I don't always "get" math.

There was really nothing scarier to me in grade school than when I went from breaking records on a race car math game (just addition and subtraction) to falling to the middle of the pack when we hit long division and then stalling out completely when we got to anything beyond that.

Algebra? Eh. Geometry? Nope. Trig? Haha. Not a chance. I mean, sure, I was a B-minus level math student, so I didn't flirt with failure, but I could just never get things to click with me in high school.

Yet, here I am writing for numberFire. Because sports make sense to me.

And, sure, I believe in the eye test to support (or combat) what the statistics have to say, but it's silly not to account for something like regression -- especially when we can rely on JJ Zachariason, did last year, I'm going to compare Reception NEP to touchdowns to see which players "should have" earned fewer touchdowns in 2016 than they actually produced.

The Process

By combining a player's Reception NEP (expected points added just on catches) and comparing it to his actual touchdown number, we can essentially find "expected touchdowns." So that's what I did (from 2012 to 2016).

Here's how the chart and trendline look.

So, a receiver with a Reception NEP of 100.00 would be expected to score around 8 touchdowns, and a Reception NEP of 60.00 should be tied to a 4-touchdown season.

2016 Red-Flag Candidates

If we go back and apply this to the 2015 players who should have scored fewer touchdowns than they actually did in 2015, we find that they did, in fact, get worse at scoring touchdowns in 2016 in terms of touchdowns per catch (if we look at players who had at least 25 catches in 2016 so as not to make any conclusions from Rueben Randle).

None of this is to say these guys were bad. It's just that they regressed, just like the model suggested they should have.

It's tough to outperform math consistently, even for some of these top-tier receivers. Of the three players who scored 14 times in 2015, only Baldwin got to even half of that mark in 2016.

And if you're wondering where Odell Beckham (13 touchdowns in 2015), DeAndre Hopkins (11), A.J. Green (10), and Antonio Brown (10) are on this list, first of all, that's one hell of a memory you have, but second of all, they didn't make this list because their totals were a little less outlierish and they deserved to score more than the above list did.

Each of those players (Beckham, Hopkins, Green, and Brown) had a top-seven Reception NEP in 2015.

Of the list that should've regressed, only Brandon Marshall (130.72) and Allen Robinson (123.81) were top-seven by Reception NEP in 2015, and well, they didn't produce a Reception NEP that suggested they should have scored 14 times (which -- based on this model -- would have been around 190.00).

So, it's important not to see a player who scored double-digit touchdowns and expect them to regress negatively the following year. That's not the point in all of this. The point is that not every player with a big touchdown total should have gotten there, according to NEP.

With all of that said, who should be expected to take a step backward in the touchdown column in 2017?

2017 Regression Candidates

This is the list of the 14 receivers who outproduced their touchdown total by at least two full scores, according to this method.

Now, we know that certain players such as Jordy Nelson, Dez Bryant, and Antonio Brown have had a longstanding ability to score touchdowns at an above-average rate, but for some possible early-round fantasy picks -- Davante Adams, Donte Moncrief, Tyreek Hill, and Michael Thomas -- banking on them to repeat in 2017 could be a losing gamble.

Adams and Nelson are both tied to Aaron Rodgers and one of the best passing attacks in football, but each outperformed their expected touchdown total by at least 4.5 scores. Nelson -- as usual -- was in the mix for roughly double-digit scores, but Adams' touchdown per catch rate went from 0.05 in his first two seasons to 0.16 in 2016.

They each were one of just seven wide receivers to see at least 20 red zone targets in 2016. Again, in the Green Bay Packers offense, you can make the case that they see elevated red zone looks again, but if not, their touchdown totals will come down significantly.

Moncrief had 10 red zone targets, and since 2009, no receiver managed to produce 7 or more touchdowns on such a low Reception NEP. In fact, Marlon Brown's 48.55 in 2013 was the only mark below 50.00. Moncrief's was 34.68.

Hill is stepping into a larger role with the Kansas City Chiefs this year, so things could be a bit different. Still, of the 236 receivers with at least 6 receiving scores in a season since 2009, only five have had a Reception NEP below 50.00. Moncrief is still last on the list, but Hill (44.82) is third after Dezmon Briscoe (40.77 Reception NEP, 6 touchdowns in 2011).

Thomas isn't quite so concerning, as the New Orleans Saints are losing Brandin Cooks and Thomas at least produced a hefty Reception NEP sum in 2016. He drew 19 red zone looks of his own, and that surely plays a part in this method (because a receiver isn't necessarily producing to get his team into the red zone but does benefit from being there).

As for the players who should expect positive touchdown regression in 2017, we have you covered.