No matter what you say, decisions are driven by emotions and whatever you may think, you are not in total control of your emotions and your emotions are not consistent. Nor is your world view an unbiased true picture of reality.
The emotional state we’re in impacts our choices and predictions as to cause them to be inconsistent. On top of that, we are fully loaded with biases that skew our decisions in ways that are hard to control.
Biases and emotions are wired into our decision making process like a circuit board with different circuit breakers, resistors, inductors and transistors. Sure, when looking at it and studying it, all components seem to behave in predictable ways but when we, ourselves are the subjects, we have a tendency to fall for every trick in the book. Even the oldest ones.
Yes, there is evidence that educated and well informed investors will make more rational investment decisions, meaning more consistent with financial theory, than less experienced investors. But, cognitive biases are hard wired into our system and if we manage to gain control of one of them and go against some of our natural instincts, there will be other biases in line ready to skew our judgment and cause us to make stupid choices.
In the article Clinical Versus Mechanical Prediction, a meta-study (a study of studies) by Grove et al published in Psychological Assessment (2000), 136 different studies were compared to evaluate the efficiency of mechanical prediction vs human experts. These 136 studies covered anything and everything from college academic performance, psychiatric diagnosis to prediction of IQ. The mechanical predictions were equal to or better than experts 94% of the time. Experts beat models only 6% of the time.
Other studies show that when the experts use a model, the accuracy of the experts’ predictions increases. But they are still not on average better than a mechanical prediction.
Explanations in the study and in the literature suggest 3 core reasons for this:
Mechanical models are consistent and aren’t affected by emotions. Models treat the same data the same way every time.
Mechanical models are better at handling statistics, e.g. assigning weights to different factors and taking correlations into account. Humans tend to overweight the importance of insignificant data, simply because it’s available to them.
Studies show that experts don’t always get adequate feedback that would help them adjust their judgements. In fact, experienced psychologists scored similar to psychology graduate students.
So, what’s the use for experts? Are they useless? No, experts are the ones programming the algorithms and models. Experts need to design the models but computers need to implement the model!
It seems like we’re doomed to be flawed in our judgments, but that’s ok! So what if you made a stupid decision? Life is all about learning by making mistakes. That seems to be the only way to re-wire ideas that aren’t working in our favor. If you haven’t made a stupid choice lately, you might want to check your pulse and see if you’re alive.
Nov 17, 2016
- Posted by Anders Stenkrona
- On November 14, 2016
- 0 Comments