Is ‘Smart Money’ Really Smart in NBA Betting?

How to analyze point spread movements in the NBA

278
Is 'Smart Money' Really Smart in NBA Betting?

One of the most popular beliefs in gambling is that “smart money” rules. Whether it’s the “whales” in horse racing or the “sharps” in the NBA, many people believe that what, where, and when these pros bet can make all the difference to one’s bottom line.

In racing, this makes a lot of sense, as most markets are pari-mutuel, which means that a player with an edge can raise the effective takeout for other bettors in the pool if their wagers are substantial enough to move the market. (And, generally, they are — the term “whale” isn’t bestowed upon those searching underneath the couch cushions for loose change to plunk down on the favorite.)

But sports betting is a different animal. First, once you place your bet, the point spread or odds don’t change. Second, the market size is somewhat immaterial in that the sportsbooks control all the action, including the size of individual wagers. If the sport doesn’t generate a lot of interest from bettors, like the WNBA, large bets will simply not be accepted. As a result, most big line movements occur due to late-breaking news — a freak injury, coaching change, or adverse playing conditions.

What do point spread movements mean in the NBA?

Still, do these point spread movements matter, especially in a league like the NBA, where the grind of an 82-game season often plays a more significant role than other, more traditional handicapping factors? (For instance, a study conducted by researchers at The Wharton School of the University of Pennsylvania found that lack of rest was worth about 10% of the home-court advantage NBA teams enjoy.)

To answer this question, I turn to my database consisting of over 35,000 NBA regular- and post-season games played from 2007 to 2020. 

Here’s what I found:

No change in the point spread

Bets:6,762
Wins:3,381
Win Rate:50.0%
ROI:-4.55%

This is what we would expect — the results are neutral. Now let’s examine teams that saw a decrease in their point spread; in other words, their straight-up win probability went up.

Line decreased by ½-point or more

Bets:14,155
Wins:6,993
Win Rate:49.4%
ROI:-5.69%

Line decreased by 3 points or more

Bets:761
Wins:372
Win Rate:48.9%
ROI:-6.68%

Show me the money

While it’s not accurate to say that these line movements were necessarily the result of “smart money,” it is interesting to note that the greater point spread reduction (three points or more) produced more red ink than the smaller one (half-point or more). At the very least, this suggests that the popular notion of following the money may be overrated when betting on the NBA.

This hypothesis is bolstered by a 2011 study, which concluded that bettors are far more likely to bet on favorites than underdogs. In a season-long experiment, researchers adjusted the point spreads of NFL games to give an advantage to the underdog. Despite this, the study participants — all of whom claimed to follow the NFL season “extremely closely” — still picked the favorite 83% of the time.

One of the study’s authors, Leif Nelson, a Haas School of Business professor at the University of California, Berkeley, explains the phenomenon.

“When people decide how to bet on a game, first they identify who is going to win. The faster and easier it is, the less concerned they are with correcting that intuition when answering the more difficult question of whether the favorite is going to beat the point spread.”

From this standpoint, the declining point spread is confirmation that the team is “best.” Add to this the fact that 70% of the NBA teams getting increased action are favorites, and it’s easy to see why the numbers are so poor.

Looking at teams whose point spreads increased, i.e. their SU win probability went down, we observe the opposite effect.

Line increased by ½-point or more

Bets:8,566
Wins:4,324
Win Rate:50.5%
ROI:-3.63%

Line increased by 3 points or more

Bets:523
Wins:273
Win Rate:52.2%
ROI:-0.35%

If we throw the venue into the equation, the results get more divergent. We’ll start by taking a peek at the teams getting late betting action and observing whether they were home or away.

Line decreased by ½-point or more (visitor)

Bets:5,501
Wins:2,702
Win Rate:49.1%
ROI:-6.23%

Line decreased by ½-point or more (home)

Bets:1,500
Wins:729
Win Rate:48.6%
ROI:-7.22%

Line decreased by 3 points or more (visitor)

Bets:232
Wins:117
Win Rate:50.4%
ROI:-4.10%

Line decreased by 3 points or more (home)

Bets:84
Wins:43
Win Rate:51.2%
ROI:-2.50%

Location, location, location

At first glance — well, second glance too — these stats appear out of sync with what we witnessed earlier. However, I would point out that we’re dealing with a minuscule sample, so the results need to be taken with a grain of salt. Consider that the qualifying 84 teams in the last table come from 14 years of NBA play — that’s just six games a year.

With that in mind, let’s look at the teams whose point spreads increased and consider the location of those games.

Line increased by ½-point or more (visitor)

Bets:5,358
Wins:2,726
Win Rate:50.9%
ROI:-2.87%

Line increased by ½-point or more (home)

Bets:3,110
Wins:1,551
Win Rate:49.9%
ROI:-4.79%

Line increased by 3 points or more (visitor)

Bets:312
Wins:156
Win Rate:50.0%
ROI:-4.55%

Line increased by 3 points or more (home)

Bets:195
Wins:110
Win Rate:56.4%
ROI:+7.69%

Concluding thoughts

Finally, some profits, although I stress again, the latter samples are far too small to be reliable. Still, the results reinforce the overriding theme of this article — mainly, that point spread movements, at least in the NBA, tell us less about the sharps than they do about the betting crowd as a whole.

I know from my experiences with horse racing that many people take more comfort in losing on a 4-5 shot than they do in losing on a 45-1 shot. After all, it’s one thing to get beat; it’s quite another to do so alone.

References:

Entine, O., & Small, D. S. (2007, November 30). The Role of Rest in the NBA Home-Court Advantage. The Wharton School of the University of Pennsylvania. https://faculty.wharton.upenn.edu/wp-content/uploads/2012/04/Nba.pdf

Simmons, J. P., Nelson, L. D., Galak, J., & Fredrick, S. (2011). Intuitive Biases in Choice Versus Estimation: Implications for the Wisdom of Crowds. Journal of Consumer Research, 38 (1), 1-15. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.953.7478&rep=rep1&type=pdf

Leave a Reply