fb-pixel
michael silverman

How do NFL oddsmakers set point spreads in a COVID-riddled season?

Wayne Parry/Associated Press

Tracking the ups and downs of an NFL point spread over the course of a week can be equal parts head-scratching and head-spinning.

Sometimes the spread is set once, never to budge.

Other times the number fluctuates for days, expanding or shrinking in half-point increments for no obvious reason.

Then there’s that random Thursday morning when you return from a walk without your phone and find your 4-point favorite has morphed into a 2-point underdog.

What gives?

How does the number get set each week?

What’s making that number rise and fall?

And what impact has this disjointed COVID-19 season had on the process?

Advertisement



We asked three top oddsmakers — Nick Bogdanovich, director of trading at William Hill US (Caesars), Johnny Avello, director of bookmaking at DraftKings, and Jason Scott, vice president of trading at BetMGM — to pull back the curtain and take some of the mystery out of how their 2020 point spreads are determined.

While Scott’s approach differs from the system Avello and Bogdanovich use, they all rely to some extent on the “zillion different factors” that Bogdanovich said affect a spread.



Avello and Bogdanovich begin with their own power ratings, in which all 32 NFL teams are ranked. When the wins and loses and any significant injuries are accounted for, those rankings are tweaked. The tweaks become more subtle as the season wears on and powerhouses and patsies settle into place.

The challenges of what happens in a sporting event during the COVID-10 era have been a challenge to companies like DraftKings.
The challenges of what happens in a sporting event during the COVID-10 era have been a challenge to companies like DraftKings.Charles Krupa/Associated Press

Avello uses a mathematical formula that includes the competitive strengths of each team, points scored, points against, strength of schedule, and margin of victory to come up with a rating system with a ceiling of 100. He says that number is reserved only for the great teams.

“Say you have the Colts-at-Browns game, the Colts with a 91 power rating and the Browns at 88.5 — that’s a 2.5-point differential, which is what the line would be on a neutral field," said Avello, who has been posting spreads in Las Vegas for 41 years.

Advertisement





As a result of the COVID-19 restrictions on fans attending games, only 15 of the 32 teams allow partial capacity in their stadiums (less than 1 percent to 24 percent).

“Under normal circumstances, the Browns have a 2.5-point home-field advantage,” said Avello. "But they’re going to have maybe about 12,000 fans or so, so I’ll give them a 1.5-point advantage, which brings them up to about 90 [power ranking], so the Colts would open with a 1-point advantage.''

In a normal year, teams such as Seattle, Kansas City, and Denver have larger-than-average home-field advantages, according to Bogdanovich. Others such as Jacksonville, the New York Jets, and Miami have close to zero advantage playing at home. Still, even in an empty stadium, a slight edge should exist for the home team simply because it knows its stadium better and the visiting team has to deal with the stresses of travel.

Among the other power-rating factors: coaches' reputations, head-to-head results in recent years, the playing surface (grass or turf), the start time (day or night), and motivations such as the last meeting between the teams, what happened in each team’s last game, or whether the team is contending for a playoff spot.

Injuries and other factors

The variance among bookmakers is low, almost nil, at the start of a week. Avello and Bogdanovich said they will look at other lines to make sure they’re near the same number and adjust if necessary.

Advertisement



If an early number is way off, said Bogdanovich, “You get straightened out pretty fast. If sharps [professional bettors] see a weak number, they’re going to attack it, whether it’s for $5,000 or $50,000."

According to Bogdanovich, early betting on a Sunday night is “very minimal. It’ll be .0001 percent of it. It’s nothing.”

In a case such as the Patriots-Broncos game, which was pushed back from last Sunday because of COVID-19 positives with New England, lines will be taken down or simply not posted until there’s clarity.

Impactful information about the upcoming week’s game is scarce right after a game; but the action, both literally and figuratively, quickly picks up.

For bettors, the absence of Cam Newton as of late has made it tougher to try and get a handle on where the Patriots stand.
For bettors, the absence of Cam Newton as of late has made it tougher to try and get a handle on where the Patriots stand.Barry Chin/Globe Staff

Weather plays a role as the season deepens. Scoring overall should go down as the weather worsens (important for over/unders — the total predicted points for a game — but not so much for the spread) but a home-field advantage might increase for teams in the Northeast or Midwest if a West Coast or dome team is visiting during a snowstorm or extreme cold spell.

Rain alone is usually not a big deal, said Avello. A team with a strong running game can gain an edge with sharp-cutting backs who know where they want to go and are able to more easily evade tacklers who can’t react as quickly on a muddy track.

Offenses reliant on an aerial attack can expect less success on a windy day, especially visiting quarterbacks and receivers unfamiliar with the wind patterns in a stadium.

Advertisement



Avello waits at least 48-72 hours prior to kickoff before factoring in temperature, precipitation, wind, and, say, a brewing nor’easter or hurricane.

Injuries, of course, are a key factor.

Your fantasy team is a different story, but unless a star player (usually a quarterback) is hurt, even an injury to a premier running back or wide receiver may not cause a significant shift in the number.

Being out because of COVID-19 is 2020′s unique twist.

To Avello and Bogdanovich, the reason (COVID-19) for Patriots quarterback Cam Newton’s absence before the Week 4 game against the Chiefs was irrelevant for point-spread purposes. The task for an oddsmaker is to determine the dropoff in performance between Newton and his backups, Jarrett Stidham and Brian Hoyer, and convert that into points.

‘“With Cam Newton, you have maybe a 4-point difference between him and his backups, Hoyer or Stidham — is it 4 points, is it 3? You have to ask, ‘What’s the situation? Are you playing a cupcake?’ Then it’s no big deal.”’

Nick Bogdanovich, director of trading at William Hill US

“With Cam Newton, you have maybe a 4-point difference between him and his backups, Hoyer or Stidham — is it 4 points, is it 3?" said Bogdanovich. “You have to ask, ‘What’s the situation? Are you playing a cupcake?’ Then it’s no big deal.”

William Hill waited until Wednesday to put up a line on the Patriots-Broncos game because of COVID-19 uncertainty, opening at Patriots minus-9.5 points. That line moved up to 10 Thursday when Newton was at practice.

The quarterback’s availability had already been factored into the opening line.

“I would say Newton is worth 3 points, that would be my guess,” said Bogdanovich.

Advertisement



Even though they’re new, COVID-19 absences are not alarming, based on Avello’s reasoning.

“The difference is that there is a relatively good chance that Cam Newton with COVID-19 will be back the next week, and if not the next week, then certainly the week after,” said Avello. “A major injury can put a guy out maybe 3-4 weeks, maybe an entire year.”

The timing of the disclosure of injury information is key. In these days of instantaneous information from TV, Twitter, and other forms of social media, the lag time between a team’s revelation of injuries and decisions on players' availability is growing shorter and shorter. The sportsbooks rely either on receiving tips from services with a track record of having reliable and early information or reacting immediately to breaking news.

‘“I would say Newton is worth 3 points, that would be my guess.”.’

Nick Bogdanovich, director of trading at William Hill US,on how Cam Newton moves the needle for gamblers

“I’m hoping to catch those injuries before Joe Q. Public finds them,” said Avello. “It’s not hours before, it’s 15 minutes, if that. Five minutes, one minute, 30 seconds — I don’t care.”

The information does not always pan out as expected, but nobody’s going to risk ignoring it.

Achieving balance

Given all this preparation and analysis, BetMGM’s Scott is modest enough to say he isn’t smart enough to know when and how much to move a point spread.

Instead of employing people to scour news sources and dive deep into football analytics, his digital team focuses on the betting side of the business: They develop computer algorithms to identify how bettors using BetMGM’s digital platform are betting on the point spread.

“Rather than us trying to be experts on [football analysis], we’ve identified a group of customers who, when they bet, the lines tend to move their way,” said Scott, who said that BetMGM’s system is common in European markets and more common in the US than some sportsbooks will acknowledge. "We use that data. Those bets will change the price. We will then be in front of the market.”

Scott is trying to find what he calls the “true price” of any NFL game, that point “where the sharps aren’t picking on either side because they don’t think they’ve got an advantage. They’ve taught us how to get there."

Bogdanovich called equal action a “pipe dream,” while Avello considers it to be “very rare” when a bookmaker can balance out a game.

Regardless of their system and philosophies, the bookmakers know that they tend to come out ahead by the end of any season thanks to the “vigorish,” which is essentially a fee the house collects from bettors for allowing them to bet.

That’s one point of information that’s usually left out of the point-spread equation for the average customer.



Michael Silverman can be reached at michael.silverman@globe.com. Follow him on Twitter: @MikeSilvermanBB.