Skip To The Main Content

When Being Smart Isn’t Enough: Texas Sharpshooter Fallacy and Seeing Patterns in Data - Part 3

APICS Senior Director, Professional Development

Tuesday December 6, 2016


As I’ve shown in previous blog postings, cognitive biases are like built-in software. You don’t have to purchase them or download them. They come pre-loaded in your subconscious.

One of these pre-loaded predispositions is the need to seek out and identify patterns that help us understand the world around us; to find causes for things we can’t explain. Our brains are very good at seeking out patterns. It is a human survival skill. We have a need to make sense of the events in our lives, and so look for patterns to help us explain them. Thousands of years ago, those inhabitants on the Serengeti plain that could quickly recognize when the shadows behind the bushes didn’t look right and leave the area quickly, lived long enough to become our ancestors.

Pattern recognition is an important ability. For example, reading is a form of pattern recognition. Neuroscientists tell us that we don’t actually read words, not even letters, but that we recognize shapes (patterns) that the brain has learned to decode as words and letters (https://www.scientificamerican.com/article/when-we-read-we-recognize-words-as-pictures-and-hear-them-spoken-aloud/).

Yet we often impose patterns where they don’t exist. That was OK for those ancestors mentioned above. Running away when there was no predator behind the bush was far less dangerous than not running away when there was.

An interesting example of imposing patterns where they don’t exist is called the Texas Sharpshooter Fallacy.

“If hindsight bias and confirmation bias had a baby, it would be the Texas sharpshooter fallacy.” - David McRaney, “You Are Not So Smart.” (Page 42)

We discussed the Confirmation Bias in our last blog installment. It is that tendency to unconsciously seek information that agrees with our existing views and ignore information that does not.

What is Hindsight Bias? “The hindsight bias reflects a tendency to overestimate your own ability to have predicted or foreseen an event after learning about the outcome.” (http://www.psychologyandsociety.com/hindsightbias.html)

For example, when I was preparing for the APICS 2016 Conference presentation than underlie this short series of blogs on cognitive bias, I expressed some trepidation that the presentation was cohesive enough, that it would make sense as a flowing narrative, that my examples were neither too simple nor too complex.

After the presentation was given, and, to all appearances, successfully, I started saying something like: “Well, that went well. I knew it would.” My friends took some small pleasure in reminding me that before giving the presentation, I had, in fact, not been certain of success, but was somewhat doubtful. My change in attitude after the fact was hindsight bias in full play.

So, what is the Texas Sharpshooter Fallacy?

Imagine a Texan sharpshooter who is standing facing the side of a barn. Now, imagine that he starts shooting at the barn. Over time, the side of the barn becomes riddled with holes. In some places there are lots of them, in others there are few.

Now imagine this cowboy walks up the side of the barn, examines the arrangement of the holes carefully and then paints a bull’s-eye over a spot where his bullet holes clustered together. it looks like he is pretty good with a gun. By painting a bull’s-eye over a cluster of existing bullet holes, the cowboy places artificial order over natural random chance.

If you have a human brain, you do this all of the time. Picking out clusters of coincidence is a predictable malfunction of normal human logic. It happens when we cherry pick data, or look for a pattern that will fit an existing presumption, or come to a large amount of data unsure what we are looking for. It is one of the ways we insert meaning

A Swedish study in 1992 tried to determine whether power lines caused some kind of poor health effects. The researchers surveyed everyone living within 300 meters of high-voltage power lines over a 25-year period and looked for statistically significant increases in rates of over 800 ailments. The study found that the incidence of childhood leukemia was four times higher among those that lived closest to the power lines, and it spurred calls to action by the Swedish government. (https://www.ncbi.nlm.nih.gov/pubmed/8213751; https://www.ncbi.nlm.nih.gov/pubmed/11744509

The problem with the conclusion, however, was that the number of potential ailments, i.e. over 800, was so large that it created a high probability that at least one ailment would exhibit the appearance of a statistically significant difference by chance alone; i.e. the multiple comparisons problem. Subsequent studies failed to show any links between power lines and childhood leukemia, neither in causation nor even in correlation.

The insidious thing about this pattern seeking ability is that, once we see a pattern, it is very hard for us to ignore it. And we can’t turn off the pattern seeking machine in our head.

How do we overcome pattern seeking and the Texas Sharpshooter Fallacy? You cannot keep it from happening, so become a skeptic – not a “I don’t believe anything you say” skeptic, but a “I’d like to see the corroborating evidence, if you please” type of skeptic – before committing to the pattern that you initially see.

Second, become comfortable with uncertainty. Easier said than done, I know, but consider what you do and don’t yet know, and how you know what you know, and use your prefrontal cortex – the home of your Executive Function – to review the quick and easy answer before accepting it.

More posts in this series

When Being Smart Isn’t Enough: The Blind Spot Bias - Part 1
When Being Smart Isn’t Enough: Shephard’s Tabletop Illusion - Part 2

All comments will be published pending approval. Read the APICS Comment Policy.

1 comment

Leave a comment
  1. Gary Smith December 17, 2016, 09:52 AM
    Great article, Bob.  I see this type of false pattern recognition all the time.  This is why I always try to, "let the data tell the story."  While we should always start with a hypothesis, proving or disproving it should not be the objective.  The objective should be the advancement of knowledge.

    Leave a comment

    1. New code



    Website Alert

    The apics.org website will have limited functionality 8 p.m. to 2 a.m. CT, Sunday, 12/17/17. Purchases and changes cannot be made during this time. We apologize for the inconvenience.
    XClose