Madden Effect

The following was written by a friend of mine who teaches an advanced psychology course.  Her description of this psychology trend also helps to explain post hoc ergo propter hoc.  Read the following and comment - what are your reactions, beliefs, etc?

The ability to explore alternative explanations for a phenomenon is an important skill for introductory psychology students to develop. People who lack this skill tend to fall prey to mistaken beliefs and superstitions. An interesting example of this is the so-called “Madden Curse.” The Madden Curse refers to the tendency for football players featured on the cover of the Madden NFL videogame to perform poorly the following season, sometimes as the result of injury. For example, following his appearance on the videogame’s cover, quarterback Daunte Culpepper went from one of the league’s best players to struggling with injuries and poor performance. Similarly, Donovan McNabb, Shaun Alexander, Michael Vick, and Troy Polamalu all underperformed or suffered injury after their cover appearances.

Although few people take the Madden curse too seriously, it is a great example of the logical fallacy post hoc ergo propter hoc (i.e., one event followed another, therefore, the first event caused the second). In other words, the curse would have it that appearing on the cover causes the subsequent decline in performance. Of course, alternative explanations for the phenomenon are seldom considered.

So what are some of the alternative explanations for the Madden Curse? One likely candidate is regression to the mean: when a subject with an extreme score is remeasured, it is likely that the score will be less extreme (i.e., closer to average). For example, a person who scores extremely fast on a reaction time task will most likely have a slower time when retested the following week. This can create the illusion that something has affected the person’s performance, when regression to the mean is simply the result of having an extreme score to begin with. Because only players who have extremely good seasons are selected for the following year’s Madden NFL cover, it is likely that these players will “regress to the mean” and have a less stellar season the following year. Importantly, this does not mean that players featured on the cover are any more likely to have a poor season than other players in the league, but that they are simply unlikely to perform as well as they did the previous season.

In addition to regression to the mean, another explanation for the Madden Curse is the confirmation bias. The confirmation bias is the tendency for people to pay more attention to evidence that supports their beliefs, and less attention to evidence which contradicts their beliefs. Although it is true that many of the players featured on the Madden NFL cover have fallen victim to the curse, some players have not. For example, linebacker Ray Lewis actually went to the Pro Bowl following his cover appearance. These counterexamples tend to be given less consideration by those who believe in the Madden Curse. Meanwhile, each time a featured player’s performance declines, discussion of the curse begins anew.