Box scores unlock regression to the mean

By Adam Burke  ( 


To try to predict the future, you have to consider the past. 

Studying box scores is a big part of the handicapping process for me. It can lead you in the right direction for a wager, and it can help you find some line value as well.

The notion of “what goes up, must come down” can certainly apply to college basketball. The best part is it doesn’t take an intimate knowledge of the teams or a lot of extensive work. All you’re looking for in previous box scores is a hint that there will be a regression to the mean.

Let’s look at a few stats you’ll want to consider with your box-score study.

3-point shooting

Weber State vs. Idaho State: On Monday, I isolated the Weber State-Idaho State game as a bet. The reason? Idaho State had just defeated a Division I opponent for the first time this season and needed a 14-of-29 performance from 3-point range to pull off its 81-74 victory against Idaho.

That game featured the Bengals’ second-highest number of made 3s and third-highest number of attempts this season. For a team shooting 27.7 percent from deep against Division I opponents, it seemed pretty safe that a similar performance wouldn’t happen the next time out. While the Bengals do shoot a lot of 3s, they also throw away a lot of possessions with a 23 percent turnover rate. A surefire recipe for a bad offense is taking a lot of 3s, missing a lot of 3s and turning over the ball too much.

The line went from Weber State -9.5 up to as high as -12. Obviously, the bottom line matters in terms of winning or losing, but all you can do is try to get the best number possible and hope it works out. This one did, as Weber State won by 17 while Idaho State shot 22.7 percent from 3.

Monmouth vs. Marist: On the flip side, Monmouth-Marist wound up being the opposite kind of game on Sunday. Monmouth lost 84-48, which is undoubtedly going to catch some attention. The Hawks shot 2-of-20 from 3, while the Red Foxes went 10-of-23. The Hawks had an awful shooting night, which happens, and while they’re not a great offensive team, a 36-point loss is far from the norm for King Rice’s group. Monmouth is an example of a team I’d look to bet in the next game.

Utah vs. Arizona: One last example is Utah-Arizona from Saturday. Arizona won by 18 points but went 3-of-18 from deep. The game fell close to the expectation, so it won’t get a lot of attention, but the Wildcats had an awful shooting performance from 3 and still won handily. That says a lot about a team.

I mentioned the Wildcats’ style last week because they play at a fast pace with tremendous efficiency and should be able to cover big spreads. While they didn’t cover the 19.5 points against Utah, the fact they came close despite their 3-point shooting suggests my read on them is right.

Free-throw discrepancy

One of my favorite angles is to fade a team that had a favorable free-throw discrepancy in its previous game but won by a small margin or lost outright. 

East Carolina vs. Memphis: East Carolina defeated Memphis 72-71 on Saturday, outscoring the Tigers by 12 points at the free-throw line with 10 more attempts. ECU actually needed a bucket at the buzzer to win, even though Memphis was in significant foul trouble and only went 4-of-11 at the free-throw line.

Memphis is a bad free-throw shooting team, so I wouldn’t rush to back them. But ECU is a team I’d look to bet against in the next game.

Rutgers vs. Maryland: Maryland is the opposite example. The Terrapins lost to Rutgers 70-59 on Saturday despite outscoring the Scarlet Knights by 11 points from the line with 10 more attempts. Maryland made 18 shots from the floor and 17 from the line, indicative of how bad they are offensively. I wouldn’t expect that to change.

An important note for this section: Game state matters because teams will foul late in hopes of adding possessions, so you have to apply some context when it comes to free-throw discrepancy.

General regression to the mean

At some point, every team will play its best game and its worst game. If something happens that is way out of the ordinary, take a deeper look and find out why. It doesn’t have to be a detailed breakdown. It can be something right on the surface that you can file away in your mind for upcoming games.

Think about these stats from Saturday:

— LSU had .843 points per possession in its loss to Arkansas, the Tigers’ second-lowest output of the season.

— Texas Tech had its highest TO percentage against Kansas State at 27.3% in its 11-point loss.

— Stephen F. Austin scored 41 points despite forcing 18 turnovers. The Lumberjacks were 15-of-57 from the floor and had .652 points per possession.

If something doesn’t fall in line with the season’s norm, it’s unlikely to continue (outside of a COVID-19 outbreak or something injury-related). Variance is a very real thing, but the most extreme outliers tend to go back to the norm quickly.

back to news

Subscriber Only News  Vsin Exclusive

View All



Follow The Money: We have reached the point in the season where you must factor byes into your handicapping. Even the week before the bye is a factor (more from Steve Makinen on this topic).

VSiN Big Bets: With such small totals, be careful laying a run line in these MLB playoff games.

View tips for Pro subscribers


Matt Brown: Dolphins -3 vs Jets

Pauly Howard: Eagles 1H (-3) vs Cardinals

View new sortable Picks at a Glance page for Pro subscribers



Betting SplitsPercentage of handle & tickets from DraftKings, updated every 10 minutes. GO HERE.

NFL Power Ratings3 sets of numbers from Steve MakinenGO HERE.

CFB Power Ratings3 sets of numbers from Steve MakinenGO HERE.

View more CFB tools for Pro subscribers

View more NFL tools for Pro subscribers