Stacking.

We've all heard about it. We all know what it means with regards to fantasy football -- at least a little bit.

Pairing teammates (and sometimes opponents) together to maximize on correlated performance. Mix it all together, and you get stacking.

I'm not the first person to dive into this topic, and there have been fantasy football correlation matrices to help guide us on the journey, but those go only so far with regards to how connected positions are in certain circumstances.

It makes sense that a QB1 and a WR1 have some correlated performances, and it also tracks that a QB1 and a WR2 have correlated results. But can we go a step further into the details?

Let's dig in and see what we can find.

Note: This article was originally published on August 21st, 2021 on numberFire and was named Best Football Article of 2021 by the Fantasy Sports Writers Association.

Defining a Big Game

Firstly, I'm using half-PPR scoring (more specifically FanDuel scoring) for all of this analysis, which should help make it relatively applicable across all scoring formats.

I am examining single-game situations over the past three NFL seasons (2018, 2019, and 2020).

Each player's (or team defense's) game log is tagged with a QB1, RB1, WR1, WR2, WR3, TE1, and DEF designation. The ranks for each position are based on pre-game projections from numberFire's database. For example, whichever wide receiver is projected for the most points on his team for that given week is the WR1 in the database.

While this study won't directly account for daily fantasy salaries, it is largely implied, as it accounts for pre-game expectations. Also, bucketing players into salaries has its uses, but it can be restrictive, and this study is meant to be more of a high-level look at how players are actually interconnected within single NFL matchups.

I could also look at fantasy point averages, but that's not really the goal here. It's to see how often one "big game" is attached to another within the same game.

Based on a combination of past position-based scores in FanDuel Sunday Million winning lineups and rough percentile rankings, I have deemed a big game for each position to be as follows:

These single-game scores ultimately wind up as close to 85th-percentile outcomes across the positions if you'd rather look at it that way.

It's a good mark to cross, a strong game to post. These positional scores are common thresholds for players in lineups that won the Sunday Million. They're not too high, and they're not too low. It also is position-sensitive.

I'll now take a rather binary look at success rates, starting with what actually happens when a team's starting quarterback puts up a big game.

The Effect of a Big Game: Quarterbacks

Here is the impending hit rate of each other main position when a QB1 crosses the big-game threshold.

These are sorted by the hit rate within the particular split in which the QB1 has a breakout game.

Unsurprisingly, a quarterback's main target is the most likely beneficiary of a big game, so this isn't really surprising.

When a QB1 has a big game, his WR1 has had a big game 53.8% of the time across our sample, which is 23.8 percentage points higher than the overall probability that a WR1 qualified for a big game if we pluck a game log from the sample at random.

That a WR1 benefits from a big game shouldn't really surprise us. However, this helps show just how much more likely we are to have big days from our fantasy teams when stacks do, in fact, hit.

Notably, a quarterback's TE1 is the second most likely player to have a big game (again, 14 half-PPR points), followed by the Opposing WR1 and then the quarterback's WR2 when said QB1 passes the 25-point threshold.

This is potentially where a dynamic big-game cutoff score can be misleading.

In theory, a QB1-TE1 stack would produce fewer points when the stack does hit than a QB1-Opposing WR1 stack (simply because WR1s put up more points, on average, than TE1s).

More objectively, when a QB1 has a big game, his TE1 averages 12.0 points over the full sample, and in games in which both post a big outing, the average jumps to 20.6 points. (This helps exemplify why it's worth it to pair teammates and anticipate a double hit.)

But when a QB1 has a big game, the Opposing WR1 averages 15.3 points overall, and when the stack actually hits, that Opposing WR1 averages 25.7 half-PPR points. That's more points than when the TE1 hits, on average.

Which, then, is the better stack?

That's debatable and probably counter-productive to the study here, which is aiming to see which stacks do actually hit and which may be taken as fact but really don't pan out.

Of course, more points are better, but TE1s have a lower mark to reach for a big game, and 20.6 points from a TE1 is more valuable relative to replacement than 25.7 points from a receiver.

And, perhaps, most importantly, the odds of a breakout game from a TE1 shoot up nearly 20 percentage points when tied to a big game from his quarterback. The Opposing WR1 is just barely more likely to do so than if we rostered a WR1 from another game.

For these reasons, the QB1-TE1 stack, with context, gets a big green checkmark.

The Effect of a Big Game: Running Backs

Here is the impending hit rate of each other main position when a RB1 crosses the big-game threshold.

These are sorted by the hit rate within the particular split in which the RB1 has a breakout game. (Note: 84.3% of big games from running backs in the sample between RB1s and RB2s came from RB1s.)

This is pretty noteworthy in that big RB1 games kind of stifle eruption rates for others within that game.

The RB1-Opposing WR1 stack is most likely to work out, and while it's barely more likely than a random WR1 to have a big game (4.1 points), any small increases in odds are welcomed when we're looking at razor-thin margins in daily fantasy tournaments. (Seriously, over a full lineup, any probability boost is vital.)

Notably, the RB1-DEF stack isn't anything to be married to when building our weekly lineups, though we still see a bump over picking a random defense. It's viable but not a requirement.

We should also not be against RB1-QB1 or RB1-Opposing QB1 stacks, as those duos do hit at a higher rate than if we paired an RB1 to unrelated quarterbacks.

The Effect of a Big Game: Wide Receivers

Here is the impending hit rate of each other main position when a WR1 crosses the big-game threshold.

These are sorted by the hit rate within the particular split in which the WR1 has a breakout game. (Note: 56.8% of big games from receivers (across WR1s, WR2s, WR3s, and WR4s) came from WR1s.)

The WR1-Opposing WR1 stack is super common when we build game stacks or mini stacks, but the data does suggest that we aren't always getting a lot of leverage by forcing in-game stacks of WR1s (which is notable in case they're both at high salaries).

We're seeing another strong relationship between the WR1-QB1 stack, but it's important to remember the distinction here. Roughly 30% of the time when the WR1 blows up, his quarterback also does. But there's greater than a 50% chance that, when a QB1 erupts, his WR1 crosses the threshold. Either way, the QB1-WR1 stack is the go-to for a reason, but a WR1 is more likely to benefit from a QB1's performance than vice versa.

We actually see better-than-average odds of teammates hitting it big in WR1 breakouts for the QB, RB1, TE1, and WR2.

Here are the WR2 numbers. (Note: The odds of an WR2 breakout game are 16.0%, compared to 30.0% for WR1s. Though I think this can be viewed as a drawback to how I approached this, I do think it's important to remember that while -- yes, WR2s and WR3s can put up big games -- it's those projected for the bigger outputs that reach the number more consistently. And at a certain point, value goes only so far, and we should be hoping for massive point outputs from our receivers.)

It's obvious here with the WR2 where to start, and that's with his quarterback, which gives us a 17-point boost over expected with a big QB1 game.

Somewhat surprisingly, the WR2's depth-chart higher-up, his WR1, puts up big games more frequently than average in WR2 breakouts.

And then we get into four straight opponents -- the Opposing WR1, the Opposing QB1, the opposing WR2, and the Opposing RB1 -- who thrive more often, and I think this is where we start to gain a lot of information about stacking outside of the most common-knowledge angles.

QB1s and RB1s and WR1s can have big games because of their projected workloads, but when a WR2 actually gets there with a high point output, it's probably tied to a stronger game environment, which leads to more top-tier outings from other players in the game. That alone isn't groundbreaking if we step back and think about it, but putting odds to these numbers solidify what might otherwise be a misconception.

Yes, we have to remember that we're looking at higher-end WR2 games here than we are with more common WR1 games (because 18 points for a WR1 is a 70th-percentile performance and 18 points for a WR2 is an 84th-percentile performance), but at a certain point, points are required to succeed in fantasy football.

And now for WR3s. Note how improbable WR3s are to reach 18.0 FanDuel points -- 6.4%. I included them just because we're more likely (subjectively, at least) to consider a WR3 for stacking than we would be for an RB2 or a TE2.

It's hard to want to dig in too much here with such an improbable position to hit it big, but the most obvious notes are that we may want to skimp on WR3-WR1 or WR3-WR2 stacks and that WR3-RB1 stacks probably aren't worthwhile due to the WR3's production likely taking away specifically from scoring chances for the RB1.

The other note here is a lot of boosts to the opposing side, meaning WR3s -- like with WR2s -- could be ideal bring-back candidates when stacking the other team.

The Effect of a Big Game: Tight Ends

Here is the impending hit rate of each other main position when a TE1 crosses the big-game threshold.

These are sorted by the hit rate within the particular split in which the TE1 has a breakout game. (Note: 90.4% of big games from TE1s and TE2s have come from TE1s.)

The TE1-QB1 stack has a lot of appeal, which is to be expected, but the TE1 really doesn't correlate that well with other big games to a high degree.

The WR1 still has better-than-random probability to coincide with a big game when the TE1 hits it big, but similar to the WR2 and WR3s, perhaps the TE1 is best paired with bring-back stacks of the other team.

TE1s eruptions also have an adverse affect on RB1 breakout odds, likely due to touchdown requirements for each to explode.

The Effect of a Big Game: Defenses

Here is the impending hit rate of each other main position when a team defense crosses the big-game threshold.

These are sorted by the hit rate within the particular split in which the defense has a breakout game.

Yeah, so don't play any opponents with a team defense, and the only position with big-game boost odds when a defense goes off is with their RB1.

The Most Probable Stacks to Hit

Okay, there has been a lot to digest so far, and it can be hard to remember all the takeaways, which I'll summarize at the end the best I can.

But this might help as much as anything.

These have been the most common two-player (including defense) stacks coming from the same game.

This table also lists the probability that two players from random games would simultaneously qualify for big games so that we can see how the stack compares to picking random players.

I think it's overkill to explain these stacks in more detail at this point, and this table ultimately should help to act as a reference we can use when constructing lineups.

Stacking Summarized

Here are some of the most key findings from the piece:

- The QB1-WR1 Stack Is Actually the Gold Standard: This stack is common for a reason, and 53.8% of the time a QB1 gets 25-plus fantasy points, his WR1 gets to 18-plus.
- The QB1 Dictates a Lot About a Team: QB1 performance is always tied to offensive output, so they're naturally a key part of any stack. WR2s, WR3s, Opposing WR1s, Opposing QB1s, and Opposing WR2s see a significant boost in big-game odds when a QB1 hits.
- RB1s Are Their Own Breed: This applies both in the sense that RB1s make up 84.3% of big games from all running backs and that their performance isn't really strongly tied to other positions in the game.
- The QB1-RB1 or RB1-Opposing QB1 Stack Is Viable: Though there isn't a huge boost in big-game odds for passers tied to RB1s, there is still a slight increase in big-game hit rate when RB1s thrive. Given that this is likely to be an uncommon stack from opponents, it's a potential leverage stack we can use in daily fantasy tournaments.
- All Receivers Can Erupt, But It's Usually the WR1s Who Do: In the sample, 56.8% of big games from all receivers were from WR1s. This is obvious, but as someone who overestimates the ability for low-end receivers to put up big games, it's nice to see the data.
- WR2 and WR3 Breakouts Make for Bring-Back and Mini Stack Candidates: More so than with WR1s, WR2 and WR3 big games are tied to better stacking environments; WR1s can produce regardless, but it's the lower-end receivers who are more stack-sensitive.
- TE1 Big Games Aren't Always Great for the Offense: TE1 games likely revolve around touchdowns rather than yardage, so we should take care not to roster too many backs and receivers from the offensive side of the ball when stacking a touchdown-dependent tight end.
- Defensive Big Games Probably Ruin the Actual Game: Only RB1s see a boost (2.6 points) in big-game odds when a team defense nets 12-plus fantasy points.

The above author is a FanDuel employee and is not eligible to compete in public daily fantasy contests or place sports betting wagers on FanDuel. The advice provided by the author does not necessarily represent the views of FanDuel. Taking the author’s advice will not guarantee a successful outcome. You should use your own judgment when participating in daily fantasy contests or placing sports wagers.