The NFL Draft is an asteroid packed with valuable resources waiting to be mined. I'm not talking about the rookies, although they are fun and new, I'm talking about the data. At the NFL Draft, we get a detailed snapshot of the NFL marketplace in all its glory - there is no bend to Draft Capital, the number that a player is given never changes. I'm not sure that there is a sturdier data point: Draft Capital is universal yet specific, simple yet precise, instantaneous yet eternal. The pick that was used on a player is exactly what the market determined that individual was worth. And what the NFL determines a player is worth is immensely more valuable that what you or I think a player is worth.
I can hear it already, the voice of 10,000 armchair GMs crying out, "The scouting report I wrote on Twitter shows that Bill Belichick is wrong!" We get it. Antonio Brown was a 6th round pick. Kevin White was the #7 overall pick. The NFL is wrong, a lot. The greatest football player of all time was taken with the 199th pick for crying out loud. But this is not about success or failure. This is about investment and opportunity.
Here's a picture of 3rd rounder, Joe Montana winning his third of 4 Super Bowls, so you never forget that it doesn't matter when the Great Ones were drafted.
When a team misses on their first round pick, it kills them. So they better do everything they can to get that player to succeed. The high cost of investing in that individual means he will be given every opportunity to produce a quality ROI. The coach will focus more on him, he'll be the first to get snaps out of all the other newcomers, his draft status will break every tie as he competes for touches. The team will expect him to get on board quicker, the staff will trust him faster and his victories will be celebrated sooner.
Day 3 picks (4th-7th rounders) are nothing to a team. They're injury insurance, punt team specialists and lotto tickets. By year 3, more than half of the Day 3 picks are off the team. In a world where a new crop of fresh potential comes in every year, Day 3 picks are training camp fodder, unless they can quickly carve out a role. They come with no expectations, no investment cost and are therefore easily replaceable assets.
To help illustrate this point I've included the work compiled by Topher Doll at the Mile High Report. Below are the Retention Rates of NFL Draft Picks over the course of their first 5 years in the league.
Here we see the how quick the life cycle is for Day 3 picks and even 3rd rounders. 1st and 2nd rounders have 4-5 years to prove that they are worthless in order to get cut and everyone else gets 1-2 years to prove that they are worthwhile to avoid being cut. The presumption of innocence is only applied to those on top.
As it stands alone, this chart is just the first layer. We've only scratched the surface of our asteroid. Aren't those 1st and 2nd round picks better players after all? We should expect them to have a higher rate of retention. And so, we dig deeper.
Russell Clay wrote a foundational piece on Success Rates in relation to Draft Capital. I highly recommend taking a look. In it, he lays out an essential premise:
When a player is taken earlier in the draft, they have a higher probability of success.When a player is taken later in the draft, they have a lower probability of success.
Sounds great, looks even better. Check this baby out.
Clay's parameter for RB success in the above chart is a season where the RB gains 1000+ yards from scrimmage. I decided to expand on this work as I started my own research.
To begin, I collected the last 20 years of Fantasy Running Back Data from Pro Football Reference and organized them by Draft Capital. In order to emphasize the focus on actual draft picks, I made the data player-agnostic and built it on what the science world would call "Player-Seasons": the production statistics attributed to an individual player over the course of a single season. You can refer to them as "Performances" or "Seasons" for simplification.
My parameter for a successful performance is quantified by PPR Points Per Game. To further specify the Draft Capital research I used 3 categories of success.
An RB2 Performance scores 12+ PPG. An RB1 Performance scores 15+ PPG.An Excellent RB Performance scores 20+ PPG.
Here are the results of the 3000+ RB Performances studied:
Now it's easy to determine the probability of single season success for any player at any point in their career. All you have to know is when the player was drafted. Why do I say that this will work for any season? Because the data using 3000+ performances over the last 20 years takes the collective of all the performances in that time span. The data doesn't care if Edgerrin James was on the Colts or the Cardinals or if he was 22 years old in 2000 or 28 years old in 2006. It doesn't even care if it's Edgerrin James, Cedric Benson, Darren McFadden or Ezekiel Elliott. It only cares that the performance was accomplished by the #4 overall pick. All the hits and the misses of the NFL Draft, every reason for success and every excuse for failure are represented in this data. In the end, it doesn't matter what the excuses are, the data doesn't care. A miss is a miss, and a hit is a hit. For our purposes only 2 variables matter: Which Pick and How Many Points.
As a practice example let's take Aaron Jones, a 5th round pick. Roll up to the chart and simply find the 5th round picks. When compared with the other 5th round picks we can say that he has an 8.30% probability of turning in an RB2 performance, a 2.70% probability of an RB1 performance, and a less than 1% probability of turning in an "Excellent" RB performance (he'd be the first to do it in the last 20 years). Again, this data doesn't care why you think Jones is special or estimate how many times he's going to pull his hamstring this year, it just tells you plain and simple: here's what the other 5th round picks have done year in and year out.
Let's do another example because Rocket Science is fun. You're wondering about Boobie Miles who was taken 8th overall in the NFL Draft, what're his odds of success? Boobie's got a 54.20% chance of scoring greater than 12 PPR points per game, a 36.70% chance of scoring greater than 15 PPR points per game, and a 17.50% chance to score greater than 20 PPR points per game. Get on board, this baby is heating up.
Should we even be attempting such amazing feats with retrospective data, isn't that dangerous? You bet your mama it's dangerous. Do you think John Glenn Jr. was asking himself if it was dangerous to be the first man to orbit the earth while flying in a tin can literally pasted together with red duct tape and saran wrap? No. John asked himself one question and one question only: Is this badass? That's really the only thing that matters, ever. So buckle up, big boy we're doing it.
Our guy John getting on board Friendship 7 after seeing Friendship 1-6 spontaneously burst into flames. He required extra assistance due to the size of his massive nuts.
Yeah but JetPack, is this really that badass? Well, it depends on your definition of badass, but no, it's not badass, there's a lot cooler things out there. I literally fought with the Catholic Church over the center of the Solar System and that was too badass for them, so it's different for everyone. From my brief Google search of "Fantasy Football Draft Capital," I do think this is the first time anyone's ever done this research before, although one of you idiots could have written the very same article and called it "Draft Capitol" instead. "Capitol" is the building where the US Congress meets to discuss what type of boats they plan to buy with your money. "Capital" is an asset, like money or draft picks. It can also be the big version of letters that you use to indicate a proper noun, like Enceladus. Anyhow, you and I are pioneers of some sort and the hypothetical novelty of this research is at least entertaining.
After running the probability data for players taken from each round, I calculated average PPG for each individual pick of the draft over the last 20 years. That's right folks, using a trendline with an R^2 of 0.993 (1.000 = perfect correlation), I extrapolated the average RB PPG for every single draft pick.
RB2 Probability Per Draft Pick
Hopefully you can figure out how to use this chart on your own. But for the sake of the draft capitol guys, I will spell it out.
1. Think of a running back.
2. Find out which pick was used to draft the running back.
3. Look at the chart
4. Go undefeated in fantasy and use the winnings to woo your wife back
Maybe you're not an odds guy. Maybe you think you're the greatest talent scout on the planet and you only rely on your own hand times for measuring the 40 yard dash. Maybe you only care about the atmosphere at the Senior Bowl and are impressionable enough to take Daniel Jones #6 overall due to his performance in a meaningless All-Star Game. All you have to do is be more consistent than a coin flip with your first round picks and you're better than the NFL. But you have to remember, this is not about your opinions. This is about someone else's investment. This is about what someone else believes about that player. The GM puts his job on the line to pick the player, and his coaches are the ones giving out the touches. Because of this, Draft Capital can provide a foundational expectation or a perspective lens for an individual's success.
