Murphy's law for engineers
Murphy’s law is memorable because it sounds like a verdict. In practice it is closer to a bias you can budget for. Failure modes you never write down still exist. Naming them does not summon them, and ignoring them does not neutralize them.
Anything that can go wrong will go wrong.
Merriam-Webster defines Murphy’s law in that form. Keep the line where it helps (design reviews, launch checklists, on-call handoffs) and keep it off the pedestal. It is not a theorem and not an excuse to stop measuring.
Defensive design
Assume people and machines will pick the wrong option when two paths exist. Validate at boundaries. Prefer safe failure (drop a feature, keep data) over clever failure. Make incorrect assembly hard, not merely undocumented. The Wikipedia article on Murphy’s law traces part of the story to design-for-mishap thinking around rocket-sled tests in the late 1940s, including a narrower precaution about multiple failure-capable paths. That engineering reading is more useful than the cosmic version, whether or not you buy every anecdote in the footnotes.
Testing
Spend curiosity before production does. Exercise backwards wiring, empty inputs, partial retries, permission mistakes, and clock skew. If your suite mostly walks the happy path, you have a narrative about quality, not evidence.
Reliability work
Treat operations like a product with users. The introduction to Google’s free book Site Reliability Engineering frames that posture (automation, monitoring, incident response as first-class engineering). Murphy’s law fits there as a prompt to list what can fail before you change production, not as a substitute for SLOs, error budgets, or postmortems that chase causes instead of villains. In those reviews, Hanlon’s razor is the useful default for intent until the evidence says otherwise.
Attribution (soft)
Origin stories disagree. Wikipedia’s Murphy’s law article collects competing recollections from people involved in early tests, later gaps in documentation, and research that may push some printed uses earlier than the famous Edwards / Muroc narrative. I am not picking a single winner here. I still say “Murphy’s law” in rooms where everyone knows the shorthand, and I still separate the slogan from tall tales when the lesson is how we build, not who said what first.