Confirmation Bias -- Introduction

When it comes to sports betting and gambling in general, an understanding of confirmation bias has immense value. I'll tackle various aspects of confirmation bias in at least a dozen posts in the months ahead. I think that it's the most convenient and useful starting point from which to explore sports betting pitfalls and errors.

Many 2019 references to confirmation bias address the current political situation in the United States. Observers see the U.S. electorate divided into hardened camps almost impervious to evidence that disconfirms their respective political positions or that places favored representatives in a bad light. Common nomenclature describes Americans as existing in "echo chambers" or "silos." These terms reference social and informational spaces that are selectively permeable, like a human cell. Information that feeds the echo chamber is allowed in; conflicting data is filtered out. The more the echo chamber fills with one-sided information, the more the pressure difference grows between those inside the chamber and those outside of it.

American politics has thrust confirmation bias and all of its elements into the spotlight. I would, however, like to set aside the political aspects of confirmation bias and examine instead the awesome utility of the concept when applied to sports betting and gambling. So, for our purposes, confirmation bias will refer to the human tendency to vet information so that any resulting conclusions conform to pre-existing theories, expectations, and beliefs.

We selectively filter our acquisition of information both "out there" in the world and information catalogued in memory so that we require the least reorganization of our existing "reality." Confirmation bias is like a personal and flawed version of Occam's Razor. Instead of the simplest explanation as our default, however, we default to the explanation that requires the least adjustment to what we already believe. Consciously or not, we adhere to the premise that our beliefs are best kept crystalline and unaffected by conflicting evidence.


Cognitive Dissonance

Cognitive dissonance theory is an ancestor and close cousin of confirmation bias studies. Leon Festinger introduced cognitive dissonance theory in his classics, When Prophecy Fails (1956) and A Theory of Cognitive Dissonance (1957). When Prophecy Fails will be a focal point for another column. The book deals with a UFO cult whose leaders predicted events on specific dates. The events did not occur, but the cult members surprisingly became even more invested in the leaders and the cult itself after the prophecies had failed.

Cognitive dissonance was one of the first theories to address things like the fact that people who have recently bought one brand of automobile tend to not read ads for other brands. These tendencies are statistically predictable. While cognitive dissonance may have been a useful (and at the time, novel) tool for asking real world questions, many of the older laboratory studies seemed quite unwieldy. Today, the subject has for the most part been subsumed under the much broader helm of confirmation bias.


Overview

Our entire individual lives are a test of our ability to overcome confirmation biases and the associated narcissism. We live these 70-odd years, pinned down by whatever culture in which we're born, and our life experiences are limited. I'm talking brass tacks, bottom line, numerically limited. Despite this, we are constantly faced with drawing conclusions and making decisions based on the individual data bases of our lives. Quite often, we don't have a statistically significant sample from which to generate future decisions, but either we don't recognize this, or we don't have the ability to de-emphasize our life histories as influences. Our minds are filled with the full color narratives of ourselves. We remember every pain or pleasure without a sense of their lack of importance. The challenge for us is to recognize when our lives have not given us a statistically significant sample, to go get the info required to take statistically sound action, and to then override the flawed perspectives our lives have provided. It's a daunting task.

In my life, I have had a single terrible experience buying a Renault automobile and a single great experience buying a Toyota. These were vivid events in my life, and they are writ large. The reality is that these events have no statistical significance. I am capable of recognizing this, so I will read Consumer Reports before car shopping. If Renault, however, gets top grades, I will still have a very, very difficult time putting my life experiences aside. It's hard to subordinate the evidence of our lives, however sparse, to some third party summary that we haven't personally experienced. The job, however, is to do just that. From a statistical perspective, the idea that "what we've lived should guide us" is largely an illusion. For many things, our lives don't provide a big enough sample from which to draw meaningful conclusions.

We view the world with perceptual biases of which we are unaware. We process information in a sea of memory biases. We may be overmatched by our biases. I think, however, that in the arena of sports gambling, we have a chance to improve ourselves by recognizing and controlling these biases. We'll discuss how in the weeks ahead.


Bob Dietz -- June 20, 2019