Description vs. Prediction: Predictive Theories and Descriptive Stories

Besides all formal, scientific, epistemological, definitions of science and scientific theories, there’s a simple clear understanding of the science in our mind. Something we have learned in school and experienced later in our daily life.

Scientists observe the world. Develop theories based on their observations. Validate their theories using existing data and finally, use these theories to predict the future.

The science without the power of prediction is history at best and story at worst. Just suppose someone tells you that all the waters on the earth were boiled at 100 C till now, but we know nothing about the future!

Prediction vs Description - Predictive Models or Descriptive Stories

In this sense, something like statistics would not be part of the science as it describes what happened and predicts the future just with the ceteris paribus assumption. Such an assumption will not hold true for most of the complex systems.

Statistics can be considered as a tool helping scientists to compare theories with each other to see which one had a better prediction power. That’s all.

Actually most of the spiritual and metaphysical descriptions of the world are descriptions of the past rather than predictions of the future. They are able to vividly and firmly describe the reason behind every past event but have almost no power in predicting what will happen in the future.

As one the most complicated species ever existed on the planet, seems that our next aim shall be developing a mindset primed for predicting the future instead of merely describing the past.

Sure most of us are interested in future. The long time existence of fortunetellers from Oracle of the Delphi up to today’s palmists and phrenologists proves that there has been kind of inclination toward predicting the future. But it’s a great difference between the people with the predictive mindset and the ones with descriptive mindset who believe they are able to predict the future.

One of the best business cases would be market researchers. Most of them just provide us with the description of the current preferences and behaviors of the customers based on some statistical reports and convince us that these reports are the most reliable roadmap to the future. Sure if nothing radical happens, this kind of analysis can be valid. But why we need such an analysis if there’s nothing radical going to happen?

Sure there’s not a clear dividing line between descriptive models and predictive ones, but it’s not so hard to compare them from this aspect. The predictability power of the models also depends on the subject of the study. Sure physics can claim a more power of predictability in comparison to the psychology. As the later on studies a more complex subject.

In general I believe that for thousands of years, most of the models proposed by humans for the complex systems such as the universe, the nature, the weather, the society and themselves, has been descriptive in some sense and the future belongs to the models which are aiming at the prediction of the behaviors of these complex systems.