Murphy's Law — "anything that can go wrong will go wrong" — is a heuristic for risk and design. Treat it as a design constraint: assume that whatever can fail will fail, and build accordingly. The law does not say that everything will fail at once; it says that over time or across many instances, the things that can go wrong will show up. So plan for them.
In protecting and surviving, the model pushes you to identify failure modes before they occur. What can go wrong with this launch, this deal, this system? List the ways. Then add margin: redundancy, fail-safes, simpler paths, or explicit contingencies. The mistake is assuming the happy path. Murphy's Law says the happy path is one outcome among many; the others will happen sometimes. The discipline is to make the system robust to the ones that can kill you or the mission.
The origin is disputed — often attributed to Edward Murphy Jr. in the 1940s in the context of aerospace testing. The spirit is engineering and military: expect failure, design for it, and you reduce the chance that a single point of failure takes you out. It is not pessimism for its own sake; it is a rule of thumb that improves outcomes by forcing you to plan for the downside.
Section 2
How to See It
You see Murphy's Law in action when something that "could" go wrong does — and no one had a plan for it. The diagnostic: was this failure mode on the list? Had anyone asked "what if this breaks?" Systems and projects that assume the best case are vulnerable; those that explicitly plan for failure modes are applying the law. Look for post-mortems that say "we didn't think X could happen" — that is Murphy's Law in reverse, the failure to apply it.
Business
You're seeing Murphy's Law when a critical launch depends on one person, one vendor, or one system and that single point fails. The failure was possible; it became actual. The law says: if it can go wrong, it will — so don't rely on a single point. Redundancy, backup owners, and fallback options are the application.
Technology
You're seeing Murphy's Law when an outage is caused by a failure mode that was theoretically possible but "not supposed to happen" — a disk fills, a config is wrong, a dependency goes down. The design assumed the happy path. Murphy says: assume the disk will fill, the config will be wrong, the dependency will fail. Monitor, limit blast radius, and have a runbook.
Investing
You're seeing Murphy's Law when a thesis depends on one key assumption (e.g. "rates stay low" or "this partner delivers") and that assumption fails. The law says: if it can go wrong, it will — so stress-test the thesis. What if the key assumption breaks? Is the position still viable? Margin of safety is Murphy applied to valuation and leverage.
Markets
You're seeing Murphy's Law when a supply chain or process has no slack — one delay or shortage cascades. The design assumed everything would arrive on time. Murphy says: assume something will be late or missing. Buffer inventory, alternative suppliers, and explicit contingency plans are the application.
Section 3
How to Use It
Decision filter
"Before committing to a plan, system, or bet, ask: what can go wrong? For each plausible failure mode, ask: what happens if it does? Add margin, redundancy, or a contingency so that when something goes wrong — and it will — you are not wiped out. Assume the worst that is plausible; design for it."
As a founder
List failure modes for every critical path: launch, fundraise, key hire, major customer. What can go wrong? Single point of failure (one person, one customer, one channel)? Dependency risk? Timeline risk? For each, add margin or a backup. Murphy's Law does not mean paranoia; it means making the list and addressing the high-impact, plausible failures. The team that "didn't think it could happen" is the team that didn't apply the law.
As an investor
Stress-test theses. What can go wrong with this company, this market, this bet? If the key assumption fails, what is the downside? Margin of safety in valuation is Murphy applied: assume something will be worse than the base case. Position size and diversification are also applications — assume some bets will fail; don't let one failure wipe you out.
As a decision-maker
In any important decision, run a quick pre-mortem: "Assume we did this and it failed. Why did it fail?" The answers are your failure modes. Then ask: can we reduce the chance of those, or limit the damage if they happen? That is Murphy's Law as a process. Apply it to projects, partnerships, and systems — not to paralyse, but to harden.
Common misapplication: Treating Murphy's Law as a reason to do nothing. "Everything will go wrong anyway" is not the lesson. The lesson is: because things will go wrong, plan for them. The law is a design principle, not a justification for inaction.
Second misapplication: Planning for every conceivable failure. Some failure modes are so unlikely or so costly to guard against that the rational move is to accept them. Focus on the plausible, high-impact failures. Murphy says "what can go wrong will" — prioritise by probability and impact, then add margin where it matters most.
Grove's "only the paranoid survive" is Murphy's Law applied to strategy. Assume competitors will copy you, customers will leave, and technology will shift. Plan for those failure modes — stay paranoid about what can go wrong — and you are less likely to be surprised. Intel's focus on continuous improvement and competitive response was a form of designing for Murphy.
Buffett's margin of safety — never overpay, always assume something can go wrong — is Murphy's Law in investing. If a business can deteriorate, assume it might; buy at a price that leaves room for that. His rule "never lose money" is implemented by assuming things will go wrong and building a margin so that when they do, you still survive.
Section 6
Visual Explanation
Murphy's Law — Whatever can go wrong will go wrong. Identify failure modes, add margin and redundancy, and design so that when something fails you are not wiped out.
Section 7
Connected Models
Murphy's Law sits with models about risk, resilience, and design. The connections below either implement it (defense in depth, margin of safety, fail-safes), extend it (redundancy, normal accidents), or formalise the response (pre-mortem).
Reinforces
Defense in Depth
Defense in depth is multiple layers of protection so that if one fails, another holds. It is Murphy's Law applied to security and systems: assume one layer will fail, so add another. The reinforcement: both assume failure and design for it. Murphy says "it will go wrong"; defense in depth says "so have another layer."
Reinforces
Margin of Safety (Systems)
Margin of safety in systems is extra capacity, time, or resource so that when something goes wrong you have slack. Murphy says things will go wrong; margin of safety says leave room for that. The reinforcement: both push you to build in buffer rather than assume the best case.
Tension
[Fail-safes](/mental-models/fail-safes)
Fail-safes are designed to fail in a safe way. Murphy says what can go wrong will; fail-safes try to make the failure mode non-catastrophic. The tension: Murphy is a reason to add fail-safes, but fail-safes can themselves fail (Murphy again). So you need both: assume failure, and design failure modes to be contained.
Tension
Normal Accidents
Normal accident theory says that in complex, tightly coupled systems, some failures are inevitable — they are "normal." Murphy is the folk version: things will go wrong. The tension: normal accident theory suggests that in some systems you cannot fully Murphy-proof; you can only simplify or accept residual risk. Use Murphy to reduce failure modes; accept that in complex systems some risk remains.
Section 8
One Key Quote
"If anything can go wrong, it will."
— Edward A. Murphy Jr. (attributed)
The quote is the whole heuristic. Treat it as a design constraint: do not assume the happy path. List what can go wrong, and add margin or redundancy for the failures that would be fatal. The law is not cynicism; it is a rule that improves outcomes by forcing you to plan for the downside.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
Make the list. Before any critical launch, deal, or bet, list failure modes. What can go wrong? Single point of failure? Key person, customer, or vendor? Timeline slip? Market shift? The act of listing surfaces risks you might otherwise assume away. Murphy says they will happen sometimes; the list is your checklist for adding margin.
Add margin where it matters. You cannot harden everything. Focus on the failure modes that are both plausible and high-impact. For those, add redundancy, a backup, or a contingency. The rest you may accept. Murphy is a prioritisation tool: assume failure, then invest in preventing or surviving the failures that would kill you.
Single points of failure are Murphy magnets. If one person, one customer, one system, or one assumption can take you down, Murphy says it will eventually. Reduce single points of failure or add a backup. The team that "didn't think it could happen" usually had one.
Pre-mortem is Murphy as process. Run a pre-mortem: "We did this and it failed. Why?" The answers are your Murphy list. Then ask: can we reduce the chance of those failures or limit the damage? That is applying the law systematically.
Section 10
Test Yourself
Is this mental model at work here?
Scenario 1
A launch depends on one engineer who knows the system. The engineer leaves the week before go-live. The launch is delayed by months.
Scenario 2
A company assumes a key supplier will deliver on time. The supplier has a disruption; production stops. The company had no alternative source.
Scenario 3
A team lists 50 things that could go wrong with a project and tries to prevent all of them. The project is delayed by a year and still has risks.
Section 11
Summary & Further Reading
Summary: Murphy's Law — "anything that can go wrong will go wrong" — is a design heuristic for protecting and surviving. List failure modes, add margin and redundancy for the plausible and high-impact ones, and reduce single points of failure. Use it before critical launches, deals, and systems. Pair with defense in depth, margin of safety, fail-safes, redundancy, and pre-mortem analysis.
Perrow argues that in complex, tightly coupled systems, some failures are inevitable. Complements Murphy: use Murphy to reduce failure modes; accept that in some systems residual risk remains.
Gawande on using checklists to catch failure modes before they happen. Practical application of "what can go wrong" in high-stakes domains.
Leads-to
Redundancy
Redundancy is duplicate capacity so that if one unit fails, another takes over. It is a direct application of Murphy: assume the primary will fail, so have a backup. The lead: Murphy's Law generates the list of "what can go wrong"; redundancy is one design response for critical single points of failure.
Leads-to
Pre-Mortem Analysis
Pre-mortem is the exercise: assume the project failed; why did it fail? The answers are your Murphy failure modes. The lead: Murphy's Law says plan for failure; pre-mortem is the process for generating and prioritising those failure modes before they happen.
