Top NBA Betting Strategies Based On Academic Research

What Makes NBA Betting Different? Ongoing Inefficiencies!

No serious conversation about NBA betting strategy begins anywhere other than Haralabos Voulgaris. From the late 1990s through the mid-2000s, he ran a winning rate approaching 70% at his peak, regularly staking over $1 million per day on NBA games. 

He found a structural flaw in how books priced halftime totals. Sportsbooks were setting first-half and second-half totals as an exact 50/50 split of the game total. However, NBA games don't score evenly across halves. More points are scored in the second half as teams accumulate fouls, use timeouts deliberately, and send opponents to the free throw line. 

Voulgaris understood this before the books did, and he hammered second-half overs, particularly under coaches like Eddie Jordan, Byron Scott, and Jerry Sloan, whose game management tendencies he had closely mapped.

When the books caught up around 2004, Voulgaris built a predictive model called "Ewing" with a mathematics prodigy that simulated games at a lineup level and identified mispriced odds before the market corrected. He became a pioneer in NBA statistical analysis, eventually presenting at the MIT Sloan Analytics Conference and being hired by the Dallas Mavericks as Director of Quantitative Research.

NBA Spread Betting Strategy

A landmark 2005 study by Paul and Weinbach in the Journal of Sports Economics, analyzing seven seasons from 1995 to 2002, found systematic bettor misperceptions in the NBA point spread market. Favorites were consistently overbet by uninformed bettors, and a strategy of backing big home underdogs was profitable. ^[1] The Sport Journal replicated the finding across more than 10,000 games from 2000 to 2008, with consistent results. ^[2]

Credit: BetMGM Sportsbook – Screenshot captured by Chad Nagel on May 25

The mechanism behind this was theorized by economist Steven Levitt in a 2004 Economic Journal paper that sportsbooks don't always try to balance their book. Instead, they shade point spreads to exploit the public's favorite bias and then take positions on the opposite side themselves. ^[3]
Evidence from over 6,000 NBA games found bets on favored teams had a win probability of less than 50%. ^[4] The books were profiting from the public's predictable irrationality.

NBA Totals Betting Strategy

One of the most well-documented inefficiencies in basketball betting is that early-season totals are persistently mispriced. A 2007 study published in Finance Research Letters found that NBA totals lines are significantly biased at the start of each season. ^[5]

In Week 1, 58.2% of games finished under the total, a win rate far above the 52.38% needed to cover vig. A simple betting-the-under strategy during opening week yielded average returns of 11.1% per game. ^[6]

Credit: BetMGM Sportsbook – Screenshot captured by Chad Nagel on May 25

This is because oddsmakers rely heavily on the previous season's data to set early lines, and teams change. Roster turnover, new coaches, and preseason sample sizes give books too little reliable signal, so they anchor to stale baselines. 

NBA Moneyline Betting Strategy

Research published in The Sport Journal found that betting against a team's current momentum, either positive or negative, won at a 56.5% rate. ^[7]

The public overestimates momentum in both directions. A team on a losing streak gets bet against harder than the underlying data warrants, while a team on a winning streak gets backed more aggressively than it should. You can make money by predicting a regression to the mean and backing a struggling team to bounce back or a hot team to lose steam.