Stacking and correlation

The table above is NFL only. For baseball, hockey and basketball we have the reasoning below and no measured table yet.

What correlation means

Correlation describes how two players' fantasy scores move in relation to their own averages. Positive means that when one beats his average, the other tends to as well. Zero means one tells you nothing about the other. Negative means one player's big day tends to come with the other's bad day.

Stacking doesn't change your lineup's average score. Averages just add up. It changes the spread. Take two players who each swing about 8 points around their average. If their scores have nothing to do with each other, their combined score swings about 11.3 points (the square root of 8² + 8²). If they moved in perfect step it would swing 16. Real pairs land in between.

A wider spread means more huge nights and more awful ones. Tournaments pay for huge nights and don't care about awful ones, so stacking is a tournament tool first. In a cash game the bad nights cost you, so stack there only when the pair is also your best floor.

NFL: quarterback and pass catchers

One play pays both of them. A 40-yard touchdown catch on DraftKings is 5.6 points for the quarterback (1.6 for the yards, 4 for the touchdown) and 11 for the receiver (4 for the yards, 6 for the touchdown, 1 for the catch). That is 16.6 points from one snap. On FanDuel, with half a point for the catch, it is 16.1.

Yardage bonuses push the same way on DraftKings, which pays 3 points at 300 passing yards and 3 at 100 receiving yards. One big passing day can trigger both, which is 6 extra points. We couldn't confirm FanDuel's bonuses for this season: see the rules page.

Use the table to choose the pairing. The usual build is a quarterback with one or two of his pass catchers, and the table shows why: a lead running back's score moves far less with his quarterback's. Two cautions. Teammates compete for the same volume, so a third receiver from one offense needs a reason. And a defense facing your quarterback works against your own stack, the strongest negative pair in the table, so don't roster both.

Bring-backs

A bring-back is a player from the team facing your stack. The logic is game script: if your quarterback throws four touchdowns, the other team has to keep throwing to catch up. DraftKings' own guide calls this a game stack for shootouts: quarterback, receiver, opposing receiver.

It costs a roster spot and salary, and it moves your bet from one offense to the whole game. The opposing pairs in the table show the link is real but much weaker than a quarterback's link to his own pass catchers, so a bring-back has to earn its spot on price and role too.

MLB: hitter stacks

Runs come in bunches. A hitter on base is the next hitter's RBI, and one inning can pay several of your players at once. Five hitters from one team make ten pairs (5 x 4 / 2 = 10), so even a small link between neighbors adds up across a stack.

Stack hitters from the same team, and favor neighbors in the batting order, because the man hitting behind a player is the one who drives him in. DraftKings allows up to five hitters from one team. Don't roster hitters against your own pitcher: their good night is his bad one.

NHL: line stacks

A goal usually pays more than one skater. On DraftKings a goal with two assists is 8.5 + 5 + 5 = 18.5 points if all three players are yours, plus 1.5 for the shot on goal. That makes the same trio worth 20 on one puck.

The basic stack is a center and his two wingers from one line. A power-play stack is a second way in, and the best pairs share both a line and a power-play unit. Don't roster skaters against your own goalie. DraftKings lineups also need skaters from at least three teams, so plan a spot for a skater outside your stacks.

NBA: why stacking is weak

Basketball has one ball. Every shot a teammate takes is a shot your player doesn't, and every rebound one grabs is a rebound the other doesn't. The clean positive link is assists and points: a made basket can pay both passer and scorer. Most of the rest pulls the other way.

So in NBA, look at game environment instead of pairs. A fast game with a close expected margin means more possessions and keeps starters on the floor. When a starter sits, his minutes and shots go to teammates. That is a role change, which you handle in projections.

Reading the table on this page

Each number is a long-run average across many games, so it can't tell you what happens this Sunday. Use it to rank pairs and to check a story you already believe. A strong number on a bad player doesn't make him a good play, and a stack that costs you too much elsewhere can lower your total. Try the trade-off in the lineup builder.

Facts checked Nov 5, 2026.