Stacking is essential to success in DFS guaranteed prize pool (GPP) tournaments. The average user who submits a handful of lineups on any one DFS slate doesn’t stack nearly enough in their lineups. The art of stacking is a way to increase variance, which is the only option that offers enough upside to win a GPP tournament. An overlooked approach is the game-level stack, which takes not only a quarterback and pass-catching option but also runs it back with a pass-catcher or running back from the opposing team.
We see in our introductory article on stacking that if we hit on the correct quarterback and wide receiver to stack, a player from the opposing team is typically going to be worthwhile to pair with this combination to apply more lineup correlation into your roster builds.
This provides inherent upside to our lineups — if we hit on the correct quarterback, two or more additional lineup spots are typically hit on by the simple fact that they are highly correlated to our quarterback play. If your quarterback is a miss, the rest of your lineup is likely in a similar situation, as it is hard to pull off a quality finish in a GPP contest with a dud at the signal-caller spot.
Focusing on this approach of identifying quarterbacks who could be hits, we have built a model to project who is most likely to finish as the highest-scoring quarterback on the main slate based on fantasy projections, opponent-adjusted grades and betting market lines. Utilizing correlations for how fantasy points are distributed at the game level, we can then see the correct framework for how to approach roster construction for a game stack.
WEEK 4 REVIEW
Stack games involving the Dallas Cowboys. This doesn’t really seem that complicated. For the third straight week, the leaderboard was littered with game stacks involving Dallas. The first non-Dak Prescott, Amari Cooper and Odell Beckham Jr. game stack finished in 37th place in the DraftKings Milly Maker. Only four lineups that finished in the top 100 did not have Prescott in at quarterback. When the chalk hits, it hits hard.
Our article last week got a little too cute in trying to avoid this matchup despite what the high total was forecasting. Prescott and Cooper were in just under 4% of lineups together, with the inclusion of Beckham dropping this number to 1.25% of lineups. Even if a game stack is chalky, there are still numerous paths to utilization that can produce outsized gains when you are correct. Sometimes, it is simply better to play the best plays while the rest of the market overthinks situations.
2020 has brought a significant adjustment to game totals after increased scoring. The mean total for Week 5 has once again reset the highest mean for any one week since we began tracking in 2007. This should cause game-level stack ownership to be incredibly spread out. There is no obvious chalk, meaning that if you have conviction about a certain play or game, load up on that situation. Let’s dive into the best stacks for Week 5, according to our model predictions, for games that could involve the highest-scoring players at each position.
The total backed off the opening 55-point print down to 54.5 to be the second-highest total on the main slate. Adding intrigue is the spread, which projects a tightly contested matchup throughout. This game projects to be the best shootout on an otherwise uncertain slate.
Our predictive models find value after the line movement, with a positive expected value play on the over. This is a quality signal for game-stack potential in DFS despite these offenses not sitting among the league's best in offensive efficiency metrics.