Beneath the Surface: Using Analytics to Improve NFL Betting

Patrick Mahomes #15 of the Kansas City Chiefs signals for a two-point conversion against the New Orleans Saints. Chris Graythen/Getty Images/AFP

The analytics community has changed sports significantly over the last two decades. Football has been slower to adopt advanced metrics than the other major sports, but they’ve gradually infiltrated even the oldest-school coaching staffs.

We’ve seen attitudes about fourth down change as coaches, faced with a mountain of evidence, have begun taking more risks, rather than automatically sending out the kicking or punting units.

READ MORE: Benefits of Betting at Online Casinos

Stats have also changed the game for fans. Web sites and even TV broadcasts are incorporating more advanced statistics to enhance our enjoyment of the game. For gamblers, it also creates a trove of data to help decide which teams to pick.

While most of us don’t have the ability or time to build a simulation model or to run regressions to find new ways of evaluating players and teams, there are plenty of people out there who do, and we can take advantage of their work. Here’s a look at two of the numbers that have been shown to be most predictive of what will happen on the field.

Pressure, Not Sacks

Everyone has seen a team’s offensive game plan get derailed early, simply because the quarterback was forced to run for his life all game long.

Back in the day, we would look at sack totals to measure the impact of a team’s pass rush. As legendary Raiders’ owner and coach Al Davis said on NFL Films, “The other team’s quarterback must go down, and he must go down hard.”

READ MORE: Over 30 Million U.S. Nationals Expected to Wager on Super Bowl LVI

As it turns out, Al wasn’t completely correct. The fear of going down hard is often enough to disrupt a team. Quarterback hits and hurries may not result in a sack, but they can mess up the best-laid plans of an offensive coordinator, forcing earlier and shorter passes than were planned or flushing the quarterback from the pocket to throw on the run.

Analysis has shown that the best predictor of how many sacks a team will get in a given game isn’t the number of sacks they’ve gotten already … it’s their pressure rate. In other words, if your defense has been in the neighborhood of the quarterback, you’re probably going to get some sacks.

Yards, Not Points

This is similar to the sacks vs. pressure discussion we just outlined. We’re basically trying to predict relatively rare events. Considering that there are multiple players trying to get sacks on every passing play, the odds of getting one are pretty low.

Similarly, even the best offenses don’t get more than a few touchdowns a game. And fluke things happen every game to skew the numbers.

A defender may fall down, giving a receiver, who suddenly finds himself uncovered, a long touchdown. A quarterback may bobble the snap from center, given a defensive end an extra second to reach him and get a sack.

READ MORE: Fun Super Bowl LVI Prop Bets With Value

While it’s tough to accept at face value, looking at how often a rare thing happens isn’t the best way to predict it will happen again. That receiver will be playing a different team, on a different field, with different weather. So it’s unlikely that a defender falling down last week will hold any predictive value for next week.

Instead, researchers have found that being in the neighborhood is the best predictor of having something happen again.

If I get involved in a fender bender on the highway, that doesn’t necessarily mean I’m more likely to get involved in another one anytime soon.

READ MORE: How to Avoid the Hype in Super Bowl Prop Bets

Many accidents are unavoidable—a deer ran in front of me, or the other driver did something dumb and hit me. Others are just dumb luck. But if I enjoy going out joyriding any time it snows, or if I frequently drive drunk, I’m more likely to get into an accident in the future, regardless of how many I’ve had in the past. This works under the same rationale.

So yards per play is shown to be a better measure of an offense’s (and defense’s) strength than points scored. If a team is getting more yards per pass than anyone else, it’s going to show up on the scoreboard eventually.

Beyond the Box Score

The whole idea behind the analytics revolution is that there are things going on that haven’t been measured in the past.

That’s the idea behind these two stats, which both essentially say “ignore what actually happened in favor of what almost happened.” Look for the teams that are doing everything right, even if the results haven’t always shown up.