WADA 2026 Report: The Testing Curve Cools and the System's Financial Fracture

The opening figure: 3.2%. That was the growth rate of WADA's testing volume from 2026 to 2026, after the prior cycle had reached 12.5%. In its 2026 annual report, the world anti-doping body recorded 297,965 samples collected in 2026, an increase of roughly 9,100 over the 288,865 reported for 2026. The headline speaks of expanded testing. The data table speaks of a curve that is cooling. I read reports like this the way I read the match log of a final: I look at the trajectory before the final whistle, rather than only the scoreboard at the end. A system run on data, whether it is the swimming world or the doping-control apparatus, leaves its trail at the inflection points that the ordinary eye skips over. And the inflection point of this report lies in this: growth is decelerating while the detection rate sits still. Some context is needed to read the report correctly. WADA coordinates global anti-doping rules and standards; it does not run any competition itself. It sets a common rulebook, the World Anti-Doping Code, and national organizations and sports federations are responsible for compliance. In swimming, this chain runs from WADA down to World Aquatics and then down to national anti-doping organizations. A young swimmer in a country with a tight budget answers to that national body, not to WADA directly. This means the quality of clean-sport assurance for a swimmer depends largely on the capacity of the machinery back home. The report is labeled 2026, but most of its testing data belongs to 2026 and 2026. A reader of numbers should note this: there is a lag between the document's time label and the data's time label. Failing to separate these two layers leads to attributing results from one cycle to another. In my own analysis work, this is the most common error: reading a table without checking which season it describes. I once worked with GPS data from 29 players at a Saigon club when the pandemic halted every competition. I learned something then: data does not rest when the stands are empty. And curves that seemed meaningless during the interruption became the baseline for the next season's decisions. That principle applies to reading the WADA report now. Every figure here, to be usable, must sit beside a multi-year baseline. A single metric, pulled out of its series, is only noise. Start with testing volume, since it is the figure in the headline. The 2026 to 2026 series is continuous, and that is the report's strongest methodological feature. Continuity allows us to read both the absolute level and the rate of change. Absolute level: 288,865 samples in 2026, 297,965 in 2026, a gap of about 9,100. Rate: from 2026 to 2026 it grew 12.5%; from 2026 to 2026 it grew 3.2%. Put simply: the system accelerated hard in the early phase, then eased off the gas. When a curve takes this shape in movement data, I usually ask about the saturation threshold. Decelerating growth may be because the base has grown large, or because resources are starting to run dry. The report does not separate these two hypotheses, but the financial section at the end of the document offers a notable hint that I will return to. The second metric, more important to me: the rate of samples with an adverse analytical finding, AAF, meaning an official positive test. Over the last three years, this figure ran 0.77%, 0.80%, and 0.78%. The swing is only about 0.03 percentage points. This is the point I consider the most robust finding of the whole report. A rate stable across three years, despite rising sample volume, admits two possibilities. First, the actual violation rate in the tested athlete population is relatively stable. Second, the testing system has expanded correctly, reaching a population with positivity density similar to the old group. The report lacks the data to distinguish these possibilities. I use only the phrase "is linked to" rather than a causal conclusion, because I need a concrete mechanism before saying one thing produces another. One thing should be clear: a stable detection rate does not mean everything is fine. It only means that what we measure has not changed. There may be a volume of violations outside the measurable range, and the system is not measuring it. The third metric is the section I most recommend reading closely: compliance audit results. The average number of non-conformities fell from 25.5 to 20 across audits. Reading this, the report presents it as a step forward. But the number of audits is only nine. Nine samples in statistics is a thin number. If you follow long-distance swimming, you know nine swims are not enough to judge form across a season; they are enough to sketch a trend, not to confirm a conclusion. I will not use these nine samples to congratulate anyone. And here is the contradiction buried deep in this section: while the total number of non-conformities fell, the group of non-conformities related to the sample-collection process rose 11%. This group accounted for 45% of all findings, and 23% of them were classified as critical. In other words, the system fixed some administrative defects but weakened at its core function: collecting samples correctly. To a data analyst, this is a familiar pattern: the headline metric improves through gains in secondary items, while the pivotal item worsens. Reading only the first row of the table loses the signal. Sample collection is the stage on which every downstream conclusion depends. A sample collected the wrong way is unusable no matter how accurate the lab is. So when errors in this stage rise, the whole chain behind it is affected. The fourth metric, and the most worrying: the financial position. Revenue fell $8.3 million short of plan. Of that, the United States is late on $3.8 million and Russia $3.9 million. This is not the story of a club short of a sponsor; it is the story of two countries with important political standing inside the system simultaneously lagging on their obligations. The report still shows a net surplus of $1.2 million, suggesting the system cut spending or deferred some items. But a net surplus alongside a large revenue shortfall is a fragile structure: it holds balance through tightening, not through growing income. A contract is not a signature; it is a hypothesis that has been signed. And a deferred contribution obligation is also a hypothesis about the parties' priorities, one being tested right now. The fifth metric, the strategic shift: WADA describes a move from traditional drug testing toward investigations and prevention. The accompanying figures: more than 250 investigation operations, 88 laboratories dismantled, more than 90 tonnes of product seized. This is a direction that favors investigative quality over sample quantity, and strategically it makes sense under tightened resources. But it should be read alongside the 3.2% growth figure above: if testing volume slows just as the system pivots to investigations, the average number of samples per sport may fall even while the total budget holds. Finally, the date to remember: the 2027 World Anti-Doping Code was adopted in Busan, South Korea, and takes effect on January 1, 2027. Ahead of that, WADA and the Athlete Council ran a consultation process with more than 600 athletes and commissions from more than 60 countries and more than 70 sports. This is a milestone with a clear date, meaning every federation can plan for it. A counterintuitive angle. What is easily missed in this report is the window between now and 2027. When a new code has been adopted but not yet enforced, there exists a transition zone I call the variance window. In that zone, old behavior persists, because the compliance cost of the new code is not yet mandatory. For a system dependent on national organizations, this window may run longer where resources are thin. A country with a lean anti-doping apparatus cannot shift to the new standard within a single year, however much it may want to. There is a notable correlation among three data series: decelerating sample-volume growth, rising non-conformities in sample collection, and delayed funds from major contributors. It is tempting to read these three co-moving series as a linear causal chain: less money produces weaker testing quality. But to say that, I need to point to a concrete mechanism linking a funding gap to a specific collection error in a specific country. The report does not provide that level of detail. What I can state more precisely: three series move in the same direction, and that direction is unfavorable to the goal of guaranteeing a level playing field. I also want to be clear about the limits of the picture. Part of the variance in this data cannot be explained by numbers. Countries with political motives for late payment leave no quantitative trace in the report. Small national organizations that lose connection to the global testing system leave no clear figure. When the stands fall silent, home advantage dissolves into a number near zero; and when money slows, what is lost is not a number but a portion of trust that quietly slips out of the system. I write this to remind myself that in every model there is an unreadable zone, and a good data analyst is one who marks that zone rather than filling it with conjecture. One further point deserves attention: the stable AAF rate of about 0.78% may be misread as a sign of "no problem." But a stable detection rate does not measure how clean a sport is; it measures how well the detection system performs. These are different quantities. If detection capacity rises, the rate might fall while nothing is actually better. Conversely, if a system weakens, the rate may hold steady while more violations slip through. This is why I never conclude anything about a system from a single metric. To close with a signal for the next cycle. Every shock has its own probability. We call it a shock only when we have not yet checked the table. In the WADA case, the surprise does not lie in the figure of 297,965 but in this: the system is slowing just when it needs to accelerate, holding its detection rate nearly motionless, and has yet to resolve the financial fracture at its largest contributors. The window from now to January 1, 2027, will be the most watchable stretch, not because of any single case, but because of a new code that has been signed but that no one is yet obliged to live with. Based on my experience tracking sports data cycles, the best indicator to read the next cycle is not total samples but the non-conformity rate in sample collection. If that 45% figure continues rising in next year's report, it will signal that resources are being pulled out of the core function, regardless of whether total sample volume grows. The question for the next cycle: if next year's report shows sample volume stalling or falling, which indicator will the system cut first? That is the real test that will be legible.