Texas Sharpshooter Fallacy
The Texas sharpshooter fallacy selects a pattern after seeing the data and then treats the selected matches as unusually strong evidence. Ignored misses and alternative patterns can make the apparent fit misleading.
THE IDEA
Texas Sharpshooter Fallacy
(also known as: clustering illusion)
Description: Selecting similarities after inspecting the data can create an apparent pattern while hiding misses or alternative comparisons. The target is effectively drawn around the hits.
Example #1: The “prophet” Nostradamus wrote about 500 years ago. The actual passage claimed to predict Hitler is not supplied here.
Explanation: When you focus on just that prediction, then it might seem that way. But don’t forget, Nostradamus made over 1000 “predictions”, most (all?) of which are vague nonsense. With so many predictions and flexible interpretations, a retrospective match need not be surprising. Its probability cannot be calculated without a defined matching rule. Again, if you ignore the noise (the predictions that do not make any sense), it looks amazing. By the way, “Hister” is the Latin name for the Danube River.
Example #2: The original dating-service claim is missing.
Explanation: What SuperCyberDate.con did not take into consideration was the 245 other likes and dislikes that were very different for both Sally and Billy – like the fact that Billy is gay.
Exception: Focusing on similarities can be legitimate when the comparison is defined in advance. Check whether the matches were selected after the result and whether the misses change the conclusion.
When to use it
Analyzing financial market patterns to avoid seeing false trends that may lead to poor investment decisions.
Evaluating scientific studies that claim significant results without considering multiple comparisons or the number of hypotheses tested.
Interpreting coincidences in daily life without jumping to conclusions about fate, design, or supernatural intervention.
Reviewing customer feedback data to distinguish genuine patterns from random noise before making business decisions.
FROM THEORY TO EVERYDAY LIFE
In practice
Nostradamus
The “prophet” Nostradamus wrote about 500 years ago: Surely he must have had some vision of Hitler! Explanation: When you focus on just that prediction, then it might seem that way. But don’t forget, Nostradamus made over 1000 “predictions”, most (all?) of which are vague nonsense. Many vague predictions offer opportunities for retrospective matches. A claim of statistical impossibility would require a specified matching rule and probability model. Again, if you ignore the noise (the predictions that do not make any sense), it looks amazing. By the way, “Hister” is the Latin name for the Danube River.
SuperCyberDate.con
The supplied commentary says SuperCyberDate.con overlooked 245 differences between Sally and Billy, including Billy’s sexual orientation. The original matching claim is absent, so the selected similarities cannot be compared with those differences here.
A MOMENT TO REFLECT
Put it to work.
Think of a time when you focused only on data points that supported your belief. How might the Texas Sharpshooter Fallacy have influenced your conclusion?
How can you avoid the clustering illusion when interpreting statistical data?
Why is it important to consider the total number of predictions or data points before concluding a pattern is meaningful?
CONTINUE THE CONVERSATION
Source texts
- Logically Fallacious - The Ultimate Collection of Over 300 Logical Fallacies