In my country, whenever there's a protest, authorities and protesters always give different figures relative to the number of people who participated in the protest. The difference is sometimes significantly large (e.g. a few years ago, I saw the authorities saying 10,000 people, when the organizers were saying 100,000). To the best of my knowledge, no one ever disclose how they estimate these numbers. (My suspicion here is that no one uses a reliable method for this estimation, so they don't disclose it out of fear of being discredited). My question is: Are there any reliable statistical methods to estimate the number of people in a protest? I guess a major problem is the sometimes chaotic nature of a protest that can interfere with setting up a proper protocol. On the other hand, it would surprise me if no one ever tried to develop a reliable method for this purpose.

Full article content could not be extracted automatically. Read the original below.