Baloney Detection 101: the Clustering Illusion

It doesn’t seem to make sense that truly random events would bunch together, or cluster, and this, the Clustering Illusion, is the perception that such events are non-random, unusual, significant, and meaningful.

This is based on a false supposition, derived from the psychological phenomenon of subjective validation, also known as selective thinking, or selection bias, a tendency to remember the events that stand out, or “hits,” and to dismiss from one’s mind and thus to forget the “misses,” those events that do not appear to get one’s attention or notice.

For example, while it seemingly makes no immediate sense that in a series of 20 coin flips that there is a 50% likelihood of getting a result of 4 heads in a row, and the fact that in any particular community, there might be a statistically significant number of those diagnosed with cancer, it’s the math, not our intuition that is correct, and once you actually do the math, it makes a lot more sense.

It would be unusual, highly unexpected, and highly improbable that each of only 20 coin flips would be the opposite of the previous flip. And it is even less likely that in any given sequence of random coin flips, that short runs will give what would logically be expected.

In any small sequence of random events, a wide range of probabilities, even and especially those that run counter to what we would consider sensible, can and should be expected to happen. Statistically odd events not only do happen, they can be expected to happen by the laws governing chance alone, without the need to invoke anything out of the ordinary.

Just because an apparently paranormal event happens more frequently than chance would seem to indicate, it does not logically follow that it is not due to chance. The laws of probability as we know them can and do predict such events, and these clustered events are random, even if they seem to be immediately unexplainable and non-random.

Anomalous cognition researchers often mistakenly interpret a run of apparent successes by their test-subjects as evidence for psychic ability, or seemingly statistically significant failures as evidence of ‘psi-missing‘ or Antipsi, and that such varies over time.

This is derived from simply ignoring or ignorance of perfectly ordinary random probabilities. The clustering illusion is also known in logic as the Texas Sharpshooter Fallacy, the Belief in the Law of Small Numbers, and the Division Fallacy, the erroneous assumption that parts of a whole are identical to the whole.

Related Articles

• I Knew You Were Going to Say That (randi.org)