"Distrust conclusions drawn from very few data points Tao recounts how astronomer Johann Bode fit a curve to planetary distances using roughly six data points, correctly predicted the existence of the asteroid belt and matched Uranus, and was celebrated — until Neptune was discovered and 'was completely way off,' revealing the whole thing was 'just a numerical fluke.' Tao notes that Kepler himself may have instinctively sensed this danger, which is why he 'didn't highlight his third law as much as the first two.' Before acting on a pattern you've spotted in a small sample, actively ask whether that sample is large enough to rule out a lucky coincidence. Source: Terence Tao, Dwarkesh Podcast, Terence Tao – Hardest Math Problems, AI's Limits, and Scientific Progress"