Devonta Freeman made me look like an idiot last year.

For a lot of reasons, I was on #TeamTevin entering the the fantasy football season last August. He (Tevin Coleman) was an interesting prospect who was hand-picked by a new Atlanta regime with a high-ish equity draft pick, Freeman was a pretty inefficient runner during his rookie campaign the previous season, and the situation in Atlanta was a favorable one from a fantasy football perspective.

The only correct part of my process was the fact that Atlanta was a good spot for a running back to thrive. Unfortunately for my Twitter mentions, that running back was Freeman, not Coleman.

I'll always believe Freeman's season was somewhat of an anomaly, but there's no doubt that I gave up on him too early. Kyle Shanahan entered the picture, Coleman got hurt, and Freeman sucked up all the volume in a fruitful backfield. And Freeman caught passes -- which he did well the previous season, for what it's worth -- which certainly helped his cause.

Fantasy owners who drafted Freeman more than likely found success last year, not just because he scored so many points, but because the cost in acquiring him was pretty minimal -- he far exceeded expectations.

And that led me to think in a little more abstract manner -- how rare of a find was Devonta Freeman last year?

Meeting Expectations

To help solve this question, I looked at the last five years worth of average draft position data from real, 12-team PPR leagues that held drafts after August 1st on  MyFantasyLeague.com. Using this data, I plotted the number of fantasy points scored (Weeks 1 through 17) by a running back versus his average draft position.

Not surprisingly, the data has a downward trend, as the higher the average draft position (the lower the actual draft pick number), the more fantasy point scored. 

What the trendline on the graph shows, essentially, is expectation -- it's the number of points players have scored over the last five years given a particular average draft position spot.

For example, if you picked a running back 25th overall, we'd expect him to score 158.71 PPR points. A running back selected 67th in a draft has an expected fantasy points score of 115.53.

Remember, the goal here is to find how valuable Devonta Freeman was last year for fantasy owners. But while we're at it, we can take a look at a larger sample than just him.

Crazy-Good Value

Freeman scored 316.9 PPR points last year, which was by far the most from any running back in fantasy football. But he was selected, on average, at pick 113.43. A running back over the last five years selected at that slot has scored, according to our trendline, 99.97 PPR points. In other words, Freeman outproduced his expected PPR point value by 216.93 points.

And that, ladies and gents, made him the biggest value of the last five years at the position.

Before we dig in, now is probably a good time to say that not all breakout running backs -- like 2014's Justin Forsett -- are part of this data set. While the average draft position information spans a little over 20 rounds, a player like Forsett wasn't even on fantasy owners' radars entering 2014. That really didn't happen often, but it needs to be noted.

Now, the table above predictably shows a lot of late-round picks. Why? Because late-round picks need to have a great season in order to far outperform expectation, while an early-round one would need an elite season to outperform cost dramatically.

That's what makes Jamaal Charles' 2013 or even Ray Rice's 2011 so special. They were drafted, on average, in the top five of drafts, but were still able to return an insane amount of value. 

Other players like Le'Veon Bell, Adrian Peterson, and Matt Forte were selected in the second round and were able to return nice value, too -- if you look at the graph above, you can see how steep the drop-off is as you move to about pick 50. In other words, high-end picks are simply expected to do a lot given their past performances.

Another thing to point out is the lack of true handcuffs on this list. We've written about why handcuffing running backs is a losing strategy in fantasy football, and as you can see from the list above, the only backs to have full-season value that were selected as handcuffs were DeAngelo Williams and Michael Bush. And a huge reason folks were drafting D-Will wasn't so much the handcuff factor, but because Le'Veon Bell was suspended to start last season.

You handcuff your running backs later in the season, not during your draft. This is just more evidence for that.

If there's any sort of trend in finding a late-round value, it seems like it's two-fold. First, you'll want to, unsurprisingly, target backs in ambiguous backfields. Guys like C.J. Anderson and Knowshon Moreno entered their particular valuable seasons behind unproven runners. So did Freeman, if we want to consider Tevin Coleman as ahead of him in the depth chart -- he at least had a higher cost in acquiring.

And then there's the always-underrated pass-catching backs like Danny Woodhead and Fred Jackson. To be honest, when drafting, your best bet is probably to target these types of players in hopes for their ground-game opportunity to rise. That's what I didn't end up doing with Freeman which, as I mentioned earlier, was a big mistake.

Actually, it was the biggest mistake I could've made at the position over the last five years.