Who must prove what? Burden of proof answers that question. The party making a claim bears the obligation to support it with evidence. The party defending the status quo or denying the claim does not have to prove a negative. In criminal law the state must prove guilt beyond reasonable doubt; the accused need not prove innocence. In civil litigation the plaintiff must show the preponderance of evidence. In science the claimant must supply reproducible evidence; sceptics need only point to gaps or counterexamples. Shifting the burden — demanding that others disprove your assertion — is a rhetorical move, not a logical one.
The default matters. If the default is "innocent until proven guilty," the cost of wrong conviction is weighted heavily and the burden sits on the accuser. If the default is "guilty until you prove otherwise," the system is stacked the other way. In organisational life the default is often implicit: "we keep doing this until someone proves it's broken" versus "we change nothing until someone proves the new way works." Whoever sets the default shapes where the burden lies and who must marshal evidence. The strategic question is always: what is the default here, and who benefits from it?
The burden scales with the claim. Extraordinary claims require extraordinary evidence. "The meeting is at 3pm" needs a calendar entry. "This drug cures cancer" needs randomised trials. "Markets are always efficient" needs more than a few decades of data. Calibrating burden to the stakes and the strangeness of the claim prevents both credulity and paralysing scepticism. The discipline is asking: what would it take to convince a reasonable sceptic? If the answer is "nothing would," the burden has been misplaced.
Section 2
How to See It
Burden of proof reveals itself in who is asked to produce evidence. When someone says "prove me wrong," they are usually shifting the burden onto you. When a policy is adopted because "no one proved it would fail," the burden was placed on opponents. When a scientific paper is rejected for lack of controls, the burden is correctly on the authors. Look for the pattern: who must bring data, and what standard must they meet? That allocation determines whose job is harder and who wins when evidence is ambiguous.
Business
You're seeing Burden of Proof when a vendor insists their solution works and asks you to prove it does not. The burden belongs on the vendor: they must demonstrate performance, security, and fit. Accepting "try it and see" or "disprove our claims" inverts the burden and puts the buyer at a disadvantage. Strong procurement and evaluation put the burden on the claimant — the seller — and define the standard of proof in advance.
Technology
You're seeing Burden of Proof when a team proposes a new architecture and says the current system must be proven inadequate before changing. The burden can sit on the proposers (show that the new design is better) or on the incumbents (show that the current design is sufficient). Where it sits determines whether the organisation tends toward change or inertia. Explicit burden rules prevent endless debates where neither side has to prove anything.
Investing
You're seeing Burden of Proof when a promoter claims a company will 10x and challenges sceptics to prove it will not. The burden should be on the promoter: what evidence supports the 10x case? Valuation, comparables, unit economics. Shifting the burden to the sceptic favours hype. Serious analysis places the burden on the bull case and asks what would have to be true — and what evidence exists for that.
Markets
You're seeing Burden of Proof when regulation assumes a product is safe until harm is shown, versus assuming it is risky until safety is demonstrated. Precautionary approaches place the burden on the innovator; laissez-faire approaches place it on the regulator or victim. Where the burden sits shapes which products reach the market and who bears the cost of uncertainty.
Section 3
How to Use It
Decision filter
"Before accepting or rejecting a claim, ask: who has the burden of proof, and have they met it? If the burden is on the claimant and they haven't supplied evidence, the default is disbelief. If someone is trying to put the burden on you to disprove them, name the move and reset the burden."
As a founder
Place the burden where it belongs. When you make a pitch — to investors, partners, or customers — you bear the burden. Bring evidence: traction, benchmarks, comparables. When you evaluate an internal proposal, decide explicitly whether the proposer must prove the change is worth it or the status quo must prove it is sufficient. Defaults favour one side; make the default conscious. When a counterparty says "prove we're wrong," don't accept it. Require them to prove they're right.
As an investor
The promoter bears the burden. Management must show why the thesis holds: growth drivers, margin durability, competitive moat. Your job is to assess whether they have met that burden, not to disprove them. When management shifts the burden — "short sellers haven't proven fraud" — treat it as a red flag. Ask: what evidence would convince a sceptic? If they have it, they should offer it. If they don't, the burden has not been met.
As a decision-maker
Calibrate burden to the claim. Routine claims need routine evidence. Transformational claims need strong evidence. When someone makes an extraordinary claim and demands you disprove it, the burden stays on them. In meetings, when consensus is "we need more data," ask: who needs to produce it, and to what standard? Assigning burden explicitly cuts short circular arguments and forces the right party to do the work.
Common misapplication: Assuming the burden is on the sceptic. "Prove me wrong" and "you can't prove it won't work" are burden-shifting. The person asserting the positive claim bears the burden. Accepting that shift lets weak claims survive because disproving every possible claim is impossible.
Second misapplication: Demanding impossible proof. Requiring the other side to prove a negative (e.g. "prove no harm") or to meet a standard no one could meet (e.g. "prove with 100% certainty") is a way to block action while appearing rational. The burden should be proportionate: strong evidence for strong claims, not infinite evidence for any claim.
Section 4
The Mechanism
Section 5
Founders & Leaders in Action
Charlie MungerVice Chairman, Berkshire Hathaway, 1978–2023
Munger insisted that the person making an investment case bears the burden. "I don't want to hear why it might work. I want to hear why it will work and what could go wrong." He placed the burden on the promoter to show evidence and to steelman the bear case. When someone said "you can't prove the market is wrong," Munger's response was that the burden was not on him — the bull had to prove the market was wrong. That discipline kept Berkshire from chasing stories that had not met their burden of proof.
Franklin applied burden of proof to claims about nature and society. In his view the claimant had to supply evidence; scepticism was the default until proof was offered. He distinguished between "moral certainty" (enough for practical action) and "demonstrative certainty" (proof in the strict sense). That calibration — burden on the claimant, standard proportionate to the claim — shaped both his scientific work and his approach to public debate.
Section 6
Visual Explanation
Burden of Proof — The claimant must supply evidence; the default favours the status quo or the negative. Shifting the burden ('prove me wrong') inverts the logic and favours the unevidenced claim.
Section 7
Connected Models
Burden of proof sits at the intersection of logic, law, and decision-making. The models below either define related standards (presumption of innocence, reasonable doubt), extend the discipline of evidence (falsification, steelmanning), or describe failures when burden is misapplied (confirmation bias, Hanlon's razor).
Reinforces
Presumption of Innocence
The accused is presumed innocent; the state bears the burden of proving guilt. Burden of proof is the general principle; presumption of innocence is its application in criminal law. The allocation — burden on the accuser — protects the individual from the power of the state.
Reinforces
Reasonable Doubt
The standard the prosecution must meet in criminal cases. Burden says who must prove; reasonable doubt says how much proof is required. Together they define the criminal trial: the state has the burden and must meet the highest standard.
Reinforces
Principle of Falsification
Scientific claims must be falsifiable; the burden is on the claimant to offer testable evidence. Falsification specifies how the burden is met in science: not by piling up confirmations but by surviving attempted disconfirmation.
Leads-to
Steelmanning
To test a claim, steelman it — give it the strongest form — then see if it meets the burden. Steelmanning is a way to apply burden of proof fairly: don't refute the weak version; demand the strong version and see if the claimant can support it.
The standard of proof scales with the claim. Routine claims need routine evidence. Claims that contradict established knowledge or imply vast consequences need proportionately stronger evidence. Sagan's formulation blocks the move where someone makes a wild claim and then demands that sceptics disprove it. The burden stays on the claimant, and the bar rises with the strangeness of the claim.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
In any disagreement, the first question is: who has the burden? If you're making the claim, you supply the evidence. If the other side is making the claim, they supply it. "Prove me wrong" is not a valid move — it shifts the burden to the sceptic, who cannot prove a negative. Name it and refuse.
Defaults determine who wins when evidence is thin. When the default is "no change," the burden is on the proposer. When the default is "change," the burden is on the defender of the status quo. Many organisations never make this explicit, so the same burden allocation that favours inertia is treated as neutral.
Calibrate the standard to the claim. "The meeting is at 3" needs a calendar. "This strategy will double revenue" needs a model and comparables. "This asset is mispriced" needs a clear thesis and catalysts. Asking for the same level of proof for every claim is either credulous or paralysing. Match the standard to the stakes and the strangeness of the claim.
In negotiations and procurement, keep the burden on the claimant. Vendors prove their solution works. Promoters prove their deal is fair. Don't accept "you haven't shown it's bad" as a reason to proceed. The burden to show good sits on the side asking for the commitment.
Section 10
Test Yourself
Is this mental model at work here?
Scenario 1
A founder says their product will capture 50% market share and challenges the board to prove it won't.
Scenario 2
A regulator requires a new chemical to be proven safe before sale; the manufacturer objects that harm has not been proven.
Popper's treatment of falsifiability and the burden of evidence in science. The scientist makes a claim; the burden is to offer testable, falsifiable hypotheses. The sceptic need not prove the claim false — the claimant must offer a claim that could be shown false.
Foundational treatment of burden of proof in law. Wigmore distinguished burden of producing evidence from burden of persuasion and analysed how allocation shapes outcomes.
Sagan's "extraordinary claims require extraordinary evidence" in context. Accessible application of burden-of-proof thinking to pseudoscience and public debate.
Tension
Confirmation Bias
We seek evidence that confirms our view and avoid evidence that disconfirms. That tendency shifts the effective burden: we require less proof for what we already believe and more for what we doubt. Explicit burden discipline counteracts confirmation bias.
Tension
Hanlon's Razor
"Never attribute to malice what can be explained by stupidity." Hanlon's razor suggests we should not place the burden on the "malice" hypothesis without evidence. The principle is consistent with burden of proof: the person claiming malice bears the burden.
