Correction, 30 September 2026. Technical wording rewritten in shorter sentences under the readability rule; the meaning is unchanged.
Correction, 1 October 2026. The outcome label now follows the study's value record, which counts this study as having changed a rule or a plan.
Correction, 19 October 2026. This text was reviewed after the study finished, in a catch-up of the earliest studies; the result, outcome and dates are unchanged.
Rules fixed before the result
The rules for each possible answer were written down and fixed before any result was seen.
Data and method
Each split session was replaced by an ordinary session of the same company, and a fall averaging 1.75 percentage points over the 128 events was planted in the companies of 25 random groups at each size of 47, 24, 12 and 6 companies, and once in all 94. Over 4,000 repetitions per group we counted how often the rank-based test and the test on the average difference each detected the fall, and compared the two rates group by group.
What remains uncertain
How concentrated a real effect would be cannot be seen in this study. Only one planted size, 1.75 percentage points, and the five-session window were tested, and the companies carrying the fall were chosen at random, whereas a real effect may follow company traits. The simulation draws ordinary sessions independently, so it does not treat companies that move together on the same dates.