Our Blog


Designing an NFL Funnel Defense Metric

The term “funnel defense” has become popular within the fantasy football community, and it refers to defenses that are so good versus either the run or pass and so bad in the other that they tend to ‘funnel’ offensive production toward the area in which they are poor. It’s a simple concept: Why try to throw all over the Denver “No Fly Zone” secondary? Why not employ a run-heavy scheme instead? In fact, that is what offenses have done in recent years, and the Denver defense is notorious for funneling offensive production toward the running game.

This type of analysis is valuable, but there haven’t been many attempts to put data to it, perhaps because several factors would need to be measured at once.

  1. How good is a defense versus the running game and passing game? We can use a metric like Football Outsiders’ Defense-Adjusted Value Over Average (DVOA) to measure this.
  2. What is the difference between DVOA production? This would show the likelihood or probability of a funnel.
  3. How good is an offense in the running game and passing game? We can use DVOA here as well.
  4. What is the difference between offensive DVOA production?
  5. What are the offense’s typical run/pass splits?

There are other factors that can be added in, like the spread of the game — a team is much likelier to employ a run-heavy strategy when leading — but at the basic level we need answers to the five questions listed above. Let’s start with the defense and give a couple examples of funnel defenses through Week 4.

2017 Defensive Funnels

The two biggest funnel defenses so far — and they’re on the opposite end of the spectrum — are the Miami Dolphins and Jacksonville Jaguars. Here are their DVOA marks for both the running game and passing game this season:

  • Dolphins: +53.3 percent pass DVOA (31st), -33.6 percent run DVOA (2nd)
  • Jaguars: -45.3 percent pass DVOA (1st), +13.8 percent run DVOA (32nd)

Even though the Dolphins aren’t last in pass DVOA — that title belongs to the New England Patriots — the difference between their pass DVOA and run DVOA makes them the strongest pass-flowing funnel. In contrast, the Jaguars are impressively first in pass DVOA thanks to their elite corners and strong pass rush and dead last in run DVOA. They are perhaps a historically strong run-flowing funnel through the first four weeks of the season, which is a big reason why Le’Veon Bell is projected to be chalky this weekend in our Models. In this defensive variable of our metric, the Dolphins (on a 0-100 scale) have a 100 rating as a pass funnel and a zero as a run funnel. The Jaguars are the exact opposite.

2017 Offensive Desires

For the offensive side, we not only have to blend offensive production using DVOA but also factor in offensive play-calling tendencies. For instance, the Giants are starting Wayne Gallman in Week 5 at the running back spot. Even if they are facing a run funnel, they seem unlikely to run at a high rate. The Giants have been the second-most pass-heavy team this season, throwing on 69.88 percent of their plays. Thus, the offensive part of the metric accounts for both DVOA production and play-calling desires. Here are some examples of teams on either side:

  • Colts: -44.3 percent pass DVOA, -23.2 percent run DVOA, 47.7 percent run rate, 52.30 percent pass rate
  • Giants: +16.7 percent pass DVOA, -34.6 percent run DVOA, 30.12 percent run rate, 69.88 percent pass rate

The Giants are an extreme example of a team that could easily be funneled into the passing game but would not be funneled into the run game. They are good at passing and their offense does it a lot. The Colts, meanwhile, have a more even blend of running and passing, but they’re so terrible in the pass game that they could theoretically be funneled into running. Of course, that may not be true given their likelihood of being down in games, but, again, this metric will not yet account for Vegas data. Hopefully later.

Week 5 Run Funnel Predictions

Let’s get to Week 5. Here are the offensive ratings, defensive ratings, and then funnel ratings for each team. It is designed to be on a 0-100 scale showing how much a team should be funneled toward the run or pass. Take the Giants again: As mentioned above, they have a strong desire to pass and they’re good at it. The Chargers don’t have a strong funnel defense either way, but because of the Giants’ offensive marks they have a strong pass funnel prediction. On the other side, the Steelers are balanced in their play-calling and solid on both sides of the ball. But because the Jags are such a strong run funnel the Steelers are likely to be funneled toward the run. One more example: The Tennessee Titans rate strongly on offense in the run game, but they face a strong pass funnel. Thus, they have a moderately high pass funnel rating. Here are the marks for all main slate matchups (click to expand).

A couple notes about the flaws of this original metric idea: First, this does not incorporate historical data. Ideally, I would rate all of these teams based on the entire history of DVOA data to show historically how strong a funnel each team is in the broad perspective. The Jaguars have a 100 run funnel rating, but that may not be the case historically — perhaps they’re a 90 or even lower. Second, as mentioned above, this does not yet account for game script. Based on this metric, the Jaguars are likely to be funneled toward the run game, but they’re also +7.5 road dogs in Pittsburgh. The offensive and defensive tendencies/marks of these two teams dictate a run-heavy strategy, but that may be thrown out the window if the Jags get down big early. Incorporating Vegas data will likely improve the predictive power of the metric.

We’ll see how things go in Week 5. While this metric highlights some things we already sense — the Pittsburgh Steelers are likely to use Bell a ton — it also highlights less obvious outcomes. For example, if the Giants are likely to be funneled heavily toward the passing game but Odell Beckham Jr. is also likely to see shadow coverage from the Chargers’ Casey Hayward, perhaps taking shots on Evan Engram in guaranteed prize pools would be wise. Or maybe being heavier on Leonard Fournette than the field could pay off, even though he’s a road dog. Or perhaps you should bet the over on the Giants-Chargers game given the likelihood of a ton of passing from both teams. We’ll see. Good luck this week, and use our suite of Tools to research these players for yourself.

 

The term “funnel defense” has become popular within the fantasy football community, and it refers to defenses that are so good versus either the run or pass and so bad in the other that they tend to ‘funnel’ offensive production toward the area in which they are poor. It’s a simple concept: Why try to throw all over the Denver “No Fly Zone” secondary? Why not employ a run-heavy scheme instead? In fact, that is what offenses have done in recent years, and the Denver defense is notorious for funneling offensive production toward the running game.

This type of analysis is valuable, but there haven’t been many attempts to put data to it, perhaps because several factors would need to be measured at once.

  1. How good is a defense versus the running game and passing game? We can use a metric like Football Outsiders’ Defense-Adjusted Value Over Average (DVOA) to measure this.
  2. What is the difference between DVOA production? This would show the likelihood or probability of a funnel.
  3. How good is an offense in the running game and passing game? We can use DVOA here as well.
  4. What is the difference between offensive DVOA production?
  5. What are the offense’s typical run/pass splits?

There are other factors that can be added in, like the spread of the game — a team is much likelier to employ a run-heavy strategy when leading — but at the basic level we need answers to the five questions listed above. Let’s start with the defense and give a couple examples of funnel defenses through Week 4.

2017 Defensive Funnels

The two biggest funnel defenses so far — and they’re on the opposite end of the spectrum — are the Miami Dolphins and Jacksonville Jaguars. Here are their DVOA marks for both the running game and passing game this season:

  • Dolphins: +53.3 percent pass DVOA (31st), -33.6 percent run DVOA (2nd)
  • Jaguars: -45.3 percent pass DVOA (1st), +13.8 percent run DVOA (32nd)

Even though the Dolphins aren’t last in pass DVOA — that title belongs to the New England Patriots — the difference between their pass DVOA and run DVOA makes them the strongest pass-flowing funnel. In contrast, the Jaguars are impressively first in pass DVOA thanks to their elite corners and strong pass rush and dead last in run DVOA. They are perhaps a historically strong run-flowing funnel through the first four weeks of the season, which is a big reason why Le’Veon Bell is projected to be chalky this weekend in our Models. In this defensive variable of our metric, the Dolphins (on a 0-100 scale) have a 100 rating as a pass funnel and a zero as a run funnel. The Jaguars are the exact opposite.

2017 Offensive Desires

For the offensive side, we not only have to blend offensive production using DVOA but also factor in offensive play-calling tendencies. For instance, the Giants are starting Wayne Gallman in Week 5 at the running back spot. Even if they are facing a run funnel, they seem unlikely to run at a high rate. The Giants have been the second-most pass-heavy team this season, throwing on 69.88 percent of their plays. Thus, the offensive part of the metric accounts for both DVOA production and play-calling desires. Here are some examples of teams on either side:

  • Colts: -44.3 percent pass DVOA, -23.2 percent run DVOA, 47.7 percent run rate, 52.30 percent pass rate
  • Giants: +16.7 percent pass DVOA, -34.6 percent run DVOA, 30.12 percent run rate, 69.88 percent pass rate

The Giants are an extreme example of a team that could easily be funneled into the passing game but would not be funneled into the run game. They are good at passing and their offense does it a lot. The Colts, meanwhile, have a more even blend of running and passing, but they’re so terrible in the pass game that they could theoretically be funneled into running. Of course, that may not be true given their likelihood of being down in games, but, again, this metric will not yet account for Vegas data. Hopefully later.

Week 5 Run Funnel Predictions

Let’s get to Week 5. Here are the offensive ratings, defensive ratings, and then funnel ratings for each team. It is designed to be on a 0-100 scale showing how much a team should be funneled toward the run or pass. Take the Giants again: As mentioned above, they have a strong desire to pass and they’re good at it. The Chargers don’t have a strong funnel defense either way, but because of the Giants’ offensive marks they have a strong pass funnel prediction. On the other side, the Steelers are balanced in their play-calling and solid on both sides of the ball. But because the Jags are such a strong run funnel the Steelers are likely to be funneled toward the run. One more example: The Tennessee Titans rate strongly on offense in the run game, but they face a strong pass funnel. Thus, they have a moderately high pass funnel rating. Here are the marks for all main slate matchups (click to expand).

A couple notes about the flaws of this original metric idea: First, this does not incorporate historical data. Ideally, I would rate all of these teams based on the entire history of DVOA data to show historically how strong a funnel each team is in the broad perspective. The Jaguars have a 100 run funnel rating, but that may not be the case historically — perhaps they’re a 90 or even lower. Second, as mentioned above, this does not yet account for game script. Based on this metric, the Jaguars are likely to be funneled toward the run game, but they’re also +7.5 road dogs in Pittsburgh. The offensive and defensive tendencies/marks of these two teams dictate a run-heavy strategy, but that may be thrown out the window if the Jags get down big early. Incorporating Vegas data will likely improve the predictive power of the metric.

We’ll see how things go in Week 5. While this metric highlights some things we already sense — the Pittsburgh Steelers are likely to use Bell a ton — it also highlights less obvious outcomes. For example, if the Giants are likely to be funneled heavily toward the passing game but Odell Beckham Jr. is also likely to see shadow coverage from the Chargers’ Casey Hayward, perhaps taking shots on Evan Engram in guaranteed prize pools would be wise. Or maybe being heavier on Leonard Fournette than the field could pay off, even though he’s a road dog. Or perhaps you should bet the over on the Giants-Chargers game given the likelihood of a ton of passing from both teams. We’ll see. Good luck this week, and use our suite of Tools to research these players for yourself.