My Way of Thinking Does Not Make a Very Good Prediction Tool

The older I get and the more experience I receive in this world, the more I find that I cannot adequately or consistently predict the behavior of other people. My own viewpoint, my own knowledge, and my own experience that I have gleaned over the years are far too limited and far too narrow in scope to provide an adequate estimation of human behavior overall. I assume that the way I think, that the way I view the world is the only way to do so and that everyone
simply must view and think the exact same way. So I make predictions and judgments based on this assumption and am annoyed when the predicted behavior does not occur because it means that I was obviously in error. I forget that the knowledge and experiences that shape my predictions and judgments are unique to myself. No one else in the world has experienced the world in quite the same way as I, and no one else has had the same interests or learned the
same things or seen quite the same things as I. So my way of seeing the world is unique, and I cannot expect anyone else to behave or think the same way I do. Additionally, I make the common mistake of relying on anecdotal evidence to support my predictions and judgments. The trouble with doing this is that anecdotal evidence is typically not representative of the general population. It is merely one example of human behavior that may support a given idea or assumption. It is very problematic when people rely too heavily on anecdotal evidence to support decisions or judgments. Making snap judgments is also dangerous for another reason — psychological studies have shown that human memory both decays rapidly and is frequently rewritten unconsciously over time, effectively altering or even corrupting the very information upon which we rely so heavily.

The beautiful thing about statistics is that it often corrects for the shortcomings in human judgment by aggregating a large pool of information into relatively simple descriptive numbers. A good predictive statistic will have a good sample of data that is representative of the population at large. If the sample is representative enough, the results of the analysis will generalize well to everybody, meaning that it has a much better chance of accurately predicting behavior by providing a probability value for the occurrence of a particular behavior. Naturally, there will always be those individuals who are so far different from the general populace as to be outliers, and no statistic will completely describe any single individual. However, statistics are meant to be descriptive of large populations and give the inquisitive mind a better chance at accurately predicting how people will behave. There will also be
those unscrupulous individuals who will bias the data in such a way as to serve their own purposes. But for the wise and those genuinely seeking to learn, the use of statistical sampling can counteract the shortcomings of the human condition. It has been proven time and again that statistics can predict human behavior much more accurately, effectively, and consistently than can human predictions based solely on head knowledge.

All this is to caution you against relying too heavily on your own snap judgments (and even against well-thought-out, yet unsubstantiated judgments) because chances are good that your judgments and predictions will be erroneous. The numbers are useful tools to have in one’s
‘toolkit’ and should be used frequently to avoid making crucial, critical mistakes. Nothing replaces good, sound research in the endeavor to make good decisions.

Have anything to add to the conversation?