How to bet the NFL: Using team ratings to calculate point spreads
We’ve got to start our journey here somewhere, and this is as close as we can come to “the beginning” of a sports year.
Many sports betting content sites have a “Beginner’s guide to sports betting” section now. Cool, but what about an “Intermediate’s guide to sports betting” area? On this Substack, we’ll try to fill the gap between “What’s a moneyline?”, and running a multi-tiered sports betting syndicate.
We’re less than two weeks away from the NFL season opener, but point spreads have been available for months. So, as we take our first step on the long road that is the 2024 NFL season, let’s begin with a reminder about not what a point spread is, but how it works.
Two key terms you’ll want to remember: Opening line and Market discovery
How NFL point spreads are made
Step 1: Win total markets open based on team schedules
NFL fans don’t get their team’s exact schedule until mid-May, when the Titans social media staff hits the streets for pure entertainment.
People who understand how the NFL schedule works know that the matchups are known long before that.
Oddsmakers - whomever is first to market - create projected win total “over/unders” for every team and release those to the public for betting with very small betting limits. An accumulation of small bets - with particular attention given to those by sharp bettors - give oddsmakers an idea about what the betting market thinks of their original numbers.
If a team is assigned a win total of 8.5, they’re deemed a .500 team or 50/100. If bettors pile on the over, that tells oddsmakers that they’ve underrated the team relative to the market. Maybe the oddsmakers end up being right, or maybe the market does.
As we read the current win total over/unders, there’s two ways to get to a current number. If the first sportsbook to offer win totals opened the Cardinals at 5.5 (-110) wins, and bettors pounded the over, that could be how the number got to 6.5 (-160) to the over, as it is at Bet365. In turn, they could have also opened them at 7.5 and seen more under money come in to bring the total down to 6.5, but not far enough to get down to an equal price of Cardinals -110 on either side of that total. Either way, this is called market discovery.
Step 2: Translating a win total into a rating
Since the Cardinals’ win total is not truly 7.5 or 6.5, it’s somewhere in-between. Technically, Arizona is expected to win 7.25 games this season.
There’s some slightly more complicated math that you’re likely not interested in, but one win is worth about 6% in a 17-game season.
This is important because we need to take the market’s first collective opinion of all the teams and create a rating for each team. Since grade school, we’ve always lived life on a scale out of 100, so for simplicity we talk about each team’s rating in that way.
Per the Cardinals, if they’re expected to win 7.25 out of 17 games, or 42.6% of their games, then they’re 42.6% likely to beat an average (50/100) team on any given Sunday.
Alas, it’s not THAT simple. Since not every team has the same strength of schedule. The Cardinals play in a slightly above-average division, the NFC West. But what would their win total look like if they played in the tough AFC North? Or, in the easiest division, the NFC South?
Step 3: How do we turn team ratings into a point spread?
You can’t take the Ravens (win total: 10.5) 61.8 rating and simply subtract the Cardinals 42.6 to create a point spread.
Why? Because teams aren’t expected to score 100 points.
So you need to estimate the number of points a team needs to win a football game, or what the highest possible point spread is for this season if the worst team ever played the best team ever. You can make that number whatever you want, but my number is 27. Three touchdowns and two field goals should win you an NFL football game more often than not, and things have to get awfully good for one team and bad for another to see an NFL game with a line of -27.
Therefore, 27.0 is the number I’m using to divide a team’s rating into.
Ex:
Ravens (0.618 * 27 = 16.68)
Cardinals (0.426 * 27 = 11.5)
Using this formula, the Ravens are 5.18 points better than the Cardinals on a neutral field. If they played in Baltimore, with two points of home-field advantage, the Ravens would be -7.15. In Arizona, with just a point of home-field advantage for the Cardinals, the Ravens would be -4.
Step 4: Oddsmakers risk management
Power rating formulas are just the starting point for an oddsmaker. Market factors come into play as well. For example, since -7.15 is not a real point spread, they can decide whether to open -7 or -7.5.
Either of those numbers opens up vulnerability to 6-point teasers, where bettors can bring the Ravens down under a field goal. Plus, oddsmakers know their audience, and the possibility that they’ll want to bet the favored Ravens. As a result, I’d expect the oddsmaker to open Baltimore at -9, daring bettors to take the value in Arizona, knowing that “Ravens by 8” is the only result where that value comes into play.
NFL Team Betting Ratings
Now that we know how a point spread is made, we have a better chance of knowing what the market thinks of each team. Having that knowledge is critical in betting, since most bets are based on a disagreement with how the market feels about one, or both teams, in a given matchup.
Taking into account win totals that have been available to bet for months, here are the market ratings for each team going into the season, with my personal adjustment for strength of schedule factored in.
Each week - on Tuesday - we’ll work backwards, taking the widely available closing lines of the past week’s games and use that information to tweak these ratings. As we’ll remind each other every week, this is an estimation about how the market as a whole feels about each team going into that game, not how we do.
Player absence (almost always quarterbacks) will affect the closing line, causing an artificial drop in a team’s rating, which we’ll note in that article along with any other reasons why as my good friend, Sheldon Alexander, might say “the math ain’t mathin’.”
Beyond using this exercise to determine what value actually means in betting, it’s a better alternative than those tired “power rankings” articles the mainstream media offers up as clickbait.
Essentially, if we can speak the same language when discussing team strength, we’ll have a better chance to win bets on a league where winning and losing often comes down to a half-point.
Matt Russell was the lead betting analyst at theScore, and hosts THE WINDOW: Sports betting podcast. If there’s a bad beat to found, Matt will find it.
can I assume the same type of math for NCAA FB point spreads??