What this shows: the grade and the numbers behind it, over the five test seasons. A typical miss is how many points the line was off from the final margin in an ordinary game (statisticians call it RMSE). Difference is ours minus the market’s, and luck is how much it could wobble by chance. Beyond the market asks whether our line knows anything the market’s doesn’t: 0 means no.
What this shows: the settings Method A worked out for itself from past games: what home court, a back-to-back, rest and travel are worth in points, how far the rest of Division I plays under the big six conferences, and how much of last season a rating keeps. Spread of misses says how far a real score usually lands from our line, which is how a line becomes a win chance.
What this shows: November is where we trail the market most. Our ratings start from last season, and college rosters change more than any league’s: freshmen arrive, transfers switch teams, and a bench player whose seniors graduated suddenly plays 30 minutes. The market knows about that in November; our ratings learn it from the first few weeks of scores. Games between a big-six team and the rest are hard too: most come in November and December, when ratings know least.
What this shows: each dot is one of the settings we tried for how fast ratings move, how much of last season they keep, and how much a closing line counts. We picked the gold dot because it did best on the tune seasons. If the tuning had fit old seasons too closely, the gold dot would sit high on the test seasons. College basketball changes a lot from year to year, so the grid is coarse on purpose and only three settings are tuned.
What this shows: against the closing line our LEANs lose, as on the other Lines pages: the close already knows what our ratings know. Against the opener, the first line posted, it is different: when our number was a few points from it, the close usually moved toward us. That movement is closing line value, the leading sign of an edge. Whether it pays is what the Picks tab is measuring.
What this shows: how far off we were each season, next to the market. Difference is ours minus the market’s; green means within the half-point goal. Tune seasons set the model up; Test seasons are the exam it never saw. 2019-20 and 2020-21 are missing: the first was cut short in March and the second was the COVID season, both rated but not scored. A season is named by the year it ends.
What this shows: each dot is a group of games where we gave the home team about the same chance to win; bigger dots hold more games. Dots on the dashed line mean our percentages can be trusted:
What this shows: the ten test games where our line and the market’s were furthest apart, with the final score. Most are early in the season, or a team missing its best player: news the market has and our ratings don’t.
What this shows: the grade and the numbers behind it, over the five test seasons. A typical miss is how far our total was from the real total (both teams’ points) in an ordinary game. Difference is ours minus the market’s closing total, and beyond the market asks whether our total knows anything the market’s doesn’t: 0 means no.
What this shows: how Method B learns: how much of a team’s scoring tendency carries over the summer, how much more a closing total counts than a final total (it is a sharper reading), and the straightening line, fitted on the tune seasons, that turns our raw total into the one we price.
What this shows: totals are hardest in November. How fast a team plays and how well it shoots depend on who is on the floor, and in November that is the part of a roster our tendencies know least.
What this shows: higher totals scatter more: a game expected to hold 165 points can land further from it than one expected to hold 125. So our over and under chances use a spread of results that grows with the total (), where the NBA and NFL pages use one number.
What this shows: the same check as Method A’s, for totals. The gold dot is the setting we chose on the tune seasons.
What this shows: how far the closing line missed in an ordinary game, for spreads and totals, over the test seasons. The spread of misses is the width of the bell curve around the close that we use to turn a line into a chance.
What this shows: the gold bars are how often test-season games ended by each margin. Unlike football, basketball has no big key numbers: a bell curve around each game’s closing line (the dashed line) follows the bars closely. The one gap: games end by a single point less often than the curve says, and by 3 to 5 a little more often. In the last minute the trailing team fouls, the leader makes free throws, and close games stretch out.
What this shows: games that closed at a whole-number spread, and how often the final margin landed exactly on it, a push that returns the bet. In basketball that happens about one time in thirty. The bell curve says a little more on lines of 1 and 2 (one-point games are rarer than it thinks) and about right from 3 on.
What this shows: how far the closing spread missed in each kind of game, next to our own miss on the same games. The big six are the ACC, Big East, Big Ten, Big 12 and SEC, and the Pac-12 through 2023-24; the rest of Division I is the other conferences.
What this shows: the exit test and the counts behind it. Test games are the Division I games the grades are scored on (2021-22 to 2025-26); 95% of them need a closing spread. The tune seasons (2013-14 to 2018-19) set the models up; their lines are thinner, and they are scored on the games that have one.
What this shows: each season’s Division I games and how many have a closing spread, an opener and moneylines. Green means close to every game. Lines start in 2013-14 and openers in 2023-24; before 2018-19 the sources skip many games between two small-conference teams. Books per game is how many sources our close is the middle of.
What this shows: how many teams played in each tier each season. The big six are the ACC, Big East, Big Ten, Big 12 and SEC, with the Pac-12 through 2023-24; the rest of Division I is the other 26 or so conferences. Games against teams below Division I are left out.
What this shows: when a team moves between tiers (Houston, Cincinnati, UCF and BYU into the Big 12, or Washington State and Oregon State left behind when the Pac-12 broke up), its rating moves with it, so its strength carries over the move unchanged.
What this shows: where every number comes from. All of it is free; only CollegeBasketballData needs a key, and its free tier of 1,000 calls a month covers the daily runs.
For information and entertainment only. Nothing on this page is betting, financial or legal advice. Our lines, chances, EVs and stakes are estimates from statistical models: they can be wrong, and past results do not predict future ones. Paper picks and Kelly stakes illustrate the math; they are not recommendations. Any bet is your own decision and your own risk: only where sports betting is legal, only if you are 21 or older (or the legal age where you are), and only with money you can afford to lose.
Not affiliated with or endorsed by the NBA, the NFL, the NCAA, any team or conference, ESPN, Action Network or any sportsbook. Team names belong to their owners. Book lines come from third parties and can be late or wrong: check your book before you bet.
Gambling problem? Call 1-800-GAMBLER.