Predicting World Cup results: Do goals seem more likely when they pay off?

In a series of experiments, Bar-Hillel and Budescu (1995) failed to find a desirability bias in probability estimation. The World Cup soccer tournament (of 2002 and 2006) provided an opportunity to revisit the phenomenon, in a context where wishful thinking and desirability bias are notoriously rampant (e.g., Babad, 1991). Participants estimated the probabilities of various teams to win their upcoming games. They were promised money if one particular team, randomly designated by the experimenter, would win its upcoming game. Participants judged their target team more likely to win than other participants, whose promised monetary reward was contingent on the victory of its rival team. Prima facie this seems to be a desirability bias. However, in a follow-up study we made one team salient, without promising monetary rewards, by simply stating that it is "of special interest". Again participants judged their target team more likely to win than other participants, whose "team of special interest" was the rival team. Moreover, the magnitude of the two effects was very similar. On grounds of parsimony, we conclude that what seemed like a desirability bias may just be a salience/marking effect, and -- though optimism is a robust and ubiquitous human phenomenon -- wishful thinking still remains elusive.