Your boss didn’t reply to your message. It’s been six hours. Your brain, helpful as ever, has already written three endings: they’re furious with me, I’m about to be fired, the whole team is talking about me behind my back. None of these has any evidence. All of them feel true. And the boring fourth option — they’re busy and haven’t opened it — barely got a vote.
This happens because the world hands you events, not explanations. The message goes unanswered; the website goes down; the friend doesn’t show. Each event fans out into a crowd of possible causes, and you can’t chase all of them. You need a fast, cheap way to decide which explanation to bet on first — before the evidence is in, before you’ve spent an afternoon investigating the wrong thing.
What a “razor” actually is
A razor is a rule of thumb that “shaves off” unlikely explanations so you can focus on the promising ones. The name is a metaphor: you’re trimming away the explanations that probably don’t belong, the way a razor trims away what you don’t want to keep.
Here is the single most important thing about a razor, and we’re going to repeat it until it’s annoying: a razor gives you a prior, not a proof. A prior is a smart starting bet — your best guess before the full evidence arrives. A proof is a settled conclusion. A razor only ever does the first job. It tells you where to point your attention first; it never tells you that you’re right. Forget this, and a razor stops being a tool and becomes a blindfold — a story you defend instead of a guess you test.
Before you read — take a guess
Before we define anything — what is a 'razor' (like Occam's) actually for?
The two razors you’re here to learn
This course teaches the two most useful razors ever invented. They trim different kinds of clutter.
Occam’s razor trims complexity. Among explanations that fit the facts equally well, it says prefer the one that requires the fewest assumptions. The website went down right after you pushed a deploy? “The deploy broke it” needs one assumption; “a hacker timed a zero-day attack to your deploy window” needs three or four. Bet on the deploy first.
Hanlon’s razor trims paranoia. It says never attribute to malice what is adequately explained by incompetence or plain carelessness. Your boss’s silence? “They’re plotting against me” assumes intent, coordination, and a reason. “They’re swamped” assumes a busy person. Bet on busy first.
The one-sentence version of each
Occam: prefer the explanation with the fewest moving parts. Hanlon: prefer “someone messed up” over “someone’s out to get me.” Both are bets you make first and then update — never hills you die on.
Notice they’re cousins. Hanlon’s razor is really just Occam’s razor pointed at human behavior: a malicious plot is an assumption-heavy explanation (it needs intent, competence, secrecy, and coordination all at once), while a simple screw-up needs almost nothing. We’ll make that link explicit — and quantitative — as we go.
The map of the course
Four short teaching lessons, then an exam you can’t undo. The route:
- Occam’s Razor — what it really says (and the two things everyone gets wrong about it), how to count the assumptions an explanation smuggles in, and why “simpler” never means “simplest at all costs.”
- Why Fewer Assumptions Win — the arithmetic under the razor. Every assumption is another thing that has to be true, and probabilities multiply downward, so more moving parts means more ways to be wrong. This is where the razor stops being a vibe and becomes math.
- Hanlon’s Razor — incompetence before malice, why it’s the antidote to two specific thinking errors, and the guardrail that stops it from excusing real bad actors.
- When the Razors Lie to You — the most important lesson: exactly where each razor fails, so you can use them without being used by them.
Then a Final Exam — graded, one question at a time, and one-way: once you answer, it locks. No back button, no retries.
A coworker sends you a blunt, comma-free one-line email and you feel snubbed. Which response uses Hanlon's razor?
How to use this course
One rule does most of the work: guess before you peek. When you hit an exercise, commit to an answer in your head before revealing anything. The little sting of being wrong is what makes an idea stick — a smooth, agreeable read teaches you almost nothing. The exercises aren’t a quiz on the lesson; they are the lesson.
Next up: lesson 1, where we put Occam’s razor under the microscope and learn the one move it’s really asking you to make — counting assumptions.