How does the press estimate crowd size at rallies?

KINIGUIDE Crowd sizes at political rallies have become an important yardstick of the success of the organisers and to assess support from the public, which is why there has always been contention over crowd size estimates.

This can be due to many factors, ranging from human error to vested interests.

For example, organisers of a rally would have a reason to estimate their crowd size in the higher range.

As such, the police will usually provide one figure, while the organisers will provide another.

To make matters more confusing, the press also tends to cite different numbers of crowd size for any particular rally.

A good example of the disparity in crowd size estimates was over the Bersih 4 rally, which took place in August 2015.

The organisers had claimed that at its peak, there were 500,000 people in attendance at the Bersih 4 rally, while the police gave a much more conservative estimate of 50,000.

Meanwhile, various media outlets provided different estimates of the crowd size, ranging from 22,000 to 100,000.

With such significant discrepancies in estimates provided, it is no wonder that people have questioned the best way to determine crowd size at a rally.

This question was brought up again after the first major rally of the year was held last Saturday, in support of amendments to the Syariah Court (Criminal Jurisdiction) Act (Act 355) tabled by PAS president Abdul Hadi Awang.

So what is the best way to determine crowd size at a rally?

The Jacobs method, invented by University of California Berkeley journalism professor Herbert Jacobs in the late 1960s, is the most widely accepted technique to estimate crowd sizes.

Arizona State University data journalism professor Steve Doig said in an MSNBC article that Jacobs came up with this method when he was trying to estimate the size of the Vietnam War protests outside his office window.

He realised the ground outside was gridded so he could count the number of people in one square of the grid and then multiply that by the number of squares.

While this method would work best in a space that is naturally laid out in a grid-like pattern, it would also work in other settings as long as there is an estimate of the total area size and crowd density.

How does this method work?

This approach estimates crowd size by measuring how many people fit into a space and how densely packed they are into that space.

The best way to utilise this method is by using an aerial photo with a straight-down view, in order to better gauge the size of the area occupied by people as well as the crowd density.

"After gathering data on numerous demonstrations, Jacobs came up with some rules of thumb that still are used today by those serious about crowd estimation," said Doig.

In a loose crowd, one person occupies roughly one square metre, about an arm's length from the body of the next person.

In a more tightly packed crowd, about 2.5 people can fit in one square metre.

At peak density, also known as the "mosh pit density", up to four people can fit in one square metre.

To come up with a more accurate crowd size estimate, you would need to measure the square feet in the total area occupied by the crowd and divide it by the crowd density.

How reliable is this method?

Though the Jacobs method is fairly straightforward, crowd size estimates can still vary, depending on calculation choices made while using this technique, according to American astrophysicist Sabrina Stierwalt in a column for the Scientific American.

Among the examples she cited, Stierwalt said crowd density can vary depending on those doing the calculations.

The boundaries of the area occupied by the crowd can also be difficult to determine, she said, especially for very large crowds where it can be hard to separate rally-goers and non-attendees.

Not only that, within a set boundary of area occupied by the crowd, crowd density can also vary from high to low.

Changes in terrain can also be deceiving in terms of determining crowd density from aerial photography.

How to counter the issues with this method?

Despite the issues with the method, Stierwalt said, they can be resolved by using technology like 

This can give estimates of crowd size usually within 10 to 20 percent accuracy, she said.

With all these information in mind, what then should be the estimated crowd size for the rally last Saturday?

By incorporating this aerial footage and Google Earth to work out the area where people were seated, ground reports and a very optimistic assumption that there were 2.5 people per sq-metre, that would work out to about 44,538 people on Padang Merbok alone when it was at capacity.

Fortunately, this retrospective estimate of crowd size was pretty close to the estimate of 55,000 produced by Malaysiakini journalists covering the event, who adopted the grid and "eyeballing" methods.- Mkini