The Convergence
One Pattern, Seven Disguises
The Pattern
You've been a doctor reading test results, a radio astronomer scanning for signals, a data journalist calling an election, a Bayesian reasoner weighing evidence, a trader parsing market noise, and a juror evaluating witnesses. Six different rooms, six different problems. And in every room, the same invisible structure determined who was right and who was fooled.
The structure is always the same: a prior probability, a piece of evidence, and a question — how much should this evidence change what I believe? The equation that answers this question has been hiding in plain sight through every chapter.
The Proof
Drag the shared parameter slider. As it moves from left to right, watch all four visualizations respond. The dot grid shifts from gray to red as the base rate rises. The waterfall display crosses more thresholds. The election curve slides past 50%. The belief meter fills.
One number. Four transformations. Same equation. The slider maps to a different parameter in each model — base rate, detection threshold, poll average, prior belief — but the mathematical structure underneath is identical. Bayes' theorem doesn't care whether you're diagnosing patients or detecting aliens.
Why It Matters
This isn't a formula to memorize for an exam. It's a way of seeing. Every time you encounter a test result, a surprising headline, a confident prediction, or a persuasive argument, the same questions apply: What was the prior? How strong is the evidence? And how much should your belief actually move?
The chapters taught you the failure modes. Base rate neglect makes accurate tests misleading. Low signal-to-noise ratios make thresholds treacherous. Systematic bias defeats aggregation. Correlated evidence masquerades as independent proof. These aren't abstract risks — they're the specific ways that intelligent people get fooled every day.
The next time someone tells you a test is 95% accurate, or a stock moved 3%, or two witnesses agree — you'll know the question they forgot to ask. And you'll know the math that answers it.