Information Cascades in Fantasy Sports and Professional Scouting
The first article linked below is from a fantasy football website in which the author warns of how information cascades can lead to incorrect conclusions. A relatively recent invention, fantasy sports allow the common fan to essentially absorb the responsibilities of a general manager and draft and trade players to put together a competitive team. However, unlike the actual GM’s, most who partake in fantasy sports have lives and jobs outside of sports, so they cannot watch ever game and attend every practice, and likewise, they likely do not have the skills to easily evaluate players. As a result, many sites that sponsor fantasy sports will pay so-called ‘experts’ to evaluate the players and predict their future successes and failures. This often results in an information cascade wherein a fantasy GM will not only take into account his own diagnosis of a particular player, but also that of these experts. This cascade is further compounded by readily available data on how owners in other leagues have drafted, added, dropped and started particular players, which allows a fantasy GM to go against his own analysis and use that of others. While each person may be performing in his best interest, the total result may be further from the truth than if each person acted individually.
This result can be expanded from fantasy sports to actual sports leagues, especially in respect to how players are scouted. For example, when a Major League Baseball team goes to draft or sign a new player, they must search through the collegiate level, high schools, international leagues, and foreign club teams. As a result, even exceptional talents can be passed over, as event the exceptional ones are seen by only a few scouts. One example of this occurring is the scouting of Mike Trout, an outfielder for the Los Angeles Angels who through full three seasons won the 2014 AL MVP, and was twice runner up. Having had arguably one of the greatest starts to career in baseball history, how did Trout fall all the way to the 25th pick of the 2009 draft? One would think that such an exceptionally talented player would be easily recognizable to a scout, especially considering his immediate success. I contribute this in part to information cascades. Trout, who was drafted right out high school, played and attended school in New Jersey, where cold weather significantly shortens the season, and gives scouts even less time to evaluate New Jersey players than their peers in California, Texas, Florida and other warm weather states. As a result, an incorrect consensus is much easier to come by, and when a higher-level scout comes by to evaluate a player like Trout, he has fewer independent analyses from lower level scouts, so each scout’s evaluation gains significantly more value. Eventually, by the time the evaluation is sent to high-level decision makers, an information cascade could have formed, possibly resulting in an incorrect result, as was obviously the case with Mike Trout.
Sources:
http://thefakefootball.com/meta-gabeing-unconscious-recall-and-information-cascades/