Rally-Tally

Be Counted

Rally-Tally

By Rally Tally Editorial Team

Editorial Team

Last updated: 21 May 2026

Rally Tally publishes practical guidance on privacy-first attendance tracking, crowd-counting methods, protest mapping, and turnout reporting for organisers, journalists, activists, and researchers.

Why crowd counts are disputed

Crowd estimates are often treated as facts before anyone explains how the number was produced. Organisers may report one figure, public authorities another, and media outlets may repeat whichever estimate reaches them first.

That happens because open public events are difficult to measure well. People move. Entry points are uncontrolled. Routes change. Groups arrive late, leave early, or split across multiple locations. A rough visual impression can be badly wrong, especially once a crowd stretches along streets, parks, or squares.

If you want a more credible crowd estimate, the first step is to stop treating crowd size as a guess and start treating it as a measurement problem.

Real examples of why method matters

This problem shows up repeatedly in public life:

  • On 3 August 2025, the Sydney Harbour Bridge "March for Humanity" produced estimates ranging from roughly 90,000 by police to around 300,000 by organisers, with independent expert analysis published by The Guardian placing the crowd in the 225,000 to 300,000 range.
  • On 16 June 2019, a Hong Kong anti-extradition protest produced another major gap: organisers said nearly 2 million attended, while police said 338,000 were counted on the route at the peak.

These examples are useful because they show that crowd-size disputes are usually method disputes. Once the method changes, the number changes too.

The difference between a guess and a method

A useful crowd count is not just a number. It is a number that comes from a method someone can explain.

At a minimum, a stronger crowd estimate should answer questions like:

  • What area was measured?
  • At what time was it measured?
  • Was the estimate a peak snapshot, a running total, or a final attendance count?
  • How dense was the crowd assumed to be?
  • Were there multiple locations, stages, routes, or waves of attendance?

Without that context, two people can look at the same event and produce wildly different estimates while both sounding confident.

Common ways people estimate crowd size

There is no single perfect method for every public event. The right approach depends on whether the event is static, moving, ticketed, unticketed, local, or spread across multiple places.

The most common approaches are:

  • Area-density estimation: estimate the physical area occupied by the crowd, then apply a density assumption such as sparse, moderate, or dense.
  • Entry or checkpoint counting: count people passing one or more points over time.
  • Section-by-section counting: divide the crowd into smaller zones and estimate each zone separately.
  • Attendance records: use check-ins, registrations, scans, or other direct participation signals.
  • Composite estimation: combine several methods to reduce the weaknesses of any one source.

For unticketed protests, vigils, rallies, marches, and public gatherings, the strongest approach is usually a combination rather than a single number pulled from a photo.

Area and density: the classic method

The most widely used public crowd-counting method is area multiplied by density.

In simple terms:

  • measure the space occupied by the crowd
  • divide it into zones if density varies
  • estimate how tightly packed each zone is
  • total the results

This is more useful than a broad visual guess, but it still depends on judgment. A crowd that looks "packed" from one angle may be loose from another. A plaza may have open gaps, trees, barriers, stages, media platforms, and road furniture that reduce usable standing space.

That is why the method works best when:

  • the mapped area is explicit
  • assumptions are stated clearly
  • different density zones are treated separately
  • estimates are revised as new observations arrive
Why peak crowd and total attendance are not the same

One of the biggest sources of confusion is the difference between peak crowd size and total attendance.

Peak crowd size means the number of people present at one moment.

Total attendance means the number of people who participated over the full duration of the event.

These can be very different. A march may never have 40,000 people in one place at the same time, but 40,000 people may still participate across several hours and locations. A rally may hit a sharp peak and then shrink quickly. A decentralised action may involve multiple city events whose totals need to be combined.

If a report does not say whether it is describing peak presence or cumulative participation, the figure is incomplete.

Why open public events are especially hard to count

Open events do not behave like stadium events or ticketed conferences.

They often include:

  • multiple entry points
  • fluid movement
  • partial attendance
  • route changes
  • overflow areas
  • simultaneous or linked events in other cities
  • participants who want to be counted without being identified

That is exactly why crowd counting at protests and public gatherings is contentious. The event format makes traditional counting methods weaker, while the public importance of the final number makes disputes more likely.

Better crowd counting starts with better event data

Even if you use area and density estimates, better event data makes the result more credible.

Useful inputs include:

  • a clearly named event
  • date and time
  • start and end locations
  • route information or waypoints
  • event type and category
  • visible changes in turnout over time
  • a count of direct participant check-ins where available

This does two things. First, it makes the estimate easier to explain. Second, it creates a stronger historical record for later reporting, comparison, and verification.

A simple way to explain your final number

If you publish a crowd estimate, include a short method note with it.

For example:

"We estimate peak attendance at approximately 18,000 to 22,000 people, based on zone-by-zone area and density estimates taken between 2:15pm and 2:45pm, supported by route observation and participant check-ins."

That one sentence does far more for credibility than a larger unsupported figure.

Why anonymous attendance tracking matters

For many public events, especially protests, attendance is politically or socially sensitive. People may want their presence counted without creating a personal identity trail.

That creates a real measurement challenge. If the only way to improve crowd counts is to collect names, accounts, or personal profiles, many participants will reasonably opt out.

A privacy-first attendance tracker changes that trade-off. It allows people to contribute to a stronger participation record without turning attendance into a surveillance system.

That is one reason Rally Tally focuses on anonymous event check-ins. For open public gatherings, a privacy-respecting direct participation signal can be more useful than another round of unsupported crowd guesses.

A practical method for counting crowd sizes more accurately

For organisers, journalists, researchers, or campaign teams, a practical approach looks like this:

1. Define the counting question

Decide whether you are trying to measure:

  • peak presence
  • total attendance over time
  • turnout across multiple linked events

You cannot choose the right method without choosing the right question first.

2. Map the event structure

Document:

  • event title
  • date
  • start and end times
  • start and end locations
  • key route points or zones

That gives you a framework for analysing where people are and when they are there.

3. Break the event into zones or stages

Do not treat the entire crowd as one block if density varies. A square, march route, and spillover street should not all receive the same density assumption.

4. Use direct attendance signals where possible

If participants can check in anonymously, that gives you a stronger factual participation signal than visual estimates alone.

It will not replace every other method, but it improves the evidence base.

5. Distinguish physical and remote support

For some events, virtual support matters too. That should be recorded separately rather than mixed into a single attendance figure.

6. State the method with the number

A good final report does not just say "20,000 attended." It explains how that figure was reached and what it represents.

Mistakes that make crowd estimates worse

Some common mistakes undermine credibility immediately:

  • using one photograph to estimate a whole event
  • ignoring time variation across the event
  • confusing route length with crowd density
  • counting total attendance as if it were peak presence
  • using a single density assumption for clearly mixed conditions
  • combining physical attendees and online supporters into one number
  • publishing a number with no method at all

If the methodology cannot be explained in plain language, the resulting figure is weak no matter how precise it looks.

What a stronger crowd record looks like

A stronger crowd record is transparent, contextual, and revisable.

It should ideally include:

  • what was counted
  • when it was counted
  • how it was counted
  • what assumptions were used
  • whether the figure is peak or cumulative
  • whether the count refers to one place or multiple linked events

For public events that matter historically, that level of clarity is far more useful than an impressive-sounding number with no method behind it.

How Rally Tally fits into this

Rally Tally is not a replacement for every crowd-counting method. It is a tool that improves one of the hardest parts of the problem: recording real participation at unticketed public events.

It helps organisers and participants:

  • list public events clearly
  • define event locations and timing
  • support anonymous physical check-ins
  • record virtual support separately
  • build a more defensible attendance record

If you are already trying to estimate crowd size using maps, routes, media observation, or on-the-ground reporting, a privacy-first attendance signal can make the final record stronger.

Start with a method, not just a number

If you want to count crowd sizes accurately, the goal is not false precision. The goal is a better method.

For protests, marches, vigils, festivals, memorials, and other public gatherings, the best results usually come from combining:

  • clear event structure
  • route or area awareness
  • time-based context
  • direct participant attendance signals
  • transparent reporting

That is the standard worth aiming for. Better crowd counts do not come from louder claims. They come from clearer methods and better evidence.

If you want a privacy-first way to support that process, explore Rally Tally’s rally attendance tracker and protest attendance tracker guides.

External reading and references