TedEd: Should you trust unanimous decisions?

A TedEd by Derek Abbott

Many of us are inclined to believe unanimous decisions. But although they seem very strong, they aren't as infallible as we think. Much of our society relies on majority consensus, which can be a good thing sometimes. But at a certain point, getting closer to consensus gets us farther from the actual results. This is the Paradox of Unanimity. In cases with an obvious answer, unanimity is normal. But when there is some natural variance, there should also be some varied distribution. For instance, a bunch of coin tosses is expected to come up heads around 50% of the time. But it would be suspicious if they started nearing 100%, and you might suspect tampering with the coin. In fact, a study in the 1990's showed that nearly half of witnesses pick the wrong suspect. This is because our short term memory is only so good, and our brain automatically fills in gaps. There can also be systemic errors aside from human judgement; for example, accidental contamination of cotton swabs used by police led to the same DNA being found in many different crime scenes. And, of course, there is outright fraud itself, often occurring in politics. Still, unanimity is useful in situations with low variation, just not in ones with high probability of variation. In the latter, there is probably an external influence.