Correlated Parlays

Sharp Sports Betting is a tool for those interested in winning money at sports betting. The book explains the most common sports bets, what all the numbers mean, and the mathematics behind the numbers.

A time to consider betting a parlay is when you can take advantage of correlation. Correlation means the outcomes of two bets are linked somehow; if one bet wins the other is likely to win, and if one bet loses the other is likely to lose. Two related terms that need to be defined are correlation and covariance. Correlation is the property of two outcomes being linked. Covariance is the mathematical measure that describes the degree of correlation. Lack of correlation is covariance of zero. A perfect correlation has covariance of 1.0. A perfect negative correlation also has covariance of 1.0. A good example of bets that are correlated is halftime results and end-of-the-same-game results. The team that is ahead at halftime is more likely to win the game than is the team behind at halftime. Sportsbooks are aware of that correlation, and will not let you parlay halftimes and totals of the same game. What you need are correlations that you are allowed to bet.

There is zero correlation when two games are independent. When Tampa Bay plays at Cleveland and Tennessee plays at Jacksonville, you probably have no reason to parlay Cleveland and Jacksonville because the results of those two games are uncorrelated. The probability of Cleveland covering the spread is the same whether Jacksonville wins or loses.

One situation that can cause correlation is when there is a single random variable that is overly important in deciding the outcome of two bets.

A caution: Don’t use the word “correlation” in a sportsbook. If the people taking your bet think that correlation is possible, they might refuse the bet. Sportsbooks generally do not knowingly accept correlated parlays.

Example: Storm

For example, suppose a big storm is predicted to move into an area in which two outdoor games are being played, but nobody knows whether the storm will arrive early enough to affect the games. Also suppose that if the storm hits, it most likely will affect both games; it would be unlikely for the storm to hit one game and not the other. Further suppose that you have a good idea of how a big storm is likely to affect each game. That suggests making correlated parlays. How you bet each game depends on how you expect a storm to affect it. Perhaps you think that a storm will hold down the score in one game, and will favor one team in the other game. You might want to parlay the under in the first game and the appropriate side in the second game.

Conclusion

Forget about parlays unless you can find bets that are correlated or a parlay card with selections you think are strong.

This is part of an occasional series of articles.

Excerpted with permission from Sharp Sports Betting by Stanford Wong, edited for this format.