College Football Advanced Metrics Glossary: Efficiency, Scoring
Bettors in states like Ohio and Nebraska are gearing up to legally bet their first college football season. Others in Kentucky and North Carolina are just around the corner. When it comes to successfully handicapping college football odds, bettors have to be aware of advanced metrics and how they apply to the game. Despite friction against “analytics,” it’s impossible to win lon-term without embracing that side of the game. So for those new bettors — or even veteran ones — it’s worth going over advanced scoring metrics in college football.
You may have seen stats like Expected Points Added (EPA), Points Per Drive (PPD), and Success Rate (SR) thrown around on Twitter and right here at TheLines. But what do they mean? Is any one more important? Which stat should you consider when handicapping college football offenses and defenses?
Best Sportsbook Promo Codes In Your State
Before we get rolling with breaking down the advanced scoring metrics in college football, consider taking advantage of the best sportsbook promo codes below. Even if you aren’t new to the sports betting landscape, books you haven’t yet signed up for are offering great bonuses:
College Football Advanced Scoring Metrics
The football community has a whole is moving away from using total points scored and points per game. While these stats themselves are perfectly suitable for assessing how well an offense did over the season at a glance, they’re often misleading. For instance, Liberty finished No. 47 in scoring defense last season (25 points per game). But the Flames’ defense was much better than that all season long, with their final points allowed number being skewed in the final weeks. Hugh Freeze was on his way out and the team clearly resembled that, giving up 108 points in their last three games to UConn (No. 107 in points per drive), Virginia Tech (No. 119), and New Mexico State (No. 90).
One massive game or a small string of outliers can easily skew points per game. It doesn’t take into account garbage time points scored, either. So let’s look a little deeper.
Points Per Drive
This one is fairly straightforward and factors into account pace. Consider points per drive a mark of how efficient offenses or defenses are in scoring or denying points each time they or their opponent has the ball.
TCU finished the 2022 season ranked ninth in scoring (37.4 points per game) but No. 18 in points per drive. The Horned Frogs often found themselves in back-and-forth shootouts, causing an inflated total scoring number. They finished slightly above Oregon. Does that make TCU a better offense than Oregon? The Ducks finished fifth in points per drive and EPA per play, meaning they scored more frequently with the ball than TCU did.
Scoring matters in football. Rather than framing it cumulatively, breaking it down into efficiency with points per drive gives users a better snapshot of that offense. However, it still doesn’t tell the whole story.
Team Expected Points Added (EPA)
EPA is a metric that, when thrown into a conversation, often makes people’s eyes glaze over. But not only is it an invaluable tool to use, it’s also easier to digest than some may initially think. The difficult answer is that EPA is a stat that gauges a player or team’s usefulness on a per-play basis. How many points per play was a player or team worth? This stat adds context to plays. The easier answer is it factors into account down and distance, field position, and type of play.
So, revisiting the Oregon vs. TCU offense discussion, Oregon ranked fifth in EPA per play versus TCU’s No. 17 rating. This context builds onto our hypothesis that the Ducks’ offense was more effective with the football than the Frogs were — if only slightly so.
EPA is a great metric that’s growing in popularity and accessibility, yet it still has its blind spots. The biggest one is that the stat isn’t opponent adjusted. So while Georgia might lead the country in EPA per play this coming season, they may not field the country’s No. 1 offense, rather took advantage of a schedule that ranks No. 44 in difficulty.
Dr. Parker Fleming painstakingly collects and keeps EPA numbers for college football at CFB-Graphs.com.
College Football Advanced Efficiency Metrics
Individual Expected Points Added (EPA)
College Football Data offers EPA numbers on individual players, dubbed “Predicted Points Added.” There, you can exclude garbage time and narrow down EPA by down or game thanks to ample filters.
EPA works mostly the same with players as it does for teams. This metric is especially useful with quarterbacks, who see vast EPA shifts between throwing touchdowns and throwing interceptions. Skill players also have those shifts. Fumbles are a bit more left to chance (particularly with who recovers them) than interceptions. Jordan Travis last season finished with 1,100 fewer passing yards than Drake Maye and 14 fewer passing touchdowns than Sam Hartman. But Travis led the ACC in EPA and passing efficiency, again closing that perceived production gap between the three QBs.
While nitpicking a spot or two in EPA to justify one QB being better than the other is foolhardy, extremes in the metric can, and should, be factored in. For example, Blake Shapen was heralded as a playmaker for Baylor. But last season, he finished outside the top half in EPA among starting QBs (min. 150 snaps) — helped along by his tendency to turn the football over on late and critical downs.
Success rate is a metric that you’re going to have to dig a bit to find for college football (luckily, I’ll be doing the digging for you this season at TheLines). While EPA is a solid measure for efficiency and explosiveness, success rate is a strong indicator of consistency moving the football.
A “successful play,” when left undefined, is entirely subjective. Warren Sharp of Sharp Football Analysis defines a successful play for the NFL as gaining:
- At least 40% of the yards to go on first down
- At least 60% of the yards to go on second down
- 100% of the yards to go on third and fourth downs
Success rate takes into account down and distance and is a great metric to assess rushing and passing attacks. While a 15-yard run on 3rd & 20 goes into the books as a 15-yard run — something that may skew yards per attempt or yards per play — it was ultimately an unsuccessful play. On a per-play basis, that 15 yard rush was as good as a 15-yard touchdown run and actually worth more on a per-play basis than a five-yard rush on 4th-and-2 that moved the chains.
For one, Houston (36.1 points per game) vastly outscored Air Force (26.2) in 2022. Using just scoring, one might assume that Houston has a largely better offense than Air Force. However, the Falcons have a higher offensive success rate (48.1%) than the Cougars (46.9%) and that the low scoring numbers are a result of style of play. While this isn’t a cut-and-dry “Air Force’s offense is better” argument, it does suggest that the gap wasn’t all that wide.
College Football Scoring Metrics: Which Should I Use?
The answer is: all of them. Each metric above has its advantages and disadvantages and, when combined in a balanced marriage, can give a strong glimpse at a team’s offense or defense.
The old adage is, “numbers don’t lie.” But that’s a farce — numbers absolutely do lie and regularly so. That’s why it’s important to add context around those numbers. To take all of these further, opponent-adjusted numbers are really the best way to go. If opponent-adjusted EPA isn’t readily available to you (which most of the time it won’t be for college football), then you’ll have to adjust yourself by considering the slate of opponents and looking at their PPD, EPA, and SR numbers.
Check back in throughout the 2023 college football season to see these metrics in action.