Testing Wisdom of the Crowd

In his book The Wisdom of the Crowd James Surowiecki reveals that, in the past, some groups made excellent decisions while other groups failed miserably at decision making via a large amount of examples. He explains that a group can make wise decisions collectively if certain requirements are met. He calls collective wise decisions ‘wisdom of the crowd’. In his paper Making the Difference: Applying a Logic of Diversity Scott Page uses the Diversity Prediction Theorem to mathematically explain that, by increasing the diversity, the crowd’s opinion becomes more accurate compared to its average individual members. To mathematically validate James Surowiecki’s reasoning, an experiment was conducted among a crowd of 119 high school students, who fit the requirements for a crowd to be collectively wise. In the experiment the Prediction Diversity Theorem was used to decide if the crowd’s opinion was significantly more accurate than the opinion of its average individual members. The theorem compares the crowd’s opinion to that of its individual members, and in this experiment a reference crowd was used to validate that the crowd’s opinion is not only accurate compared to its average individual members, but also has real-life value. This thesis will provide evidence that wisdom of the crowd occurs when asking a crowd of 119 high school students to predict the outcome of the 2014 FIFA World Cup soccer group-stage games in Brazil. It will also show that, if you separate the crowd into a high and low knowledge group, the high knowledge group has a lower individual error on average, while the low knowledge group has a higher prediction diversity. These two findings combined show that the crowd’s opinion of the low and high knowledge group do not differ significantly. Lastly, this the-sis will show that the opinion of this crowd does not significantly differ from the predictions of a betting website, showing that the opinion of the crowd has real-life value.