Back to some behavioral economics. Thaler cites three stylized facts about horsetrack betting, they are as follows:
1. Most betters bring an amount of money that represents a small amount of their overall wealth. In terms of mental accounting, this money may represent the theory that money is non-fungible. Because they are not brining their whole wealth may be indicative of the portion of money associated with a specific label. For example, perhaps the money is thought of to carry utility in the form of entertainment. So instead of using this money to pay off debt, this money is strictly dedicated to entertainment, in this case betting.
2. The sum of money that is brought to the track is allocated the course of the betting day with the intent to be used during every opportune time to bet during that day. Continue to do so, unless the sum of money has completely depleted before the day has concluded. In terms of mental accounting, this may represent the affects of an individual applying a value function. Value function is measured in terms of gains and losses. So it may be the case that this individual has spent the time and effort to go to the track, which becomes a sunk cost, and feels that he/she must recoup the cost by partaking in the desired activities to the fullest extent. Therefore, the individual may feel obligated to spend all the money in order to feel the day was fulfilled and that the gains through utility with the gambling have offset the costs of getting there.
3. People who go with other people to form groups rarely bet among themselves even though they could essentially guarantee a zero sum game for the group by betting against each other (one looses another one gains). In terms of mental accounting, this may reflect a transaction utility. Transaction utility is the thought that the individual gets a certain amount of utility from the good but also an extra value of utility, positive or negative, through the transaction. The actions in this scenario indicate that the individual receives some unobservable amount of utility when betting as a group (transaction). This case, it may reflect group conformity and fitting in socially with peers thus compensating for the rational loss.
The common sense explanation for the third fact is that people do not get as much utility from taking money from a friend as they do taking money from the track. Another reason friends may not bet against each other is that they might feel obligated to repay the loser (for example buy them a drink because they lost money) in some way or have an inclination to not take all of the money that was bet. Another problem especially when we talk about horse racing is that the bundle of bets an individual wants to make may not be available within the group or that there is a cost to organizing the bet within the preferences of the individuals in the group. Think about how complicated it would be to organize a trifecta bet between a group of four individuals.