Why your gut is often badly wrong about what's likely and what's dangerous, and a handful of simple habits that make you think about chance far more clearly. A picture for every idea.
a scale from 0 to 100%
Probability puts a number on uncertainty. 0% means never; 100% means certain; everything else sits in between. A coin flip is 50%. Rain "likely" might be 80%. That's the whole foundation, but our instincts handle the in-between numbers surprisingly badly, which is where things get interesting.
streaks aren't "due" to end
Real randomness creates clumps and streaks that feel meaningful but aren't. Flip a coin and you'll often get several heads in a row. That's normal, not spooky. And a coin is never "due" for tails after five heads; it has no memory, so it's still 50/50 every flip. Believing otherwise (the "gambler's fallacy") drains a lot of wallets.
memorable ≠ likely
We judge danger by how easily a scary example comes to mind, not by how common it actually is. Plane crashes and shark attacks are vivid and make headlines, so they feel dangerous, yet everyday risks like car crashes and heart disease harm vastly more people while barely registering as scary. The dramatic risks grab attention; the common ones do the real damage.
base rates and scary headlines
A headline screaming "this doubles your risk!" sounds terrifying, but doubling a tiny risk still leaves a tiny risk. Going from 1-in-a-million to 2-in-a-million is "double," yet basically nothing. The missing question is always the base rate: how common was it to begin with? Without that, a percentage change is meaningless theater.
expected value
Smart decisions combine how likely something is with how big the outcome is: that's expected value. A lottery has a huge prize but such tiny odds that, on average, you lose. Insurance flips it: a small, near-certain cost protects against a rare but devastating loss. Same math, opposite verdict, because the size of the outcome matters as much as the odds.
a few habits that fix most mistakes
You don't need advanced math, just better instincts:
• Think in counts, not percentages: "3 out of 1,000" is clearer than "0.3%".
• Distrust tiny samples: a few results prove little; randomness needs room to average out.
• Separate "scary" from "likely": ask if your fear is driven by drama or by frequency.
• Always ask for the base rate: "compared to what? out of how many?"
• Weigh size × odds before any big decision or bet.
Probability is how likely, from 0% (never) to 100% (certain).
Randomness clumps; streaks are normal and nothing is ever "due."
We fear vivid rare events more than common deadly ones.
Ask "out of how many?" Doubling a tiny risk is still tiny.
Expected value: weigh how likely against how big the outcome is.
Think in counts, distrust small samples, and separate scary from likely.