How to bet on college basketball
College basketball is the one American sport where the book is at a real disadvantage on part of its own board. It has to price more than three hundred teams, and it can't know them all. The way to bet it is to find the part of the board the book knows least and learn it better than the book does.
How the number gets made
You bet a college game the way you bet the NBA: a spread at -110 on each side, a moneyline and a total, usually with a first-half line as well. The difference is where the numbers come from.
On a game between two famous programs, the line is made from power ratings, adjusted for home court, and then shaped by a lot of money from people who watch those teams. That number is hard to beat. On a Tuesday game between two small-conference schools, the book is often working from a rating and not much else, and it knows it. It opens the line with a low limit, waits for the first bets to tell it where the number should be, and moves. Most of the soft lines in college basketball are in those games, and most of them are only soft for the first few hours.
So I'd tell anyone starting out to pick one conference, ideally a small one, and follow every team in it: watch the games, know who is hurt, know who plays fast. That is a few dozen teams instead of several hundred, and it is the only way an ordinary bettor knows more than the book about anything.
Tempo decides the total
A college total is mostly a sum of possessions. Say two teams usually play about 68 possessions each and score 1.05 points per possession: that is 71.4 points a side, 142.8 for the game. Now say one of them is a slow team that holds the ball and drags the game down to 62 possessions. At the same scoring rate, that is 130.2.
That is 12.6 points of total, from the pace alone. The book knows the averages. What it doesn't always weigh properly is the matchup, and which team gets to set the pace. A small school with an unusual style (very slow, or living on threes) is a different problem for a big-conference team that hasn't seen it, and the line tends to price the bigger name rather than the problem.
The big name against the small school
The public bets brand names, and in college basketball the brand is the conference. A team from a major conference playing a mid-major draws the casual money almost by reflex, and the book knows that and shades the number toward the big school.
I wouldn't take the small school on principle. I'd take it where there's a reason: a mid-major that has already played major-conference teams close, a major team with a record built on weak opponents, or a big school with a road trip and a conference game coming that matters more to it. Records in college basketball tell you a lot less than they seem to, because schedules are so uneven.
One player matters more here, too. College benches are thinner than pro ones, and losing the best player can move a game more than the line allows for if the news is late.
February
By the last month of the regular season, the board is full of games where the two teams don't need the result equally. A team fighting for a tournament place plays a team whose season is over. The book prices the teams; it is not so good at pricing how much a result matters to each of them. That is the one month I'd say the motivation angle is worth more than usual, and it still comes second to the number.
March is its own subject, on the March Madness page.
Futures, early
Title odds for next season go up as soon as this one ends. They carry a big cut, they tie your money up for a year, and most of the time I'd leave them alone.
The exception is when you know something the public will find out later: a coaching change at a smaller program, a transfer or freshman class that people haven't heard of yet. Say you take $100 on a team at +2000, a $2,000 win if it cashes. If the public finds the team in the winter and the price drops to +800, a new bettor would win $800 on the same $100. You hold the better ticket, and you can hedge it later if you want. If the public never finds the team, the $100 is gone, and that will be the usual result.
The bankroll
There is a college game every night from November to March, and every bet pays the cut. The college basketball betting sites page works through what that costs over a season. I'd bet only the games in my own conference, at 1 or 2 percent of the bankroll each, and let the rest of the board go past.