Devonta Freeman has been an absolute beast in fantasy football the past two years. He finished as the RB6 in PPR leagues last season after coming in as the overall RB1 (by roughly 56 fantasy points) during the running back apocalypse of 2015.

However, past production does not equate future performance. There are several warning signs that should have us wary of Freeman’s outlook going forward into 2017, especially when the cost to acquire him is a late-first-round pick in PPR formats, according to Fantasy Football Calculator.

Offensive Regression

The 2017 Atlanta Falcons' offense will likely experience a fairly significant offensive regression. Not because it fits the "Super Bowl hangover" narrative or they lost Kyle Shanahan -- but because there's almost no way they can be as good as they were last season, when they were a top-10 all-time offense.

Matt Ryan has been in the league for nine years now. We have a pretty good track record on the type of player he is: a good quarterback but not the MVP-caliber player we saw last year. Ryan’s 2016 saw him set career-best marks in numerous efficiency metrics such as adjusted yards per attempt and quarterback rating. Ryan’s touchdown rate last year was 7.1%, more than double his 2015 touchdown rate of 3.4%. We can expect his 2017 touchdown rate to move closer towards his career clip of 4.7%.

Ryan is the quarterback and thus the engine of the offense, but we can look at more than just his career year to demonstrate how big of an outlier last season was for the Falcons. We'll do that using Net Expected Points (NEP). If you're unfamiliar with NEP, it's our in-house metric which shows how many expected points are lost or gained on a play. A nine-yard completion on 3rd-and-10 is different from a nine-yard pickup on 3rd-and-8 -- NEP accounts for that. (You can read more about it in our glossary.)

From 2008 (when Ryan entered the league) to 2015, the Falcons never finished with a higher schedule-adjusted NEP per play mark than 0.11. In 2016, Atlanta more than doubled that clip, posting an eye-popping and historically great 0.26 Adjusted NEP per play. For reference, this is the highest end-of-season mark in Adjusted NEP per play in our database, which goes back to 2000. The next best offense was the 2007 New England Patriots.

Seasons like the Falcons just had are historical outliers for a reason, and Atlanta's offense would have to buck some serious odds to perform at that level once again.

Naturally, if the offense is bound for regression, it will be more difficult for Freeman to score touchdowns and rack up fantasy points at the same rate he did in 2016. Team scoring matters for a running back’s upside in fantasy football, and Freeman’s fantasy point total last season was buoyed by the otherworldly Falcons' offense.

Touch Share

Some of the most correlated statistics to running back fantasy production are touches and touch market share. The Falcons' offensive success generally masked the fact that Freeman was not utilized as an every-down player in 2016. The following numbers come from Pro Football Focus' Scott Barrett.

In fact, Freeman saw about 60% of the Falcons running-back snaps and opportunities. While this is still well above average in today's NFL, Freeman’s 60% mark is below backs being taken in his range in drafts -- guys like DeMarco Murray and Melvin Gordon. Even Jordan Howard and Jay Ajayi, both of whom didn’t start the first few weeks of the season, had higher snap and opportunity shares than Freeman last year.

With offensive regression expected, Freeman likely needs to dominate snaps like his counterparts in order to live up to his draft cost . Based on his market share numbers from 2016, this is not something we can count on.

Declining Usage in the Passing Game

The most concerning part of Freeman’s usage last year was his role in the passing game. Even in standard formats, passing game involvement is crucial for fantasy points. Via a target is worth more than a carry in terms of expected fantasy points.

The chart below depicts the fantasy points per rushing attempt (FP/Attempt) at the running back position -- all running backs -- as well as the fantasy points per target at the position.

As JJ stated, for the sake of clarity, the fantasy points per attempt column isn't taking the total number of fantasy points scored by a running back and dividing it by the number of carries. Instead, it's looking at specifically rushing -- the amount of points a running back is adding only when he gets a carry. Naturally, because the back isn't catching a pass, there's no need to designate the difference in average in a PPR and non-PPR league. (The Tevin Coleman.

Again via Scott Barrett, Freeman actually saw less than half the snaps when the Falcons were in obvious passing situations (3rd and at least 5 yards to go). This was an optimal decision for the Falcons, as well, as Coleman (0.78) nearly doubled Freeman (0.40) in terms of Reception NEP per target.

Bear in mind when considering Freeman’s snap shares that these include the three games Coleman missed. Considering Coleman’s hyper efficiency last year, it will be hard for the Falcons to keep him off the field in 2017, especially in clear passing situations.

Conclusion

Despite these concerns, Freeman is being drafted as a strong RB1 (top-12 running back). He is going 11th overall in PPR leagues as the 6th running back off the board. However, his declining usage in an offense that is extremely unlikely to match last year’s numbers indicates that a late-first is a steep price to pay for Freeman.

While Freeman's red zone involvement always gives him touchdown upside, drafting a committee back -- even one as skilled as Freeman -- in the first round is a risky proposition.