This article is interesting for several reasons:
First, point spreads are not perfectly competitive markets. The spread is roughly 50% higher for favorites. This means there is an implicit decrease in profitability for betting on favorites. This markup has nothing to do with incoming bets. The odds makers inflate the spread for additional profit on their end.
Second, people generally bet on favorites because they automatically choose who is likely to win and then ignore the spread when betting. This is irrational because the spread is supposed to (theoretically) make the odds 50/50 for both teams.
Third, the data clearly supports that favorites and underdogs win approximately in equal proportion after the spread is taken into account.
Fourth, these biases are irrelevant if point spreads don’t exist.
MY OPINION: This article is useful for betting on boxing. Instead of point spreads, we have adjusted odds. An example: Mayweather/Maidana is -1400/+850. That’s a 65% markup on the betting line for Mayweather (or, to put it in another way, you’re making 65% less if Mayweather wins all else equal). A competitive market would have equal odds: -1125/+1125.
If you plan on making multiple bets on a wide array of fights you may want to bet more often on the underdogs. You will maximize your profits. However, because there is no point spread, you can still come out positive if you consistently bet on favorites (you just won’t make as much).
Note: FiveThirtyEight is an independent site affiliated with ESPN (similar to Grantland). They specialize in quantitative analysis of many different topics. Sports is one of their favorites (along with politics and economics).