Regression Fallacy

The regression fallacy is a logical error that occurs when an individual mistakenly assumes that a phenomenon occurring outside of the normal range of data would continue to do so. It is a flawed reasoning where an extreme observation is expected to remain extreme in future observations, disregarding the tendency of data to return to its average or mean over time. This fallacy is based on a misunderstanding of the concept of regression to the mean, which is a statistical phenomenon.


Regression to the mean is a statistical concept that suggests that if a variable is extreme on its first measurement, it is likely to be less extreme on a subsequent measurement. In other words, extreme observations are more likely to move closer to the average or mean when measured again. The regression fallacy occurs when individuals fail to account for this statistical principle and incorrectly assume that extreme observations will remain extreme.


Example 1: A football player scores several goals in a particular match, far surpassing his average performance. Some fans may mistakenly expect the player to consistently score such a high number of goals in future matches, disregarding the statistical regression that suggests he will likely return to his average goal-scoring performance.

Example 2: A stock investor witnesses an exceptional, one-time increase in their portfolio value due to a particular investment. While they may attribute their success to their investment strategy, this could be a case of the regression fallacy if they fail to consider that such exceptional gains are unlikely to be sustained in the long run.


While the regression fallacy may lead to incorrect expectations and predictions, it is important to note that regression to the mean is a statistical probability, not a certainty. In some cases, extreme observations can indeed be indicative of a significant underlying change or trend. Therefore, it is crucial to consider other factors and gather additional evidence before drawing conclusions solely based on the regression fallacy.