Behind the process: Why Odell Beckham Jr.'s 2018 TD projection is so low

So, you think Odell Beckham Jr. is a sure thing for double-digit touchdowns in 2018? Our projections aren't so convinced. AP Photo/Bill Kostroun, File

"Why so low on Odell Beckham Jr.?"

I've been posting player and team projections on Twitter for several years, but not many have fueled a reaction like the one I got when I dropped this graphic showing 2018 receiving touchdown projections:

Not only were fans wondering, "Why is my favorite team's wide receiver not here?" (often in all caps), but players also jumped into the conversation. Davante Adams (projected first with 10) was pleased. DeAndre Hopkins (also with 10) still felt he was too low. Ha Ha Clinton-Dix seemed to think I was disrespecting the top guys (more on that later).

Perhaps the biggest complaint was related to Beckham's projection.

Beckham showed up 15th on the list with seven projected touchdowns. This is a curious number for a player who reached 10 scores in each of his first three NFL seasons and who had three in four games in 2017. Admittedly, the number surprised me as well when I did my initial projection run. It made sense after a lengthier look.

And when I say "lengthier," I'm not underselling it. There is a lot that goes into every projection I compile and post here at ESPN and on Twitter. The fact is, a player's past production is only one piece of the puzzle. In Beckham's case, the drastic dip has raised questions -- I'm a reasonable guy so I totally get it -- which is why this seems like as good a time as any to take you through my process.

This piece will break down the path to Beckham's 2018 touchdown projection and hopefully give you a better understanding of how the projections that you see each week when setting your lineups are compiled.

Scoring a lot of touchdowns is hard

The best way I can kick off any analysis of touchdowns is to link to this piece I wrote back in 2015. If you take one thing away from this article, it should be the conclusion that touchdowns are almost completely fueled by opportunity, not skill. Skill gets you the snaps, routes and targets necessary to score touchdowns, but, with few exceptions, "scoring touchdowns" is not a special skill.

This is important to understand in the evaluation of all players, but becomes especially important when projecting a player who has beaten the odds in the touchdown category throughout his career. Beckham caught 12 touchdowns as a rookie, 13 in 2015 and 10 in 2016, so, sure, it's easy to simply project him for 11 or 12 scores in 2018 and wipe your hands clean. It's much tougher to acknowledge that (A) three seasons is actually a small sample; and (B) history suggests his current pace is unsustainable.

Let's get into that history.

Since 2007, there have been only 104 instances of an NFL player reaching double-digit receiving touchdowns during the regular season. That works out to 10.6 per season, though only three players achieved that goal in 2017 and five in 2016. During the 11-year span, only one player has done it five times (Rob Gronkowski) and four have achieved it four times (Larry Fitzgerald, Calvin Johnson, Brandon Marshall, Jimmy Graham). Even the great Antonio Brown -- fantasy's top-scoring wide receiver each of the past four seasons and a top-five scorer the prior two years -- has only three under his belt. He joins seven other receivers in that department, meaning a grand total of 13 players have posted more than two 10-plus touchdown seasons during the past 11 years.

Because Beckham posted three of these seasons in a row, I dove a little deeper to see how close together players were able to pack campaigns with double-digit scores. Since 2007, there have been only two instances in which a player managed four double-digit receiving touchdown seasons over a five-year period. Both were Gronkowski (2010-14, 2011-15). If Beckham hits double digits in 2018, he'll be the third. That's right, Beckham and Gronk alone at the top.

By the way, if I do this same "four in five years" exercise but lower the threshold to nine touchdowns, eight seasons fit the bill. For a minimum of eight scores, its 19. If it's not already apparent, consistently posting big touchdown totals is difficult.

There's more to it than skill

Of course, to this point, Beckham has gotten the job done. He has exceeded his "expected" touchdown total by a drastic amount. His career regular-season OTD -- a number derived by evaluating the likelihood of a touchdown based on the line of scrimmage and target location -- sits at 27.5. Beckham has 38 touchdowns during the span. He's caught 20 of 48 end zone targets, whereas an average player would have caught 16.5.

I know what you're thinking: Beckham is so good that it's reasonable to expect him to exceed expectations in the touchdown department. The link I shared earlier explains why that's a dangerous assumption, but I took an even deeper look at this.

Beckham has posted OTDs of 8.8, 8.0, 8.9 and 1.9 with corresponding touchdown totals of 12, 13, 10 and 3, respectively. In other words, he exceeded his expected touchdown total by at least 1.1 all four years. Since 2007, there have been only four instances in which a player exceeded his expected touchdown total by more than one score in five consecutive seasons (Dez Bryant, Vernon Davis and Antonio Gates twice). Again, over time, opportunity is the variable that matters and players are simply unable to sustain high totals without it.

There's also the elephant in the room pointing to the fact that there are many elite or top-end players who have underwhelmed in the touchdown department (think Andre Johnson and Julio Jones), even when in situations with good quarterbacks. Despite massive target volume and elite skill, both Johnson and Jones went through stretches of their careers where they simply didn't get the ball much near the goal line, which led to disappointing touchdown totals.

Speaking of which, a quick aside. I also got a lot of pushback on Jones, whom I have projected for eight touchdowns in 2018. Jones scored only three touchdowns last season, so eight does sound like a lot on the surface, especially considering he reached eight only once over the past five years. However, a closer look shows that Jones totaled a 20.3 OTD and managed 20 end zone targets when he scored 22 touchdowns from 2013-16. That's right: he actually exceeded expectations, but wasn't getting enough volume near the goal line. Last season, Jones hit a career-high 8.6 OTD (fourth highest among wide receivers) and handled 16 end zone targets (seventh most). The 5.6 gap between his touchdown total and OTD was easily highest in the NFL. Put another way, after years of exceeding expectations despite underwhelming volume, Jones got the volume but was extremely unlucky in 2017. History all but guarantees he'll score a lot more often if he sees similar usage in 2018.

The quarterback factor

OK, back to Beckham. A look at the players who have exceeded their OTD by the larger margins shows a pretty clear trend: elite quarterback play. Beckham has exceeded his expected total by 10.5 since he entered the NFL, which is the league's 16th-best mark since 2007. Of the 15 players above him, four played significant snaps for the Packers (Greg Jennings, Jordy Nelson, James Jones, Randall Cobb), three for the Patriots (Gronkowski, Randy Moss, Wes Welker) and three for the Saints (Marques Colston, Darren Sproles, Robert Meachem). The others are Doug Baldwin, Jeremy Maclin, Julius Thomas and Anquan Boldin. By the way, if we go further down the list past Beckham, we see the likes of Terrell Owens, Lance Moore, Kenny Stills and Jimmy Graham all in the top 23. There's some decent talent in this group, but it's certainly not a list of the best players in the league over the past decade. There's a clear trend pointing toward the league's elite quarterbacks, and Eli Manning is not in that group. In fact, since 2007 Giants receivers as a whole (including Beckham's plus-10.5) are 1.5 below their expected touchdown total.

Size and durability

The next critical aspect of Beckham's projection is his size. Beckham is listed at 5-foot-11, 198 pounds, which history shows is not a frame that play-callers will build a red zone offense around. Usage near the goal line does correlate with both height and weight. As the article notes, a player of Beckham's size in an average offense with a 20 percent target share will usually end up with an OTD (or expected touchdown total) of 7.0.

Another roadblock to high production for Beckham is his shaky durability. In four seasons, Beckham has appeared in all 16 of the Giants' regular-season games once and has missed a grand total of 17 games (16 due to injury and one to suspension). He's suffered multiple hip and hamstring injuries and two different left-ankle injuries limited him to four games in 2017. Beckham is only 25 years old, so this shouldn't be considered a major concern, but injuries are a huge issue across the league (for some more than others) and it's something that should be baked into all projections. For example, I don't project any running backs for more than 15 games played.

Team and scheme

That covers a lot of what we need to know about Beckham, specifically, but more so than any other sport, a player's usage and production is tied to his teammates and scheme.

Let's start with the Giants' recent inability to score points and their coaching changes.

New York ranked 31st in points (29th in offensive touchdowns) in 2017, and yes, I know that Beckham missed 12 of those games. But in 2016, the Giants ranked just 26th in points (22nd in offensive touchdowns) with Beckham. A trend during the Ben McAdoo area was a massive percentage of the team's touchdowns being scored via the pass. During his four seasons as offensive coordinator or head coach, McAdoo's percentages were 70, 88, 82 and 77 percent. That average of 79 percent would've been fourth highest in the NFL last season and his scheme certainly helped inflate the team's passing touchdown total. During the seven seasons leading up to the McAdoo hiring, the Giants averaged 62 percent and never exceeded 67 percent in a single season (67 percent happens to be the NFL-wide average over the past three years). New head coach Pat Shurmur's offenses average 66 percent, including 62 percent over the past six years. New OC Mike Shula's offenses averaged a 63 percent rate during five seasons in Carolina.

The change from McAdoo to Shurmur and Shula may lead to more scoring, but we can't reasonably project a massive jump after two poor seasons and no quarterback change. Additionally, a larger share of the touchdowns figure to be of the rushing variety. I have the Giants at 33 projected offensive touchdowns (27th highest in the league), 72 percent of which are passes (seventh highest). That works out to 24 passing touchdowns and our next step is to figure out who will catch them.

As the Giants' roster currently stands, Beckham will join Sterling Shepard, Brandon Marshall and Evan Engram as the team's primary targets.

Though he did score eight touchdowns as a rookie, slot man Shepard hasn't been targeted a ton near the goal line (9.0 OTD in two seasons) and isn't a major threat to Beckham at 5-foot-10, 194 pounds. With a 17 percent target share, I have him at four touchdowns.

The other two, however, are big factors near the end zone. Engram handled nine end zone targets (fourth most among tight ends), posted a 5.8 OTD (sixth) and scored six touchdowns (seventh) as a rookie. He's 6-foot-3, 234 pounds and, assuming similar 2018 usage, it's hard to imagine he falls below the five-to-seven range in the touchdown department. Giants tight ends combined for 19 touchdowns in three seasons (6.3 per year) during Beckham's three 10-plus touchdown campaigns ... and that was with Larry Donnell and Will Tye soaking up most of the work. Engram is much better and much more involved.

Marshall is more of a wild card because we don't know if he'll make the 53-man roster (he could be released once he passes a physical). Of course, since we're talking about process here, this gives us an opportunity to learn why his presence would have a large impact on Beckham's projection. Since 2007, Marshall has handled 181 end zone targets and has posted a league-high 89.1 OTD. Both are easily highest in the NFL during the span. In fact, only six other players have 100-plus end zone targets during those 11 seasons. Marshall paced the NFL in end zone targets four times over the past six seasons in which he's played in 10-plus games and finished lower than seventh in the category only once (2010) since 2007 when doing so. I currently have him down for only an 11 percent target share, but 6-foot-4, 224-pound Marshall is one of the all-time most-utilized targets near the goal line and that doesn't figure to change if he's around in 2018. If he's cut, Beckham's touchdown share will rise, but at the moment, Marshall's presence must be factored into the equation.

So, at the moment, we have Beckham with 170 targets and seven touchdowns, Shepard with 103 targets and four scores, Marshall with 67 targets and three touchdowns and Engram with six scores on 122 targets. That's 20 of our 24 touchdowns, leaving scraps for inevitable scores by other players. The NFL-wide average for receiving touchdowns by running backs is 3.1 during the past five years and the Giants' rate is 1.6 during the span. In addition to the backs, the likes of Roger Lewis, Cody Latimer, Rhett Ellison, etc. will certainly chip in with a handful of scores.

If you ask me, that's a perfectly reasonable distribution of 24 passing touchdowns. If you think the Giants' offense will be better than I do (Manning is averaging 25.6 passing touchdowns per season in his career, after all), you should feel inclined to bump up everyone's numbers. If you think Beckham will still handle upward of half of the team's touchdown receptions, you'll need to be prepared to project drastically low scoring rates for his teammates.

So there you have it. A thorough dive into most of what drives a single player projection. Granted the Giants' schedule and draft selections, injuries around Beckham on the depth chart and to opposing defenses, and other news and notes will lead to tweaks, but we've at least achieved a reasonable 2018 touchdown projection for OBJ.

To wrap up, I want to address the inevitable "I'm going to bookmark this and check back in December" commentary. I don't know the future. No one does. Projections of this form are meant to be the "most likely outcome." Sometimes the most likely outcome has a 10 percent chance of hitting right on the nose, but is still better than betting on any other possible outcome. Variance, injuries and other unpredictable factors will always loom large in the NFL, which is why ignoring the noise and focusing on what you can predict is so important. Trust the process (a good process) and don't focus too much on a variance-distorted outcome. Last year, I would've been laughed out of the room had I projected Beckham for three touchdowns. Technically, I would've been correct, but only thanks to a poor, flawed process (it's unreasonable to project a player of Beckham's caliber for more than a missed game or two, let alone 12).

Yes, Beckham may exceed (or fall short of) my 2018 projected total. He may break through historical trends, post 10-plus scores for the fourth time in five years and go on to be a "better" touchdown scorer than Gronkowski. He could also miss more time due to injury and disappoint for the second year in a row. Or he could do what thousands have done before him and settle into a reasonable -- still high, but more reasonable -- touchdown pace.

Why so low on Odell Beckham Jr.? I'd say I'm right on the money.