Game Theory behind Crowdsourcing Competitions
For many companies. hosting crowdsourcing competitions is an increasingly popular solution to solving a problem. The idea is to consult the public directly via the Internet, where individuals can contribute and collaborate on a project. In theory, crowdsourcing is a powerful and efficient method, because it takes advantage of the skills and talents of thousands of people. Furthermore, these companies can encourage individuals to compete by offering a reward. However, a recent study published by the Journal of the Royal Society Interface reveals that there is a downside to this method. Due to the shared nature of crowdsourcing, the progress that each competitor makes is open to the public. This leaves his work open to sabotage from the other players. The study shows that sabotage is actually quite frequent, and that raising the cost of sabotage counter-intuitively encourages players to attack another’s work.
Researchers analyzed player behavior in several crowdsourcing competitions and found that the game theory behind the competitions can be modeled by Prisoner’s Dilemma. Say player 1 is a strong competitor with the current best solution to the problem and player 2 is a weaker competitor with a mediocre solution. The players can either choose to “attack” or “not attack” the other player’s work. However, attacks are costly. For instance, the costs could represent the time and effort taken to coordinate the attack or the consequences of getting caught. Assuming that the attacks dramatically set back the victim’s progress (which has happened in past competitions), then the attacker wins the competition. Otherwise, the stronger player would win the competition.
The striking results of the model show that weak competitors are more likely to attack and strong competitors tend to hold back when the cost of attack is raised. Furthermore, increasing the cost only increases the likelihood that the weaker players would win the competition. From a sociological perspective, the weaker player is more willing to attack and risk the costs. If he can successfully sabotage the stronger player’s work, then his chances of winning dramatically increase. Moreover, the costs don’t mean as much to him, because if he loses, he doesn’t gain anything. However, the stronger player isn’t willing to take that risk because he doesn’t want to lose his upper hand (a better solution). The reasoning behind these choices become more pronounced as the cost increases.
Consequently, the trade-offs of crowdsourcing aren’t as well represented as they should be. When crowdsourcing turns into competition, competitors are more prone to sabotage each other instead of collaborating and building off each other’s work. Just as in Prisoner’s Dilemma, instead of cooperating and mutually benefiting from each other’s work, the players work in their best interests.