INDIANAPOLIS—It is believed by most that many brains are more powerful than one. So, when it is time, for example, to guess how many gumballs are in a jar, the average of the group’s guesses is probably better than most of the individual guesses.
But, there isn’t much out there that really explains why that is, says Ryan Rogers, Butler University Assistant Professor of Entertainment Media and Journalism.
Rogers looked into this concept using one of America’s favorite past times—filling out March Madness brackets. He wanted to find out what exactly makes collective intelligence effective.
“Yes, we know crowd sourcing is beneficial, but what are those traits, and tasks, that are going to make the group impactful in its decision-making process?” Rogers says. “What kind of group is most effective and what kinds of tasks lend itself to crowd sourcing?”
Individuals were divided up based on their backgrounds and expertise in college basketball. One group was made up of serious college basketball fans. The other group was made up of college basketball experts, for example, journalists, former players, coaches, or others with insights beyond just being an engaged fan.
Each group then filled out NCAA tournament brackets using collective intelligence software. The goal, Rogers says, was to see how group make-up would impact the effectiveness of collective intelligence, and therefore, the infamous activity of avoiding a busted bracket after, well, one round.
The results, published in the Journal of Creative Communications, showed that the experts and the fans performed similarly throughout the first few rounds of the tournament. However, the experts gained a real edge over the fans as the tournament progressed—as the task became more difficult. When it came to the later rounds—games that are typically more challenging and complicated to predict—the experts had more success in picking winners than the fans.
“There’s a passion and there’s an interest,” he says. “It is not just about having a buddy who knows basketball, but our study showed that it is about the group dynamic, and that specific traits impact how successful the group will be. In addition to the traits of a group, our study showed task matters, too. The more difficult the task, the more important the make-up of the group.”
The results are important, Rogers says, because they can be applied to many fields and subject matters much more complicated than guessing gumballs in a jar or filling out a bracket.
The experts separated themselves in the later rounds of the tournament—when the task was more complicated and collective wisdom, therefore, mattered more, Rogers says. This distinction is a crucial finding.
When it comes to solving a complex engineering problem, for example, he says, it would be important to think about getting a group of experts together. Rogers compares that to asking a bunch of stargazers to solve a complex astrophysics problem. Collective intelligence, he says, wouldn’t help that group.
“Their love of the subject matter won’t matter because the topic is highly complex,” he says. “They simply don’t have enough technical knowledge to leverage the wisdom of the crowd. That is what, essentially, this study teaches us. It is not just that many brains are better than one, but who the group is made up of that impacts its effectiveness.”
Media contact:
Rachel Stern
Director of Strategic Communications
rstern@butler.edu
914-815-5656