Our Blog


Finding DFS Value Using Play-Calling Tendencies vs. the Spread

In my last article, I showed you why the spread matters less than you think for NFL DFS.

The spread is commonly used as a proxy for expected game script, so many DFS players tend to target a team’s passing or running game depending on the spread.

However, I found that while there’s a fairly strong correlation between the spread and the rate at which teams calls passes or runs (.47), site pricing and player ownership rates offset a lot of the value in targeting players based on the spread.

Here, I’m going to show which which teams are spread-agnostic, meaning they tend to stick to their play-calling tendencies regardless of spread, and which teams are spread-responsive, meaning they tend to adjust their play-calling tendencies based on the spread. Then I’ll look at which players benefit in terms of DFS value.

Note: All data is from the 2017 season, and all DFS metrics referenced are from DraftKings. Passing plays called include all plays that end in a pass attempt or sack taken, but not those that end in a quarterback scramble.

Expected Play-Calling Tendencies vs. the Spread

First, I computed expected pass-play percentage based on the closing spread for every game last season, then I calculated which teams passed or ran disproportionately in relation to the games they were favored in.

The x-axis of the graph below shows each team’s actual pass-play percentage versus expected, with teams farther to the right passing more.

The y-axis represents how much each team’s play calling correlated to the spread, with teams higher on the graph deviating more from what the spread would suggest.

Here’s what it means for teams that fall in each quadrant, going clockwise from upper right.

Pass-Heavy and Spread-Responsive (Upper Right)

These teams pass more than anticipated, but do so mostly in situations when the spread suggests they should, as opposed to doing so in all situations. As a result, these teams’ passing attacks will usually be good targets, but especially so as underdogs.

Using our NFL Trends tool, these teams’ quarterbacks posted a +0.62 Plus/Minus as underdogs last season. While this isn’t great, they also had a 4% upside chance at only 4.5% ownership. By comparison, these quarterbacks had a 6.8% ownership with only a 2% upside chance as favorites.

(Removing Oakland from the sample leads to a +1.1 average Plus/Minus and a 5% upside chance, but their situation could be quite different with Jon Gruden in charge in 2018.)

Pass-Heavy and Spread-Agnostic (Lower Right)

These are teams that pass more than anticipated, but do so regardless of what the spread would suggest. Unlike the teams in the upper right quadrant, quarterbacks on these teams are far better values as favorites.

These team’s quarterbacks posted a +1.8 Plus/Minus as favorites but -0.10 as underdogs. Generally, these are teams that will pass the ball no matter what, so they make for better targets in situations where they are likely to have success.

For example, Ben Roethlisberger‘s average Plus/Minus was 4.6 points higher and his ownership was less than half as a favorite in 2017 compared to an underdog.

Run-Heavy and Spread-Agnostic (Lower Left)

These backs performed better as underdogs, with a +0.20 Plus/Minus compared to -0.70 when they were favored. They averaged fewer points per game as dogs (10.82 vs. 10.9 as favorites), so it seems DraftKings undervalues these backs pricing as underdogs (or overvalues them as favorites).

These teams run a lot in every game, making paying up for their backs solely due to a good matchup unnecessary. These backs hit their upside score 11% of the time at only 5% ownership as dogs.

Look to target backs like Leonard Fournette in games where he’s an underdog for GPPs; he saw a 9.2-point jump in average Plus/Minus as a dog while being owned only 0.2% higher.

Run-Heavy and Spread-Responsive (Upper Left)

Finally, we get to teams that are generally run-heavy but vary their game plan as the situation dictates.

Of these nine run-heavy teams, the only ones to post a positive running back Plus/Minus last season (regardless of the spread) were Minnesota and New Orleans. There are likely different explanations for each of them that I will discuss in my next article. Most of the teams farthest to the left (that ones that are the most run-heavy) had struggling quarterbacks and offenses in general. As a general rule, these types of teams don’t have enough success for their running backs to justify their price tags.

Final Takeaways

A team being pass-heavy doesn’t guarantee that a team’s quarterback will be a value in DFS, nor does a team being run-heavy guarantee value for a team’s running backs. It’s always important to target players in situations where their salaries aren’t overly inflated. Since DFS site pricing accounts for the spread to some extent, some of the perceived value derived from play-calling tendencies in a particular spread or game script will be completely offset by player pricing. In tournaments, the same is also true regarding player ownership.

To recap, here are the main takeaways (note that the final three are somewhat counterintuitive):

  • Quarterbacks on pass-heavy, spread-responsive teams were undervalued as underdogs.
  • Quarterbacks on pass-heavy, spread-agnostic teams were undervalued as favorites.
  • Running backs on run-heavy, spread-agnostic teams were undervalued as underdogs.
  • Running backs on run-heavy, spread-responsive teams were overvalued in all situations.

It’s also important to keep in mind that how spread-agnostic or spread-responsive a team is depends on only on how they react in the face of game flow, but how well that team plays relative to expectations. For example, the Buffalo Bills ran more than anticipated relative to the spread, but that was due not only to play-calling philosophy, but also because they generally outperformed the spread’s expectation, finishing the regular season with a 10-6-1 against-the-spread record.

In future articles I’ll look at individual teams’ situations more in depth to see what this data means for 2018. Good luck!

Pictured above: Doug Marrone (left) and Leonard Fournette (right)
Photo credit:  Kim Klement – USA TODAY Sports

In my last article, I showed you why the spread matters less than you think for NFL DFS.

The spread is commonly used as a proxy for expected game script, so many DFS players tend to target a team’s passing or running game depending on the spread.

However, I found that while there’s a fairly strong correlation between the spread and the rate at which teams calls passes or runs (.47), site pricing and player ownership rates offset a lot of the value in targeting players based on the spread.

Here, I’m going to show which which teams are spread-agnostic, meaning they tend to stick to their play-calling tendencies regardless of spread, and which teams are spread-responsive, meaning they tend to adjust their play-calling tendencies based on the spread. Then I’ll look at which players benefit in terms of DFS value.

Note: All data is from the 2017 season, and all DFS metrics referenced are from DraftKings. Passing plays called include all plays that end in a pass attempt or sack taken, but not those that end in a quarterback scramble.

Expected Play-Calling Tendencies vs. the Spread

First, I computed expected pass-play percentage based on the closing spread for every game last season, then I calculated which teams passed or ran disproportionately in relation to the games they were favored in.

The x-axis of the graph below shows each team’s actual pass-play percentage versus expected, with teams farther to the right passing more.

The y-axis represents how much each team’s play calling correlated to the spread, with teams higher on the graph deviating more from what the spread would suggest.

Here’s what it means for teams that fall in each quadrant, going clockwise from upper right.

Pass-Heavy and Spread-Responsive (Upper Right)

These teams pass more than anticipated, but do so mostly in situations when the spread suggests they should, as opposed to doing so in all situations. As a result, these teams’ passing attacks will usually be good targets, but especially so as underdogs.

Using our NFL Trends tool, these teams’ quarterbacks posted a +0.62 Plus/Minus as underdogs last season. While this isn’t great, they also had a 4% upside chance at only 4.5% ownership. By comparison, these quarterbacks had a 6.8% ownership with only a 2% upside chance as favorites.

(Removing Oakland from the sample leads to a +1.1 average Plus/Minus and a 5% upside chance, but their situation could be quite different with Jon Gruden in charge in 2018.)

Pass-Heavy and Spread-Agnostic (Lower Right)

These are teams that pass more than anticipated, but do so regardless of what the spread would suggest. Unlike the teams in the upper right quadrant, quarterbacks on these teams are far better values as favorites.

These team’s quarterbacks posted a +1.8 Plus/Minus as favorites but -0.10 as underdogs. Generally, these are teams that will pass the ball no matter what, so they make for better targets in situations where they are likely to have success.

For example, Ben Roethlisberger‘s average Plus/Minus was 4.6 points higher and his ownership was less than half as a favorite in 2017 compared to an underdog.

Run-Heavy and Spread-Agnostic (Lower Left)

These backs performed better as underdogs, with a +0.20 Plus/Minus compared to -0.70 when they were favored. They averaged fewer points per game as dogs (10.82 vs. 10.9 as favorites), so it seems DraftKings undervalues these backs pricing as underdogs (or overvalues them as favorites).

These teams run a lot in every game, making paying up for their backs solely due to a good matchup unnecessary. These backs hit their upside score 11% of the time at only 5% ownership as dogs.

Look to target backs like Leonard Fournette in games where he’s an underdog for GPPs; he saw a 9.2-point jump in average Plus/Minus as a dog while being owned only 0.2% higher.

Run-Heavy and Spread-Responsive (Upper Left)

Finally, we get to teams that are generally run-heavy but vary their game plan as the situation dictates.

Of these nine run-heavy teams, the only ones to post a positive running back Plus/Minus last season (regardless of the spread) were Minnesota and New Orleans. There are likely different explanations for each of them that I will discuss in my next article. Most of the teams farthest to the left (that ones that are the most run-heavy) had struggling quarterbacks and offenses in general. As a general rule, these types of teams don’t have enough success for their running backs to justify their price tags.

Final Takeaways

A team being pass-heavy doesn’t guarantee that a team’s quarterback will be a value in DFS, nor does a team being run-heavy guarantee value for a team’s running backs. It’s always important to target players in situations where their salaries aren’t overly inflated. Since DFS site pricing accounts for the spread to some extent, some of the perceived value derived from play-calling tendencies in a particular spread or game script will be completely offset by player pricing. In tournaments, the same is also true regarding player ownership.

To recap, here are the main takeaways (note that the final three are somewhat counterintuitive):

  • Quarterbacks on pass-heavy, spread-responsive teams were undervalued as underdogs.
  • Quarterbacks on pass-heavy, spread-agnostic teams were undervalued as favorites.
  • Running backs on run-heavy, spread-agnostic teams were undervalued as underdogs.
  • Running backs on run-heavy, spread-responsive teams were overvalued in all situations.

It’s also important to keep in mind that how spread-agnostic or spread-responsive a team is depends on only on how they react in the face of game flow, but how well that team plays relative to expectations. For example, the Buffalo Bills ran more than anticipated relative to the spread, but that was due not only to play-calling philosophy, but also because they generally outperformed the spread’s expectation, finishing the regular season with a 10-6-1 against-the-spread record.

In future articles I’ll look at individual teams’ situations more in depth to see what this data means for 2018. Good luck!

Pictured above: Doug Marrone (left) and Leonard Fournette (right)
Photo credit:  Kim Klement – USA TODAY Sports

About the Author

Billy Ward writes NFL, MLB, and UFC DFS content for FantasyLabs. He has a degree in mathematical economics and a statistics minor. Ward's data-focused education allows him to take an analytical approach to betting and fantasy sports. Prior to joining Action and FantasyLabs in 2021, he contributed as a freelancer starting in 2018. He is also a former Professional MMA fighter.