How does drug testing work in sport? The full process

Six in the morning, a door in a quiet street, and a stranger on the step with a photo pass and a cool box. That is what drug testing in sport mostly looks like. Not a stadium, not a medal ceremony, not a laboratory with blue lighting. A tired athlete in a dressing gown being told they may not go back to bed alone.

The chemistry is the part that gets written about. Mass spectrometers, isotope ratios, the detection of things present at concentrations that stagger anyone who thinks about them properly. The chemistry is genuinely brilliant and it is almost never where the arguments happen.

Anti-doping is a legal system with a laboratory attached. Almost every controversy you have ever seen about a doping case turned on a question of law or procedure: what the athlete was responsible for, what standard of proof applied, whether a rule had been followed to the letter, how much fault a tribunal was prepared to find. The analysis said what was in the bottle. Everything after that was argument.

Drug testing in sport is a legal system with a laboratory attached

The whole apparatus sits on one document, the World Anti-Doping Code, and a set of International Standards that hang off it. The Code is not law in the ordinary sense. It is a contract of adhesion: sports bodies that sign it agree to write its provisions into their own rules, and athletes agree to those rules as a condition of entering competition. Nobody is compelled to sign. Everybody who wants to compete does.

That structure explains features of anti-doping that look strange from outside. There is no jury. There is no criminal standard of proof. There is no right to silence in the way a police interview would give you one, and a refusal to give a sample is itself an offence rather than an exercise of a right. None of that is an oversight. It is what happens when the rules are contractual rather than criminal, and the sanction is exclusion from a private competition rather than a prison sentence.

The Code defines a fixed list of anti-doping rule violations, and only those. Presence of a prohibited substance in a sample. Use or attempted use. Evading, refusing or failing to submit to collection. Whereabouts failures. Tampering. Possession. Trafficking. Administration. Complicity. Prohibited association with a banned coach or doctor. Acts intended to discourage or retaliate against someone reporting to the authorities.

Note what that list contains. Only two of the eleven require a laboratory to find anything.

• 11Anti-doping rule violations defined in the Code
• 8International Standards sitting under the Code
• 3Whereabouts failures in twelve months that make a violation
• 60Minutes in the daily slot an athlete must be findable

These are counts fixed by the World Anti-Doping Code and its International Standards rather than measurements or estimates. Provisions are revised between versions of the Code, so the authoritative figures are always the ones in the version in force at the time of the conduct.

Out-of-competition testing is the part that actually deters

Testing at a competition is the version everybody pictures, and it is the less important half of the system.

Think about what a competition test can catch. It is announced by the fact of the event. Every athlete knows the date months in advance. It samples one day at the end of a preparation that took months, and it looks for substances still circulating on that day. Anything used to build capacity in the winter, cleared from the body long before the season, is invisible to it.

That is not a hypothetical weakness. It is the entire logic of how endurance and strength doping is designed to work. The drugs that matter are mostly training drugs. Anabolic agents build tissue that persists after the substance is gone. Blood boosting raises oxygen carriage during the block of work that produces the fitness. The performance benefit outlives the detectability by a wide margin, and any programme that only tested on race day would be testing for the residue of a decision taken half a year earlier.

Out-of-competition testing exists to close that gap, and it has one defining feature: no advance notice. An athlete in a testing pool can be tested on any day, at any hour, at home, at a training camp, at a hotel, at an altitude base in another country. The value of the test lies almost entirely in its unpredictability. A test you can see coming is a test you can plan around.

There is a second distinction that follows from this, and it confuses people constantly. The Prohibited List separates substances banned at all times from substances banned in competition only. Stimulants, most narcotics and glucocorticoids by certain routes sit in the second group, because their effect is acute: they help on the day, and taking one in November does nothing for a race in July. Anabolic agents, hormone and metabolic modulators, and the methods of manipulating blood sit in the first group, because their effect persists.

So a substance found out of competition may be entirely permitted, and the same substance found in a sample given after a race is a violation. The rules are drawing a line around when a benefit can be obtained, not around whether a molecule is intrinsically wicked. Understanding this removes most of the confusion around cases involving asthma medication, recreational drugs and cold remedies.

The awkward consequence is that out-of-competition testing requires the authorities to know where athletes are, all the time. Which brings us to the most disliked rule in sport.

Whereabouts is the rule athletes hate and the system cannot do without

An athlete placed in a registered testing pool has to file, every quarter, a set of information about their own life in advance: home address, overnight accommodation for every night, regular training venues and times, competition schedule, and one 60 minute window per day during which they undertake to be at a specified place and available to be tested.

They can change the filing, and in practice they change it constantly, because life is not filed in advance. A late training change, a family emergency, a flight moved, a friend's spare room. Every one of those is an update the athlete has to remember to make, on an app, often late at night, in a foreign time zone, while also being an elite athlete.

Two things can go wrong, and the rules distinguish them.

A filing failure is a failure at the paperwork stage: nothing submitted for the quarter, or information so inaccurate or vague that a doping control officer could not find the athlete using it. Filing a hotel with no room number and no phone contact can be a filing failure even if the athlete was inside the building the whole time.

A missed test is a failure at the door: the officer attends during the declared 60 minute slot, makes the attempts the rules require, and the athlete is not there. The athlete carries the whole slot, not the start of it, so someone who declares seven to eight in the morning and leaves at half past has not complied by having been present earlier.

Any three failures of either kind in a rolling twelve month period is itself an anti-doping rule violation. No positive test needed. That construction is deliberate, because a system of unannounced testing with no penalty for being unfindable would be a system of announced testing with extra steps.

The privacy objection is serious and deserves stating at full strength rather than being waved through.

An athlete in a testing pool is required to disclose, to a database, where they sleep every night for the next three months. That includes information about people who never signed anything: partners, parents, the friend whose sofa is being used. The obligation runs through injury, holiday and the off-season. It applies during pregnancy. It applies at four in the morning if that is the slot filed, and it applies to athletes who have never given a positive sample in their lives, on the basis of a category they were placed in rather than any suspicion about them personally.

Nothing else in ordinary life resembles it. People subject to criminal justice supervision are monitored to that standard. Athletes accept it as a condition of entering races.

The counter-argument is not that the intrusion is small. It is that the alternative is worse for the athletes themselves. Without whereabouts, the only credible testing is at competition, and competition testing catches training drugs poorly. Clean athletes would then be competing against a programme that the system could not see, and the value of their own results would fall accordingly. The rule protects the people it inconveniences, which is a genuinely unusual position for a rule to be in and does not make it comfortable.

Two mitigations are worth knowing. Testing pools are tiered, and the full whereabouts burden applies to a relatively small group at the top rather than to everyone with a licence. And the assessment of each failure is individual: an officer's attempt has to have met the standard set out in the rules before a missed test can be recorded, and athletes can and do challenge whether it did.

What happens between the knock on the door and the sealed bottle

The collection procedure is written in obsessive detail, and every line of it exists because something once went wrong.

1. 1SelectionThe athlete is chosen from a testing pool, by a risk-based plan, at random, or on intelligence. Test distribution plans weigh the physiology of the sport, the phase of the season and any information the organisation holds.
2. 2NotificationA doping control officer or chaperone identifies themselves, shows accreditation, and tells the athlete of their selection and their rights and obligations. The athlete confirms in writing.
3. 3Direct observation beginsFrom that moment the athlete stays in sight of the officer. Delays can be granted for a medal ceremony, media duty, cool-down or medical treatment, but the observation does not stop while they happen.
4. 4The athlete chooses the equipmentSealed collection vessels and sealed kits are offered as a choice, and the athlete picks and inspects them. Choosing your own bottle is the point at which tampering allegations mostly die.
5. 5The sample is given under observationUrine is passed in direct view of a same-sex officer. It is the most invasive moment in the process and the one that no procedural refinement has managed to soften.
6. 6Volume and suitability are checkedA minimum volume is required. Where too little is produced, the partial sample is sealed and held while the athlete waits, then combined under the same observation. Concentration is checked against a minimum so the sample is analysable.
7. 7The athlete splits and sealsThe athlete divides the sample into the A and B bottles and seals both. The seals are tamper-evident and the codes on them are recorded on the form.
8. 8The paperwork is completedSubstances and supplements used in the preceding days are declared, along with any therapeutic use exemption. The athlete records any concern about the procedure on the form itself, which is the only place a complaint later carries weight.
9. 9AnonymisationThe bottles and the laboratory copy of the form carry a code number and nothing else. The laboratory analysing the sample does not know whose it is, which is why laboratory bias is a weak argument in almost every case.
10. 10Transport under chain of custodySamples travel in sealed containers with documentation recording every transfer of possession. Any break in that chain is the strongest procedural attack available to a defence.

The sequence set out in the International Standard for Testing and Investigations. Volumes, timings and device specifications are defined in the Standard and its technical documents, which are amended from time to time, so they are described here rather than quoted.

The blood version differs in the obvious ways and in one that matters. Blood is drawn by a trained collection officer, and the rules impose a rest period after exertion and a seated period before the draw, because the markers being measured move with posture and with recent effort. A blood test taken straight off the track would produce values that mean something different from the same athlete's values at rest, which would corrupt exactly the kind of longitudinal comparison the blood is being taken for.

The newest addition to the toolkit is the dried blood spot: a few drops from the upper arm or a fingertip, absorbed onto a device, dried, and shipped without refrigeration. WADA has issued and revised a technical document governing collection devices and the analyses that may be requested from a single device, with the current revision applying from the start of 2026. The attraction is logistical rather than scientific. A dried spot is cheap, small, stable in transit and far easier to collect in places a long way from a laboratory, which addresses a real weakness discussed at the end of this piece. It complements urine and venous blood rather than replacing them, because the volume available limits what can be looked for.

The A sample accuses, the B sample confirms, and neither proves intent

Every sample is split at the moment of collection. That single design decision does a great deal of work.

The A bottle is analysed first. If the screen finds something and the confirmation procedure stands it up, the laboratory reports an adverse analytical finding, which is the technical term for what everyone else calls a positive test. The athlete is notified and offered the analysis of the B sample.

The B analysis is the athlete's right, exercised at their request. They may attend, or send a representative, and the laboratory has to allow it. Requesting it is not an admission and declining it is not one either.

What the B sample establishes is narrower than the public imagines. It comes from the same urine or blood, given on the same day, by the same person. It cannot tell you whether the substance arrived through a syringe or a contaminated batch of protein powder. It cannot tell you when it was taken, in most cases, or how much. What it rules out is that the A result was a laboratory error, a swapped sample or an analytical artefact that would not reproduce.

That is a real question and it is worth the expense of answering. If the B does not confirm the A, the case is over. Not mitigated, not reduced. Gone.

There is a further layer that matters in disputed cases. Accredited laboratories are presumed to have conducted analysis in accordance with the International Standard for Laboratories. An athlete who wants to attack the analysis has to establish a departure from that Standard which could reasonably have caused the finding. If they manage it, the burden flips, and the anti-doping organisation then has to show that the departure did not cause the result. That is the shape of most technical defences: not "the laboratory is corrupt" but "this specific step was not performed as the Standard requires, and here is why that could produce this result".

Samples can also be stored and reanalysed later, within the limitation period the Code sets. That provision is more consequential than it sounds. It means a sample given today can be tested in several years' time against methods that do not yet exist, and it converts every stored sample into a standing threat to anyone who doped and got away with it on the day. Retrospective reanalysis has reallocated more medals than any single testing programme.

Strict liability is the principle that produces every hard case

Here is the rule that generates the arguments. Under the Code, it is the athlete's personal duty to ensure no prohibited substance enters their body. Presence in a sample is a violation, and neither intent, fault, negligence nor knowing use has to be shown for that violation to be established.

Read straight, that sounds monstrous. An athlete could be sabotaged, or given a contaminated supplement, or served a drink by somebody they trusted, and still have committed a violation.

Yes. And here is why the rule survives every attempt to soften it.

Consider the alternative. If intent had to be proved to establish a violation, every case would turn on the athlete's account of their own state of mind, and there is no test for that. Anti-doping organisations would have to disprove a story invented after the fact, with no access to the athlete's kitchen, their supplier or their medical cabinet. The practical result would be that anybody with a plausible explanation and a competent lawyer would go unpunished, which is to say almost everybody.

Strict liability is a rule about the allocation of an unavoidable risk. Somebody has to bear the consequences of an unexplained substance. The Code puts that burden on the person who chose what to swallow, because they are the only person with any control over it.

The important thing, and the thing most reporting gets wrong, is that strict liability applies to the finding and not to the sanction. Intent, fault and circumstances all come back in at the second stage, and they can move the outcome from four years to nothing at all.

The Code's structure works roughly like this. A violation involving a substance not classed as specified carries a long default period, cut sharply if the athlete establishes the violation was not intentional. If the athlete shows no fault or negligence at all, the period is eliminated. If they show no significant fault or negligence, it can be reduced, subject to a floor.

Deliberate use established48months

Athlete shows the use was not intentional24months

No significant fault or negligence, reduced to the floor12months

No fault or negligence at all0months

Constructed illustration built on the structure of the World Anti-Doping Code, using months of ineligibility. These are not the outcome of any real case, and the exact periods available depend on the version of the Code in force, the classification of the substance and the tribunal's assessment of fault.

Show the numbers

Invented worked example: the same finding under four different conclusions

Item	Value
Deliberate use established	48months
Athlete shows the use was not intentional	24months
No significant fault or negligence, reduced to the floor	12months
No fault or negligence at all	0months

Three practical points about that ladder, all of which decide real cases.

Proving a negative is hard and expensive. To show the use was not intentional, an athlete usually has to identify the actual source. That means finding the sealed tub, having it analysed at their own cost, and establishing that the batch contained something not on the label. An athlete who cannot afford that analysis loses an argument they might have won, which is a real inequality inside a system that presents itself as identical for everyone.

Fault is assessed against what a careful athlete would have done. Checking a product against a supplement testing scheme, asking a team doctor, reading the label: doing those things and still being caught is a different case from doing none of them.

And the reduction is discretionary within a range. Two tribunals looking at similar facts can land in different places, which is why appeals to the Court of Arbitration for Sport so often argue about the length of a ban rather than whether there should be one.

Therapeutic use exemptions are permission, not a loophole

Athletes get ill. Some have chronic conditions that require exactly the substances the Prohibited List bans. The therapeutic use exemption is the mechanism that lets a diabetic take insulin and a person with a hormonal condition take the treatment for it without ending their career.

An application goes to a panel of independent physicians, and four conditions have to be satisfied. There must be a diagnosed condition supported by proper clinical evidence. The treatment must be the indicated one for that condition. It must be highly unlikely to produce any enhancement beyond a return to the athlete's normal state of health. And there must be no reasonable permitted alternative.

The fourth condition does the most work and is the least understood. A permitted alternative that is merely less convenient, or slightly less effective, defeats the application. The exemption is a last resort rather than a preference.

Exemptions are granted for a named substance, at a stated dose, by a stated route, for a stated period. They are not a general licence, and an athlete who holds one for one substance and tests positive for another is in exactly the same position as an athlete with no exemption at all. Applications can be made in advance or, where treatment was given in an emergency or the athlete was not in a testing pool at the time, retroactively.

The suspicion attached to exemptions is that they are a legal route to an advantage, obtained by athletes with good doctors and compliant physicians. That suspicion is not baseless, and the leaking of exemption records in the past made it a public argument rather than a private one.

Two responses are worth holding at once. The formal process is genuinely demanding and the fourth condition genuinely excludes most attempts at gaming it, because most banned substances do have permitted alternatives for most conditions. And access to the process is unequal, because assembling a successful application requires medical support that a fully funded national team can provide and an unattached athlete cannot. The exemption system is not corrupt. It is asymmetric, in the same direction as everything else here.

The biological passport tests the athlete rather than the sample

Direct detection has a structural problem. It can only find what it is looking for, in a window during which the substance or its metabolites are still present. Reduce the dose, spread it out, and the window narrows towards nothing.

The athlete biological passport attacks that problem from the other end. Instead of asking what is in this sample, it asks whether this athlete's own biology is behaving the way it should over time.

The haematological module tracks blood markers, principally haemoglobin concentration and the proportion of immature red cells, along with values derived from them. The steroidal module tracks the ratios between naturally occurring steroids in urine, which sit in a stable relationship in an individual and get disturbed by administered hormones.

The statistics are the clever part. Each athlete's own results build a personal profile, and an adaptive model produces an expected range for their next value based on their own history rather than a population average. That matters, because population ranges are wide enough to hide a great deal, while an individual's own variation is much narrower. A value that would be unremarkable in a population can be flatly impossible for a particular athlete.

When a profile goes outside the expected range, it is referred to a panel of independent experts, who see the profile without knowing whose it is. They can ask for the athlete's explanation, for medical records, and for further testing. Only if the panel agrees unanimously that doping is highly likely and that no known physiological or pathological condition explains the pattern can proceedings for use of a prohibited substance or method be brought.

The passport does something no direct test does: it makes microdosing visible. A dose small enough to clear before any officer could arrive still has to do something to the blood, or it would not be worth taking, and the passport is watching the effect rather than the cause. It also has a deterrent function that is difficult to measure and probably larger than its case count, because it turns a manageable risk into a permanent one.

Its limits should be stated plainly. It is slow, requiring a series of samples over time. It is expensive, because the samples that build a profile all cost money whether or not they show anything. Altitude exposure, illness, dehydration and blood loss all move the markers, and each needs excluding. And a passport case is an inference from a pattern rather than an identification of a substance, which makes it harder to explain and more contested at hearing than a straightforward finding.

Results management is where the case is actually decided

The moment a laboratory reports a finding, the case leaves chemistry and enters procedure.

1. 1Initial reviewBefore anything is said to the athlete, the organisation checks whether a therapeutic use exemption covers the finding and whether there was an apparent departure from the testing or laboratory standards that could have caused it.
2. 2Notification and provisional suspensionThe athlete is told of the finding and their rights. For substances not classed as specified, a provisional suspension is mandatory. For specified substances and contaminated product cases it is optional, and an athlete may accept one voluntarily.
3. 3The B sample is offeredThe athlete may have the B analysed and may attend. If the B does not confirm the A, the case ends there and any provisional suspension is lifted.
4. 4ExplanationThe athlete may give an account: a source, a prescription, a contaminated product, an argument on procedure. Anything said here shapes the fault analysis later.
5. 5ChargeIf the case proceeds, a formal assertion of a violation is issued, setting out the article said to be breached and the consequence sought.
6. 6HearingA first-instance panel independent of the organisation that brought the case decides whether the violation is established and, if so, what fault attaches. The organisation must satisfy the panel to comfortable satisfaction, a standard above the balance of probabilities and below the criminal standard.
7. 7Decision and consequencesThe panel sets the period of ineligibility, disqualifies results from the relevant period, and time already served under provisional suspension is credited against the ban.
8. 8AppealThe athlete, the anti-doping organisation, the international federation and WADA can all appeal, in most international cases to the Court of Arbitration for Sport. An appeal is a fresh hearing rather than a review, so a sanction can go up as well as down.
9. 9PublicationFinal decisions against athletes who are not minors are published. The Code's approach to disclosure where a case ends without a violation, and where the person is young, has been tightened in successive versions.

The results management sequence under the Code and the International Standard for Results Management. Deadlines, review requirements and appeal windows are set in the Standard and are revised between versions of the Code.

Two features of that sequence carry most of the practical weight.

The provisional suspension is imposed before any hearing, and for the most serious categories it is not discretionary. An athlete provisionally suspended in the spring misses the summer, and if they are eventually cleared the season does not come back. That is the cost of a system that would otherwise let a probable doper compete through the eighteen months a contested case can take. It falls hardest during a qualification window, where a missed period of competition can cost a place at a championship even after full exoneration, a mechanism worth understanding alongside how athletes actually qualify for an Olympic Games and how ranking points accumulate across a season.

The standard of proof is comfortable satisfaction, and it is not a fudge. It sits deliberately between the two standards most people know, and it is calibrated to the seriousness of the allegation: the Code says so explicitly, so a graver charge requires more. Where the burden shifts to the athlete, to establish how a substance entered their body or that there was no significant fault, the athlete's standard is the balance of probability. A defence therefore does not have to be proved to the hilt. It has to be shown to be more likely than not.

Contamination changed the sanction, not the finding

The most common genuine injustice in anti-doping is the athlete who took a legal product that contained something not on its label.

Supplement manufacturing is a poorly regulated business in most jurisdictions. Cross-contamination happens on shared production lines. Ingredients are bought from suppliers who bought them from other suppliers. Some products contain undeclared active compounds because that is what makes them work. Meat can carry residues of growth promoters in countries where those are used in agriculture. Some substances now get detected at concentrations where an environmental or incidental route is a genuine possibility rather than a lawyer's flourish.

Strict liability means every one of those is still a violation. The Code's answer is to move the argument to the sanction, where a contaminated product case can be reduced to something between a reprimand and a substantial ban depending on the fault the athlete is found to have shown.

The 2027 Code, agreed after a long consultation and taking effect at the start of that year, widens this. The definition of a contaminated source is broadened beyond products bought off a shelf to cover contaminated medication, contaminated food and drink, contamination through skin contact, and environmental contamination, which brings a set of real-world routes inside the provision that allows a reduction for no significant fault. The Code's long-standing warning that athletes take supplements and medications at their own risk has been amended so that it no longer sweeps in prescribed medication alongside supplements. Both changes point the same way: keeping strict liability for the finding while acknowledging that the world contains ways for a molecule to reach an athlete that no amount of care would prevent.

Athletes are not left without practical protection, but the protection is thin. Supplement batch testing schemes exist and are worth using, though they certify a tested batch rather than a brand. The safest position remains taking as little as possible, which is advice nobody in a performance environment enjoys hearing.

What drug testing in sport cannot do

An honest account has to end with the limits, because the system's public credibility keeps being damaged by claims it never made.

Detection windows are the first limit and the hardest. Every method has a period after use during which it can find something, and that period varies from months for some long-acting anabolic agents to a matter of hours for others. Peptide hormones are the difficult category: the body makes them too, so detection depends on distinguishing an administered version from a natural one or on catching an indirect signature, and the window can be very short indeed. A sophisticated programme is designed around those windows. This is precisely why unannounced testing and the passport exist, and why neither is a complete answer.

The second limit is that testing finds what it looks for. Analysis screens for known compounds, and a substance designed specifically to be absent from the reference libraries will pass a screen until somebody obtains it, characterises it and builds a method. Stored samples and retrospective analysis are the countermeasure, and they work backwards rather than in the present.

The third limit is money, and it is the one that most undermines the promise of a level playing field. Anti-doping is delivered by a patchwork of national organisations, international federations and independent testing bodies whose budgets differ by orders of magnitude. A test costs real money, an out-of-competition test in a remote location costs considerably more, and a passport requires many samples per athlete over years. The consequence is that the intensity of testing an athlete faces depends heavily on the sport they play and the country they compete for. Two athletes on the same start line may have been tested with wildly different frequency in the preceding year, and neither of them chose that.

The fourth limit is institutional. For most of the history of the sport, testing was run by the same federations whose commercial interests depended on their stars. The move to independent bodies, from international testing organisations to sport-specific integrity units, was an attempt to remove that conflict, and it was a real improvement. It was also a very recent one. Any judgement about performances from earlier eras has to account for who was doing the testing and what they had to lose, a pattern set out in detail in what cycling's doping era actually revealed about its own governance.

The last limit is one of category. Anti-doping regulates substances and methods, not advantage. Athletes gain from altitude tents, from equipment developed by companies with research budgets, from medical support and nutrition that most competitors cannot access. The line between a permitted advantage and a prohibited one is drawn by the Prohibited List, which admits a substance for consideration on the basis of performance enhancement potential, health risk and whether it is contrary to the spirit of sport. The last of those is a judgement rather than a measurement. That is the same argument, with different equipment, as the one about whether footwear technology should ever have been allowed, and neither argument has a scientific answer waiting to be found.

How to read a doping case without embarrassing yourself

The next time a case appears, four questions will get you further than any amount of speculation about the athlete's character. The rest of our athletics coverage is written on the same principle.

Which violation is alleged? Presence of a substance, use inferred from a passport, a whereabouts failure and a refusal to submit are four completely different things with different evidence and different defences. Reporting that says only "failed a drugs test" has usually not established which one it is.

Was the substance banned at all times or in competition only, and when was the sample taken? That single pairing resolves a large share of apparently baffling cases, and it is printed in the source documents.

What is being argued about: the finding or the fault? An athlete accepting the finding and contesting the length of the ban is making a different case from one saying the analysis is wrong, and the two get reported identically.

And where is the case in the process? A provisional suspension is not a conviction. A first-instance decision is not final while an appeal window is open, and an appeal can raise a sanction as easily as lower it. A great deal of commentary is written about cases that have not finished, and a great deal of it is wrong by the time they do.

None of that tells you whether an athlete cheated. It tells you what has actually been established, by whom, to what standard, and what remains open. Anyone who has read the decision knows how much of the public argument was conducted about a document nobody had opened.