The fifth—FIFTH—installment of the Sharknado franchise was officially announced last week. I’m sure producer David Michael Latt would prefer to be making multi-million dollar blockbusters with The Rock, but alas, he’s stuck doing straight-to-TV movies starring Tara Reid.
You know what though? Latt’s movies are popular, and they make money. The optimal path is not the only path to success.
I bring this up because sometimes in fantasy football you can’t play optimally. Everyone wants to stack a pair of stars like Andrew Luck and T.Y. Hilton, getting the highly valuable double-dip every time they connect. That’s the optimal play. The blockbuster.
On the other hand, no one ever wants to pair Luck with Frank Gore, or Hilton with Jack Doyle, due to their perceived conflicts. Every dropback for Luck takes away a carry for Gore; every pass to Hilton is a target Doyle didn’t get. These types of pairings—the Sharknados—aren’t ideal.
Unfortunately, sometimes that’s just the roster you end up with. Fortunately, in many cases, these pairings work well enough.
For this article, I looked back at 2016 to see how often, and under what circumstances, various suboptimal teammate pairings produced two starters in the same week. A “starter” is defined as a top 12 weekly finish at quarterback or tight end, and a top 24 weekly finish at running back or wide receiver.
For lack of a better term, and because it’s kind of catchy, I’m going to call these “anti-stacks.” Whereas stacking pairs players with correlated outcomes to maximize upside, anti-stacking pairs players with uncorrelated outcomes who can be started together, if necessary, in spite of that conflict.
Complete a mock draft in minutes with our free Draft Simulator 
QB/RB
In 2016, a quarterback and running back from the same team finished the week as starters 154 times. That’s pretty good—with only 12 starting quarterbacks each of the 17 weeks, a pair of QB/RB starters was only possible 204 times total. Put another way, only 50 quarterbacks finished as starters without a running back from their team also finishing as at least an RB2.
Notwithstanding the perceived civil warring between a team’s running and passing games, this makes sense. Good quarterback play is more likely to generate scoring opportunities and a lead, both positives for running back production. And with the last five seasons comprising five of the nine best scoring seasons ever, there’s been plenty to go around.
Meanwhile, the fact that a running back is stealing opportunities from his quarterback doesn’t matter because the volume isn’t particularly important for quarterbacks (case in point: Joe Flacco, Carson Wentz, Blake Bortles, and Carson Palmer were four of the top eight in passing attempts last year). This explains why teams with elite quarterbacks like Atlanta and New England routinely generated pairs of QB/RB starters, while teams with bad quarterbacks like Denver and Houston didn’t.
Boiling this all down, if you’re thinking about starting a quarterback and running back from the same team, I wouldn’t worry about whether or not to start the running back. Rather, focus on the quarterback—if it’s a shaky option, you’re probably better off streaming someone from another team.
The above debunks the perceived conflict a little, but QB/RB teammates still only finished as starters about 30% of the time. So how do we give ourselves the best chance to find the dynamic duos? The successful anti-stacks more often came from home teams (55% of the time) who won (66%) but scoring a lot of points was most important.
Half of the teams had an implied team total of 24.5 or greater, and a whopping 75% of teams scored at least 25 points. Accordingly, if you’re going to tempt fate with a QB/RB anti-stack, ideally they’re on a team that’s at home, favored to win, and has a good chance to score at least 25 points.
RB/WR & RB/TE
A starting running back came from the same team as a starting pass-catcher (WR/TE) 52% of the time last year. Since we’re defining “starters” as top 24 running backs and receivers, and there are only 32 teams in the league, it’s not surprising there was a lot of overlap. Still, it was interesting to find these supposedly conflicting positions co-existed more often than not.
It didn’t require the stars aligning, either: these instances involved the home team only 51% of the time and the winning team only 55% of the time, with the average implied team total a modest 23.5 points. Given that this isn’t a rare event—in fact, it’s the norm—the real question is whether this type of anti-stack caps your upside. Here, it’s a mixed bag.
In the RB/WR and RB/TE pairings that produced two starters, running backs and wide receivers finished 1-12 at their respective positions (RB1/WR1) about as often as they finished 13-24 (RB2/WR2), and tight ends finished 1-6 (TE1) roughly as often as they finished 7-12 (TE2).
However, of the 267 RB/pass-catcher anti-stacks that produced two starters, teammates finished as an RB1 and WR1 only about 33% of the time, and as an RB1 and TE1 only about 28% of the time. So, while these anti-stacks produce top starters with regularity, you’re more likely to land a top running back and pass-catcher by starting players from different teams.
WR/WR & WR/TE
Last year pass-catchers from the same team ended up as starters in the same game nearly 34% of the time. Not as frequent as the RB/pass-catcher anti-stacks, but obviously there’s enough passing volume to go around.
I was surprised to find WR/TE combos produced two starters much more frequently than WR/WR combos. If you think about it, though, there isn’t much of a conflict between these two positions. Only six tight ends averaged 7+ targets per game last year, leaving plenty of volume for their receiver teammates.
Additionally, the bar to be a “starting” TE is pretty low, requiring on average just 7.2 points (standard scoring). The other good news is it’s not hard to figure out where the successful WR/TE anti-stacks are coming from. There were five teams who produced WR and TE starters in the same game at least six times, and all either had a stud tight end (Carolina, Kansas City, Washington) or heavily involved their tight ends in the offense (San Diego, Indianapolis).
As for the WR/WR anti-stacks, this isn’t the place to get creative. 16 of the 32 NFL teams produced multiple starting receivers in a game just two or fewer times. The teams that most frequently supported two starting receivers (Atlanta, Green Bay, New Orleans, and Oakland) were top seven offenses last year, and except for Atlanta, had two-star receivers. A fifth team, Miami, also produced a high number of WR/WR starting pairs, which I’d chalk up to an easy schedule and fluky passing TD%.
RB/RB
This whole column has been about suboptimal teammate pairings, but with rushing volume and efficiency down in recent years, RB/RB anti-stacks are especially bad. There were just 58 times last season where two backs from the same team finished as starters, and in these instances, RB2 finishes were nearly four times more prevalent than RB1 finishes.
A handful of run-heavy teams (Atlanta, Buffalo, Oakland, Tennessee) and elite offenses utilizing pass-catching backs (New England, New Orleans, and Washington) accounted for the majority of these occurrences, so again, don’t get creative. Only start two RBs from the same team if you really must, and only if they’re on one of the few teams fitting the aforementioned categories.
Anti-Stacking & Anti-Fragility
The main purpose of this article was to consider when suboptimal pairings can be employed week-to-week, but I want to briefly touch on a season-long anti-fragility component to this concept. Most people who drafted Keenan Allen last year probably passed on Melvin Gordon a few rounds later, because were they really going to start two Chargers? For the same reason, most Alshon Jeffery owners probably didn’t go hard after Jordan Howard when he emerged as the Bears’ starter.
For those owners, Allen’s season-ending injury and Jeffery’s season-long disappointment killed them twice. Too many owners make bad decisions trying to avoid these types of situations.
I… have a friend who missed out on Odell Beckham Jr.’s historic rookie year because I, I mean he, “already had Victor Cruz.” Knowing you can sometimes roll out a pair of teammates, even if their production isn’t perfectly aligned, lets you focus on just adding talent and leaves you better prepared for the inevitable chaos of the NFL season.
Conclusion
The idea that teammates are competing for fantasy production in a zero-sum game is overstated. We’re in the midst of a historic scoring boom, where good offenses and good matchups can often support teammates supposedly in conflict. While anti-stacking isn’t something I necessarily target, these types of pairings shouldn’t be fantasy anathema either.
Subscribe: iTunes | Stitcher | SoundCloud | Google Play | TuneIn | RSS
Scott Cedar is a featured writer at FantasyPros. For more from Scott, check out his archive and follow him @scedar015.
