A medieval friar's gift to modern thinking
Around 1320, the Franciscan friar William of Ockham articulated a principle that would outlast empires: 'Entities should not be multiplied beyond necessity.' If two explanations account for the same evidence, prefer the simpler one. It's not that complex explanations are always wrong — it's that simpler ones are wrong less often. Occam's Razor doesn't prove anything. It's a filter for where to direct your limited attention and scepticism.
Simplicity as probability, not aesthetics
There's a mathematical basis for Occam's Razor that goes beyond philosophical preference. Every additional assumption in an explanation is another point of potential failure. If explanation A requires three assumptions and explanation B requires seven, explanation B has more than twice as many places where reality can diverge from the model. Bayesian probability formalises this: simpler hypotheses start with higher prior probability not because the universe prefers elegance, but because there are fewer ways for them to be wrong.
How it works in diagnostics and debugging
When your car won't start, a mechanic checks the battery before the engine computer. When a website goes down, the engineer checks if the server is running before reviewing application code. Medical training follows the same principle: 'When you hear hoofbeats, think horses, not zebras.' Dr. Theodore Woodward's aphorism captures Occam's Razor in clinical practice. The common, simple explanation (dead battery, server crash, common infection) is statistically more likely than the exotic one. Start there.
The razor's limits: when simple is too simple
Occam's Razor has a critical caveat: the explanations being compared must account for the same evidence equally well. If the simple explanation leaves anomalies unexplained, it's not simpler — it's incomplete. Darwin's theory of evolution is vastly more complex than 'God created everything as-is,' but it accounts for evidence (fossil record, DNA, geographical distribution) that the simpler claim cannot. Einstein's relativity is more complex than Newtonian mechanics, but it explains Mercury's orbit and GPS timing errors that Newton couldn't. The razor cuts away unnecessary complexity, not necessary complexity.
Applying the razor to decisions
Before accepting any explanation — for why a deal fell through, why a product failed, why a team is underperforming — list the assumptions each competing explanation requires. The one with fewer assumptions deserves investigation first. Not because it's guaranteed to be right, but because it's the most efficient use of your diagnostic effort. When your startup's growth stalls, check whether the product has become worse before theorising about macroeconomic headwinds. When a team member's performance drops, ask whether their workload changed before speculating about motivation or loyalty.