Minister Louis Farrakhan’s Million Man March filled the Mall 20 years ago. The National Park Service estimated the crowd at a paltry 400,000. Event organizers called the figure defamation and threatened to sue. Congress responded by forbidding NPS from further Mall crowd estimates.

A matter of perspective

Picture taken looking west from the Capitol

The way to make any crowd look much bigger is to shoot from a relatively low angle with a lens that puts a large area into a single image. The foreshortening effect makes spaces between the people disappear, so they seem to fill every inch.

Park Service aerial photograph looking east toward the Capitol.

Crowds

Open

Crowds

Open

Open

The Park Service was hampered by using relatively low shots from a helicopter rather than a satellite. And the helicopter was moving. And they flew off to the side, never directly overhead. So the pictures were low, angled and blurry.

Crowd counting science

Developed in the 1960s, a scientific method for crowd counting is to estimate the number of square feet occupied by the group, then divide by the average area used by a person. NPS used estimates of:

Approximately one person per square meter . . .

for a loose crowd

for a tight crowd

A little over two per square meter . . .

A little over four per square meter . . .

for a squeezed group like a mosh pit.

A new remote sensing analysis

Once the Million Man March controversy arose, the Boston University Remote Sensing lab headed by Farouk El-Baz stepped in for a recount. After testing with students in the lab, they concluded that the tightest group would be:

Six people per square meter . . .

each person would have about the space of a single newspaper page.

Boston University crowd analysis overlaid on a Park Service aerial

Using that methodology for the whole mall, Boston University estimated the crowd at 837,214 people with an error margin of plus-or-minus 20 percent. That means it was between 669,7771 and 1,004,657.

266,601

152,000

414,969

3,644

Total: 837,214 people

A matter of perspective

Picture taken looking west from the Capitol

The way to make any crowd look much bigger is to shoot from a relatively low angle with a lens that puts a large area into a single image. The foreshortening effect makes spaces between the people disappear, so they seem to fill every inch.

The Park Service was hampered by using relatively low shots from a helicopter rather than a satellite. And the helicopter was moving. And they flew off to the side, never directly overhead. So the pictures were low, angled and blurry.

Park Service aerial photograph looking east toward the Capitol.

Crowds

Open

Crowds

Open

Open

Crowd counting science

National Park Service used estimates of:

Developed in the 1960s, a scientific method for crowd counting is to estimate the number of square feet occupied by the group, then divide by the average area used by a person.

Approximately one person per square meter . . .

for a loose crowd

A little over two per square meter . . .

for a tight crowd

A little over four per square meter . . .

for a squeezed group like a mosh pit.

A new remote sensing analysis

Once the Million Man March controversy arose, the Boston University Remote Sensing lab headed by Farouk El-Baz stepped in for a recount. After testing with students in the lab, El-Baz concluded that the tightest group would be six people per square meter.

Boston University crowd analysis overlaid on a Park Service aerial

Using that methodology for the whole Mall, Boston University estimated the crowd at 837,214 people with an error margin of plus-or-minus 20 percent. That means it was between 669,7771 and 1,004,657.

152,000

266,601

CONSTITUTION AVE.

414,969

3,644

INDEPENDENCE AVE.

Total: 837,214 people

SOURCE: Sources: National Park Service, Boston University Center for Remote Sensing; images via AP and National Park Service. GRAPHIC: Dan Keating and Laris Karklis - The Washington Post. Published Oct. 9, 2015.