CFB betting how-to: Using team ratings to calculate point spreads
In the NFL, parity is encouraged. The salary cap and draft act as guard rails, resulting in the largest point spread in modern history being Jacksonville as 28-point underdogs at Denver in 2013. Meanwhile, that's not the case in college football, as the top teams in a conference are frequently more than a four-touchdown favorite, even against the bottom of their league and when they face teams in lesser-quality conferences. When spreads climb up to multiple touchdowns, how high is too high?
The first step in becoming a good bettor is to understand how the odds were made for the game being bet on, so we need to figure out what goes into the college football point spreads that get opened for betting each Sunday afternoon.
To come up with these point spreads, oddsmakers use their ratings for each team, apply those numbers to a formula, and then adjust for an estimated home-field advantage. Sprinkle in a little logic about bettors' general preferences, and they can make final tweaks to the game lines.
Those lines get bet on throughout the week and are eventually shaped into place by kickoff on Saturday, with bettors getting their last chance to weigh in. Oddsmakers should've found a number they can work with by then. Throughout thousands of games, sportsbooks hope to profit thanks to the vig (or commission) that bettors are paying to make a safe and regulated wager - most often at -110 - or by allowing exposure on the side they think will lose and winning half of those games.
Finding an edge on the market
College football betting lines shift more than in the NFL. Barring a mid-week injury to a pro quarterback, you aren't likely to see a shift of even three points in the pros. Moves of that nature - or larger - are a regular occurrence among the more than 50 FBS college games available to bet each week, meaning that there's more opportunity to make a bet on Ohio State at -13.5 on a line that closes at -17.5. Turning losses into wins - thanks to buying at a good price - is critical in achieving our goal of winning at least 52.3% (the breakeven point on -110) of our bets.
How do we beat the market to valuable numbers? We do so by having a good idea of what a point spread should be before it gets bet to that number and using past information from the betting market to determine what we expect the next point spreads to be. A simple rating system and a slightly more complicated formula is the easiest way to project that.
CFB team ratings
As we do in the NFL, we'll keep things simple by using a rating system out of 100. You're welcome to adjust your ratings in whatever way you evaluate teams. Do it by mascots, for all I care.
However, we have a summer's worth of the market weighing in on how it values each team, thanks to regular season win totals that create a starting point for how the market rates teams. Coming later this week, you'll be able to see national ratings for each team in each of the ten FBS conferences. But for now, we'll use the nation's independent teams to help us calculate projected point spreads.
| TEAM | RATING (/100) |
|---|---|
| Notre Dame | 80 |
| Army | 42 |
| UConn | 32 |
| UMass | 18 |
The other 129 FBS teams have their ratings adjusted by the strength of their conference. A 4.5-win SEC team is not the same as a 4.5-win MAC team. For the independents, a deeper dive into their schedule was required to come up with their respective ratings.
Point spread valuation/calculation
Figuring out a theoretical maximum point spread in FBS college football is much more difficult than in the NFL.
We've seen point spreads in the 70s when a lowly FCS team signs on to get run over by a big-time program. However, when discussing FBS, we'll cap the biggest point spread imaginable at nine touchdowns. A line of -63 would require a team rated at 100/100 to face a team rated at 0/100. If that hypothetical matchup happened, once the game is 63-0, what's the motivation for a team to run up the score higher?
From there, we'll take a team's rating out of 100 and multiply that percentage by 63 to give it a point rating. For example, Notre Dame's estimated rating of 80/100 gives them 50.4 points. The Irish don't face Army this season, but, if they did, Army's rating of 42/100 gives the Knights 26.5 points. Based on our estimated ratings, we'd expect Notre Dame to be -24 on a neutral field.
Elsewhere, UConn (20.2 points) close the season at UMass (11.3). That's a nine-point difference before you give the Minutemen any credit for home-field advantage.
More than in the NFL, the market shifts wildly on its opinion of college teams week-to-week as a result of college-aged players' tendencies to have inconsistent performances. Sticking to a stable median in your rating for a team is the best way to withstand the ebbs and flows of a team's play. This should help you find value by betting early in the week and by the time Saturday's games kick off.
Matt Russell is the lead betting analyst for theScore. If there's a bad beat to be had, Matt will find it. Find him on social media @mrussauthentic.