Occam's Razor: When the Simplest Explanation Is Best and When It Isn't

In the 1997 film Contact, a scientist says she has traveled through a series of wormholes and back, and her questioners remind her of Occam’s razor: the easiest explanation tends to be right, so she probably never left. By the end of the film, the audience learns her recorder caught about 18 hours of static.¹ The simple story was wrong, and the scene is a fair picture of how the razor gets misused.

Occam’s razor is the rule that, among explanations that fit the facts equally well, you should prefer the one that needs the fewest unsupported assumptions.² It tells you which explanation to check first, not which one is true: if sign-ups fall the morning after a website release, look at the tracking code before concluding that customers vanished. That order matters because in a series of experiments, psychologist Tania Lombrozo found that people already treat simpler explanations as more likely true and need disproportionately strong evidence before they favor a more complex one.³

What is Occam’s razor in simple terms?

Occam’s razor, as Encyclopaedia Britannica defines it, is a principle that gives precedence to simplicity: of two competing theories, the simpler explanation is to be preferred. Its oldest wording is about not positing more things than you need.⁴ The word razor refers to shaving away the assumptions an explanation does not require.¹

The careful version has two conditions that the popular version drops. The first is “other things being equal”. The Stanford Encyclopedia of Philosophy notes that philosophers today tend to read the razor as a principle of theory choice, and that the “other things being equal” form is compatible with simplicity counting for fairly little: closer to a tiebreaker than a trump card.² The second is what counts as simple. The same entry separates elegance, meaning fewer and shorter basic principles, from parsimony, meaning fewer kinds of things assumed to exist, and notes that the two typically pull in different directions.²

An assumption counts against an explanation only when nothing you already know supports it. “The server restarted overnight” is cheap if the logs show a restart, and expensive if you are guessing. The razor is one of the named rules in most lists of thinking tools collected as mental models, and it belongs beside the site’s other guides to making better decisions.

Before you compare explanations, write down what each one needs to be true, and cross off anything you already have evidence for. What is left is the cost the razor counts.

Where the razor came from, and what Ockham wrote

William of Ockham (c. 1287–1347) was an English Franciscan philosopher and theologian, and the principle is named after him, but the famous slogan that one “must not multiply entities beyond necessity” is nowhere in his texts.⁵ The sentiment is his. The words are a later paraphrase.

What Ockham did write, in his commentary on the Sentences of Peter Lombard, includes the line “Numquam ponenda est pluralitas sine necessitate”, or “plurality must never be posited without necessity”. The best-known wording was set down by the Irish Franciscan John Punch in 1639, and the philosopher Libert Froidmont claimed in a 1649 book to have coined the Latin nickname that became “Occam’s razor”.⁶ An AAAS blog post on the principle’s history says Ockham gets the credit largely because of the 1852 writings of the Scottish philosopher Sir William Hamilton.¹

Ockham did not invent the idea either. Aristotle had argued that a demonstration from fewer premises is better, other things being equal, and Ptolemy wrote that it is a good principle to explain things by the simplest hypothesis possible.⁶ Britannica notes that the principle carries Ockham’s name because he used it so often and so sharply.⁴

1. 4th c. BCAristotle prefers proofs with fewer premises
2. c. 1287–1347Ockham uses the principle throughout his work
3. 1639John Punch writes the famous Latin wording
4. 1649Froidmont claims to have coined the Latin nickname
5. 1852William Hamilton’s writings tie the rule to Ockham

One detail of Ockham’s own version is worth keeping. According to the Stanford Encyclopedia’s entry on him, his fullest statement of the razor lists the conditions under which you may assume more: when reason, experience or, for him, scripture requires it. If one of those is met, the less economical claim is the better one.⁵ Ockham’s rule was never “the simpler, the truer”. It was “do not assume more without a reason”. So when someone invokes the razor to end a discussion, ask whether the extra assumption has evidence or experience behind it; if it does, it is not excess, and the razor does not cut it.

Is the simplest explanation usually the right one?

Not as a rule. Occam’s razor only applies when competing explanations account for the facts equally well, and the Stanford Encyclopedia quotes the biologist Kent Holsinger’s point that such ties are comparatively rare.² An explanation that is simpler but leaves a fact unexplained has not won a tie; it has lost on the evidence.

The everyday misreading, “the simplest answer is usually right”, turns a tiebreaker into a forecast. It also hides a judgment call, because simple depends on what you count. One explanation can assume fewer things and still be less likely, and the shorter story can feel more convincing just because it is easier to tell.

Myth

Occam's razor says the simplest explanation is usually the right one.

Fact

It says that among explanations that fit the facts equally well, you should prefer and test first the one with fewer unsupported assumptions.

That pull toward short stories has been measured. In psychology experiments, simplicity is often defined as the number of separate causes an explanation needs: one illness that explains two symptoms is simpler than two unrelated illnesses.³

The study

Moderate evidence

Tania Lombrozo, Cognitive Psychology, 2007

People preferred explanations with fewer causes and judged them more likely to be true. When simplicity and probability pointed in different directions, people gave the simpler explanation a head start and needed disproportionate evidence before favoring the complex one. When the probability information clearly backed the complex explanation, the preference could be overcome.³

A 2022 study by Thalia Vrantsidis and Lombrozo, with US adults recruited online, found that simplicity still nudged people’s estimates of how likely an explanation was even after they had been given the relevant probabilities.⁷ Both measured judgments in experimental tasks, not real decisions, so they show a tendency, not a measured cost.

Your mind already applies a strong version of the razor without asking. When a single tidy cause appeals to you, that is the moment to ask what evidence would favor the messier story, and whether you have looked for it.

Occam’s razor in practice: four worked examples

Used well, Occam’s razor is a way to order your checks: list the explanations that fit, count the unsupported assumptions each needs, and test the leanest one first while keeping the others on the list. The four examples below are illustrations, not case studies, and each shows a different way the rule earns its keep.

1. Both fit the facts: the razor only applies when each explanation accounts for what you have seen
2. Count the unsupported assumptions: each block is something that must be true but that you cannot yet show
3. Test the leaner one first: it is often the quicker and cheaper check
4. Keep the other on the list: drop it when the evidence rules it out, not because it is longer

A metric drops overnight

On Monday, your sign-up dashboard shows a sharp fall since Friday. One explanation: Friday’s website release broke the tracking code, so sign-ups happened but were not counted. Another: interest in your product collapsed over a weekend, for reasons nobody can name. The first needs one assumption you can check in minutes against the release notes and a second data source. The second needs a sudden change in many separate customers at once. Check the tracking first.

A bug that only appears on one laptop

A test passes on every machine except a colleague’s. You could suspect a fault deep in the programming language, or a difference in her setup, such as an older version of a library. A fault in the language would have to affect her machine alone, while setup differences are common and easy to compare. Compare her setup with yours before filing a report against the language.

A colleague goes quiet

You sent a proposal on Tuesday and heard nothing by Friday. The stories that come to mind first are often the expensive ones: she hated it and is quietly lobbying against you. That needs assumptions about her opinion, her motive and her secrecy, none of which you have seen. She is busy, or the message got buried, needs almost none. Hanlon’s razor, a separate rule of thumb, makes a related point about people: don’t assume malice where carelessness or a simple mistake would explain it.⁸ A short, neutral follow-up tests the cheap explanation at almost no cost.

Horses, not zebras

Medicine has a well-known version of this. In the late 1940s, Theodore Woodward of the University of Maryland told his interns that when they hear hoofbeats, they should not expect a zebra; the saying is a warning to consider common conditions before rare ones.⁹ Strictly, that is a rule about base ratesbase rate: How often something happens in general, across all similar cases, before you know anything special about your own. Judgments that ignore it, a mistake called base-rate neglect, tend to be too confident about the case in front of them.Full entry in the glossary, how common each cause is, rather than about counting assumptions. The two often agree, and the example shows why: the razor works best alongside a sense of what is common.

When does Occam’s razor not apply?

Occam’s razor stops helping when explanations do not fit the facts equally well, when the simple account needs a hidden assumption to survive, or when more than one cause is genuinely at work. Wikipedia’s article on the principle notes that appeals to simplicity were used against meteorites, ball lightning and continental drift, all of which turned out to be real.⁶

Continental drift: the simple story that lost for decades

Alfred Wegener proposed in 1912 that the continents move, and set out the idea in a 1915 book. It was rejected by many geologists, largely because no known mechanism could move continents.¹⁰ Fixed continents could look like the economical view. Yet to explain similar plants and animals on continents far apart, geological maps of that era drew huge land bridges across the Atlantic and Indian oceans, an extra assumption the “simple” view quietly needed.¹⁰

New evidence settled it. From the late 1940s, surveys of the ocean floor found long ridges where new crust forms and magnetic stripes that seafloor spreading explained, and by 1967 plate tectonics had become central to geophysics.¹⁰ Simplicity cannot settle a contest when each side carries an unexplained piece; measurement has to.

Hickam’s dictum: more than one cause at once

In medicine, the usual counterweight has a name. Hickam’s dictum, attributed to a possibly apocryphal physician named Hickam, perhaps John Bamber Hickam, holds that a patient’s symptoms can come from several diseases at once, and it is described as a counterargument to misapplying Occam’s razor.¹¹ A common version says a patient can have as many diseases as they please.

The same pattern shows up at work. A project can be late because the estimate was optimistic and a key person left and a supplier slipped. Looking for one root cause is a good start; insisting on one can leave two problems unfixed.

Two warning signs suggest the razor is being misused. The first is when the simple explanation only fits because some inconvenient fact is being set aside. The second is when being wrong in one direction costs far more than being wrong in the other, so a less likely explanation still deserves a check. In both cases, ask what each story predicts you would find, then go and look.

The bottom line

When explanations fit the facts equally well, test the one that assumes less first, and keep the messier ones on your list. Ockham himself allowed more assumptions whenever reason or experience required them. Your mind already favors tidy stories, so write down what each explanation needs to be true, then go and look for the evidence.