Are NBA Basketball Games Fixed?

L.A. Lakers
Wendy/Flickr/CC BY 2.0

Former NBA official Tim Donaghy created a bit of a stir with his allegations that Game 6 of the 2002 Western Conference Finals between the Sacramento Kings and the Los Angeles Lakers was fixed to ensure a seventh game. Even though his claim is impossible to prove there are plenty of people who will give Donaghy the benefit of the doubt, as, fix or not, the game is widely recognized as the worst officiated game in recent memory.

The game was called so poorly it prompted a letter from Ralph Nader to NBA Commissioner David Stern.

In wake of Donaghy's allegations, a number of people have wondered about the integrity of basketball games and are asking if the games are really fixed. If you ask Jonathan Gibbs, a graduate student at the Stanford University, the answer is that it's highly likely.

Gibbs' Research

Gibbs drew on the research of Justin Wolfers, a professor of business and public policy at Wharton School at the University of Pennsylvania, who did a similar paper on the possibility of point-shaving taking place at the college level. Gibbs looked at NBA point spread tendencies to determine if there were any unnatural occurrences taking place. Just as Wolfers found that large favorites in college basketball didn't cover the point spread a disproportionate number of times, Gibbs found similar findings in the NBA in his paper titled "Point Shaving in the NBA: An Economic Analysis of the National Basketball Association's Point Spread Betting Market.

"The premise of this paper was to empirically examine the NBA betting market to test if, in spite of the large monetary compensation NBA players earn, whether there is evidence of cheating through point shaving. Generally, the informed bettors participating in the NBA betting market are efficient in allocating point spreads to reflect differing abilities between teams," Gibbs stated in his thesis.

"However, the asymmetric incentives, caused by the act of setting the betting line, are being exploited.

"Both data sets provide strong statistical evidence to suggest that this phenomenon, traditionally associated solely with unpaid college basketball players, exists in the professional ranks. The importance of this finding is that through publicizing point shaving's existence, the probability of its detection increases, correspondingly increasing its costs, and decreasing its incidence."

Gibbs studies NBA point spread date from 1993-94 to 2006-07 and found that underdogs covered the spread 50.05-percent of the time, winning 7822 times, losing 7805 times and there were 232 pushes. But larger underdogs, where it's generally considered easier to fix games, covered the point spread at a higher rate. Underdogs of 10 or more points covered the spread 52.15-percent of the time (1,372-1,259-39), while 13-point and greater underdogs covered the spread 53.04-percent of the time (462-409-20).

Arguments Against Gibbs

The most common complaint against Gibbs' findings is that it fails to take into consideration "garbage time" in the final few minutes of a lopsided game. These complaints say that the team who is ahead is more likely to take its starters out, while the team trailing may leave its top players in longer in hopes of a miracle comeback and can turn an 18-point deficit into an 11-point deficit in several minutes.

Naturally, the other claim is that NBA players simply make too much money to take a chance of getting caught in a point-shaving scandal. A player who is in the position of being able to influence the point spread is likely a player who makes more than $1 million a year and has even less incentive to risk it all for trying not to cover the point spread.

Arguments For Gibbs

Gibbs has plenty of supporters in the academic world on the basis of his research, including Wolfers; California State University, Bakersfield Associate Professor of Economics David Berri; Stanford Assistant Professor Nick Bloom; and University of North Carolina at Greensboro Assistant Professor of Economics Dan Rosenbaum. All weighed in on Gibbs' research for ESPN and had positive things to say about it.

Others will point to the fact that the previously stated "garbage time" typically involves the bench players from both teams, not just the team that is ahead in the final few minutes.

And those bench players often have additional incentive to do well, trying to impress coaches and many times fighting for their jobs. Additionally, a well-rested player off the bench should be able to come close to matching the performance of a player who has already played 35 minutes and is fatigued.

Could Officials Be to Blame?

Finally, there's the possibility that it isn't the players who are doing the fixing but instead is the officials. After all, a few bad calls in a blowout are going to draw less attention than a few bad calls in a game that comes down to the wire, where each possession is magnified.

Even Gibbs recognized this in his thesis.

"Additionally, a major point, untouched in this examination, is that point shaving is consistently attributed to players, while coaches and referees are no less principal characters in basketball games and could be equally culpable," his paper stated. "Further research could study substitution patterns and infractions called to test if, instead, coaches or referees are willfully affecting the final margins of games."

Fixing Games in Other Sports

Those who believe that the fix is in do have some justification for their beliefs, as match-fixing has occurred in other sports. Most people are familiar with the saga of the 1919 Chicago White Sox, which occurred in the World Series and rumors of shady tactics in the NFL are always taking place. There is an excellent article on fixing in the world of cricket on the Top Betting Reviews website.


Gibbs doesn't come right out and state that NBA games are definitely fixed but does a good job of persuading readers that the possibility certainly exists and backs it up with solid research.

While large favorites do tend to win at a lower rate against the point spread than they should, it will be difficult to prove that it is because of point-shaving, but anybody can look at the statistics and realize that something doesn't quite add up.