Reported Protest Numbers Disputed Across Media Outlets

We began by tracing where each figure came from. The 120,000 figure originates from the organisers' own press statement. The 4,000 figure comes from an unnamed police source quoted by a single outlet. Neither number was independently measured.

The square has a publicly measured area of about 22,000 square metres. Crowd scientists commonly estimate dense standing crowds at two to four people per square metre. That gives a theoretical capacity of roughly 44,000 to 88,000 for the square alone, which already makes 120,000 hard to reconcile without the surrounding streets.

We reviewed aerial footage supplied by two broadcasters and a drone operator. The crowd filled the square and parts of two adjoining avenues, but the density was uneven, with large empty areas near the edges. Footage taken at the peak moment suggests a figure well above 4,000.

We asked an independent crowd-counting analyst to review the footage. Using grid sampling, the analyst estimated a peak attendance of between 25,000 and 40,000, with a wide margin of error because of overlapping tree cover and lighting conditions.

Both extreme figures have an obvious incentive. Organisers benefit from larger numbers because they signal public support. Authorities may benefit from smaller ones because they suggest limited public concern. Neither source has a record of providing verified counts for past events.

Several outlets repeated one figure without attribution, and some headlines presented a disputed estimate as settled fact. A fair account would give the range and name the source of each number.