Regression to the Mean
Is this the new normal, or just a spike?
When you see an extreme result — exceptionally good or bad — write: Is this a pattern or a spike? What would an average outcome look like here? Before changing strategy, wait for at least one more data point.
You're tempted to overhaul your approach based on a single outstanding result.
You have strong evidence of a genuine structural change, not just a data point.
Why it works
Extreme results are statistically likely to be followed by more average ones, regardless of what you do. Overreacting to outliers is one of the most common and expensive decision errors.
After a record-breaking quarter, companies restructure to ‘capture the momentum.’ After a terrible month, they panic and change strategy. Both reactions assume the extreme result reveals a new truth about the system. Usually, it doesn’t — it’s a natural fluctuation that would have corrected itself without intervention. The cruel part is that whatever you changed gets the credit when things return to normal. Hired a new manager after a bad quarter? The next quarter improves, and you attribute it to the hire. But the improvement was probably coming anyway. This is why tracking matters: without a baseline, you can’t distinguish between a correction you caused and a regression that was inevitable.