Sinulog crowd estimates

In the last few years, estimates of the crowd that gathered for the solemn procession and the Sinulog Grand Parade have been hyperbolic, to say the least.

In 2010, then Police Regional Office (PRO) 7 director Lani-o Nerez came up with eight million. The number included those who lined up along Cebu City’s streets to watch the parade and those who turned up for the early mass at the Basilica del Sto. Niño.

A very cynical “really” was my initial reaction to the figure.

Because, let’s put it this way, the whole population of Cebu Province, and that includes the cities of Cebu, Mandaue and Lapu-Lapu, was a mere 4.6 million people, and that was even based on the 2015 census. The fact that almost double Cebu’s total population would converge in a limited area in Cebu City for that one day was almost too hard to swallow.

Apparently, I was not the only one gagging because in the succeeding years, the estimates dropped. Albeit the figures were no longer jaw-dropping and ridiculously improbable but these still raised some eyebrows, mostly mine.

Last year, the PRO 7’s estimate was a modest 3.2 million persons. As of 6 p.m. The figure was a big jump from the estimated 2.5 million revelers who joined the festivities in 2014.

Then PRO 7 director Manuel Gaerlan said “that the estimate was based on observations of nine sector commanders and the assumption of four persons for every square meter of the parade route and surrounding streets.”

But researchers Paul Yip at the University of Hong Kong and Ray Watson at the Melbourne University said the “four per square meter” principle is unrealistic, and described it as “mosh-pit density.” The two men wrote a paper on crowd estimation techniques that was published in the journal Significance.

Rob Goodier, in his article “The Curious Science of Counting a Crowd” in the publication Popular Mechanics, said that “crowd-size estimation is a murky science, positioned at the intersection of statistical precision and political sleight-of-hand, and plenty of people are motivated to either exaggerate or low-ball an event’s attendance.”

With that said, Sinulog organizers might be interested in what the Digital Design and Imaging Service plans to do. The firm wants to crowdsource head-counting aerial photos to Amazon’s Mechanical Turk, a network of people around the world who do tasks online for free.

These people, who shouldn’t have any vested interest, will do the actual headcount. Obviously, they’ll come up with different numbers, but Sinulog organizers can always throw out the outliers and average the rest.

If they do this, organizers might come up with a figure of yesterday’s and today’s crowd that is not the result of bias, political or otherwise.