusted Production Index among all 290 receivers came from Demaryius Thomas with an 11.01 (which is absolutely absurd, by the way). The next two wide receivers in the entire metric were Jordy Nelson (10.61) and Dez Bryant (8.50). Other fun top 10% scoring wide receivers include:
Tyler Lockett (4.61), Hakeem Nicks (4.42), Calvin Johnson (4.38), Justin Blackmon (4.34), Amari Cooper (3.92), Braylon Edwards (3.75), Santonio Holmes (3.59), and Tyler Boyd (3.46)
At first glance, it looks like a good thing to score highly in the adjusted production index. Obviously there are some busts in there too, but we’ll get into sifting through those shortly.
First, let’s just look at how API scores were distributed through the 290 sampled wide receivers to see what a good score looks like.
|Adjusted Production Index||Percentile|
According to the production distribution, scoring in the top 30% (or least top 40%) speaks to a positive potential outcome, historically. So, in short, in looks like a score of 0.68 or better is a good API. I wanted to see if that actually made any sense so I tried to find a couple wide receivers right around that threshold and found both Julio Jones and Stefon Diggs. There were a couple decent exceptions (Antonio Brown and Steve Smith Sr.) that scored below 0.68. But as you’ll see below there’s a quick drop off in real long-term stars below 0.68. Let’s look at average PPR per year versus Adjusted Production Index.
You may notice that there were plenty wide receivers that missed even with solid API scores too. Why is that?
Two words: draft capital. That’s your (unsurprising) answer. As many RotoViz writers have touched on before, draft capital is important. But hit rates become incredibly fun (maybe more than ever) when you simply combine API and draft capital.
I split players in two groups: top-100 draft picks, and players taken later. In most drafts that loosely translates to Day 1 and 2 prospects versus Day 3 and beyond.
Let’s look at those same 10% tiers from earlier but sorted by some draft capital.
|API Deciles (Min. Score)||Career PPR (Top 100)||Career PPR (101+)||PPR AVG (Top 100)||PPR AVG (101+)|
|6 (-0.60+)||793.86 (Steve Smith)||265.01 (Antonio Brown)||89.5||46.61|
Even with Steve Smith and Antonio Brown skewing the career totals for tier six, the average yearly numbers tell a pretty clear story. Receivers that have an API score in the top 40% (0.68 or better) and see Day 1 or 2 draft capital seem to be a near lock for significant NFL production. Receivers that score in between -1.44 and 0.68 and have early draft capital also have a decent shot at success, and perhaps maintain at least “role player” status in the NFL. Beyond that it seems the rest is just noise. You might as well rank by draft capital after that (undrafted players ranked last due to even worse hit rate).
So let’s look at thirty receivers from this year’s rookie crop and try to rank them according to the above criteria (heavily weighting API and draft capital tiers).
API Adjusted Rankings
|Rank||Player||Adjusted Production Index||API Percentile||Draft Pick|
|16||Gary Jennings Jr.||0.3||53.40%||120|
|23||David Sills V||1.86||77.80%||Undrafted|
|27||Stanley Morgan Jr.||0.61||57.60%||Undrafted|
Does this mean that Andy Isabella should be your WR1? Not necessarily. But it does mean that Isabella’s combination of adjusted production profile and draft capital put him in pretty rare company. Historically Isabella’s profile hits just about every time. Does it mean Hakeem Butler has to miss because he just missed the draft capital threshold by a few picks? Not necessarily. What API does do is give us some players to target based on their combination of meaningful production scores. When you combine that with draft capital it should be easy to find value gaps in your rookie and startup drafts that take your team to the next level!
I’ll dig more into how this metric could predict draft capital and how it fits into devy leagues here soon. For now, I hope this helps you find some players that have a good chance of hitting in the NFL given their college production.