The first variable I wanted to test for was game location. More specifically, I wanted to examine the effect that playing a game at home vs. playing a game away had on a given team’s chances at beating the spread.
In basically every team sport you can bet on, the concept of home court(/field) advantage is taken into account when setting spreads. Although it can certainly vary from team to team, in the NBA this advantage is thought to be good for roughly 2-3 points. So for example, a team that would be a 7 point favorite at a neutral location becomes a 9.5 point favorite when playing at home. One interesting side note: Many people assume that home field/home court advantage occurs because players are able to perform better in environments where they are more comfortable and the fans are cheering for them. While this probably does have some effect (even more so in college), the real reason teams perform better at home is that they get more favorable calls from the referees. Even more interesting is that this bias is not conscious on the part of the officials; rather it is likely an unconscious conditioned reaction to having every call for the home team cheered and every call against the home team booed.
So although the effect of home field advantage is already baked into every NBA spread that you see, it is certainly possible that the average bettor either overestimates this effect or underestimates this effect. With this in mind, I ran a paired samples t-test to see whether there was a significant difference between teams records against the spread when home and teams records against the spread when away.
Results: I found that the mean WATS (Wins Against The Spread) for home teams from 2009-2013 was 101.27, and the mean WATS for away teams was 106.03. This difference was significant at the .019 level, which is a significant result (percentage?) and means that the null hypothesis (there is no difference between home team WATS and away team WATS) can be rejected.
Analysis: The results indicate that generally speaking, the average bettor tends to overestimate the effect of home court advantage, leading to more favorable spreads for away teams which allow them to beat the spread more than home teams. While I was somewhat surprised to find a significant result in my first analysis, this was an encouraging sign that there are holes to be exploited in the way Vegas sets their spreads.
What could be the reason for this overestimation of home court advantage? One possible explanation is that the effect of home court advantage has been lessening in recent years, and bettors have failed to adapt. According to the statistics, the past two years have seen a devastating 36% decline in homecourt advantage, the biggest 2 year dip in NBA history.
However, the next logical question to consider is this: Could someone use this information to consistently make a gambling profit? Away teams were successful against the spread 3,181 times out of 6,229, good for a roughly 51% winning percentage. So if you had blindly placed a bet on every single away team for the past 4 years, you would not reach the 52.4% win rate needed to overcome the house take. This is not exactly surprising, as I would think that such a simple profit-yielding strategy would have been discovered by now. However, this is not to say that this information couldn’t be used successfully as part of a more complex betting strategy.