Favourite-Longshot Bias

In gambling and economics, the favorite-longshot bias is the observed tendency for bettors, on average, to overvalue longshots and undervalue favorites. In a horse race where one horse is given odds of two to one and another one hundred to one, for example, the true odds might actually be closer to one and a half to one and three hundred to one respectively, making a bet on the longshot a far worse proposition than a bet on the favorite; it is not uncommon for bettors to lose only about five percent betting favorites while losing around forty percent betting longshots over the long run. The phenomenon was first identified by a researcher named Griffith, and later researchers have proposed various explanations for why people keep making these losing bets, including risk-loving behavior, risk-averse behavior, or simple inaccuracy in estimating the true odds; several methods exist to measure the bias by converting betting odds into estimated true probabilities.