The Epidemic of Overconfidence
Everyone knows someone who thinks they’re good at everything. I can name more than a few friends of mine who never seem to be as good at things as they believe, and if I were being truly honest with myself, I’d say that I probably overestimate my skill in a few areas. This epidemic of overconfidence is nothing new, as countless studies have shown that people regularly overestimate their abilities. One study showed that 70 percent of high school students thought of themselves as above average leaders, while another showed that men consistently think they are more attractive than they really are. Professors, who some might think would have a good grasp on reality, seem to be the worst offenders, as one survey reported that a whopping 94 percent of college professors said their teaching skills were above average.
As it turns out, there is an evolutionary reason for people to believe highly of themselves. A new study has used game theory to show that overconfidence tends to have an advantage over underestimating your own skill or even having an accurate assessment of your own abilities. The game used in the study is simple. There is a resource, which in the real world corresponds to something we need, such as food. There are two players (the result still applies for three or more players actually, but for simplicity let’s say it’s two). Each player can choose to try to acquire the resource or not. If one player goes for the resource and the other does not, he gets it at no cost. If both go for it, the stronger player gets the resource, but there is a small cost to both players. If neither goes for it, there is no cost to either, but neither gets the resource. In the case that the stronger player is player one, along the left side of the matrix, we get the following payoff matrix (where the top left is both players going for the resource):
8,-1 9, 0
0, 9 0, 0
If both player’s know who is stronger, and thus know the setup of this payoff matrix, it is clear that the stronger will get the resource as the weaker sits back and lets him have it. However, in the real world, we don’t know who is truly stronger before actually battling it out – there is some level of uncertainty. The study tested the outcome of such a game with varying degrees of uncertainty. They also tested it with players who overestimated their ability, underestimated their ability, and thought accurately about their ability. What they found was that in games with higher levels of uncertainty about the opponent’s strength, those who overestimated their own ability had larger payoffs over time than those who did not.
This may not seem to make sense, as we would expect players who knew there own ability precisely to avoid the costs of entering fights they’re probably going to lose. The simple explanation is that since the cost of losing is so low compared to the payoff for winning that those who are overconfident will win more with the occasional payoff than they’ll lose from the cost of losing a battle. The evolutionary benefit of overconfidence has ensured that most of us think a little more highly of ourselves than we should. I’d like to say that next time I think about my own abilities, I’ll try to be a little more accurate, but that would just end up hurting me in the long run. I think I’ll stick with the confidence. Yeah. This blog post is pretty awesome.