How many people were at the big 9/12 “Tea Party” protest? [UPDATED]

The big 9/12 protest certainly attracted a ton of people. How many seems to be an open question. The Washington Post reported “tens of thousands,” so a conservative might reasonably expect there to have been at least a hundred thousand. The low end of estimates seems to be around 50,000–60,000, and the high end, in the London Daily Mail was two million. That’s a pretty big spread.

In the last story the Volgi read, it reported the police had declined to give any estimates (despite earlier reports they were saying 50K-60K), and the Park Police are out of the crowd-estimation business (if one recalls right, they got too much flack for saying the Million Man March was actually the Four-Hundred-Thousand-Man March).

So, let’s take a look at some pictures.

In this one, taken from the Capitol, you can see the crowd extends past the Grant Memorial (the equestrian statue in the middle ground).

In this one, also taken from the Capitol, but looking northwest, rather than due west down the Mall, you can see the Peace Monument with the William B. Bryant Annex of the U.S. Courthouse at Third and Constitution in the background, with the crowd extending up Pennsylvania Avenue.

Ok, now click over to the Flash app at the top of this USA Today article on estimating crowd sizes in anticipation of the record crowd at President Obama’s inauguration.

I’d guess that you’ve got at least a third of that 240K space filled up—so say that’s 80K, not counting whoever’s coming down Pennsylvania.

Now look at this picture:

That’s a solid line of people from Fourteenth & E all the way to the Capitol. Fourteenth Street is where the USA Today dot indicating the 1.2 million who attended LBJ’s inauguration. Of course, this isn’t exactly right, because Pennsylvania is running diagonally, so it’s actually a longer line, and because Pennsylvania is narrower than the Mall (1.25 x 0.2 mi vs. 1.18 x 0.5 mi, by my Google Earth measurements.)

So, if, for the sake of argument, the Mall to Fourteenth, at [1.18 x 0.5 =] .59 sq. mi., holds 1.2M people, and Penn. Ave. to Fourteenth is [1.25 x 0.2] = .25 sq. mi., then you’ve got something just over 500K people in that picture.

So add that to 80K, and you’re up around 600K.

Buuuuut, apparently there were people going from the White House to the Capitol “for three hours” according to the caption here.

Let’s assume the first person leaves at 0:00. At an average 3 mph, she’ll reach the Capitol in .42 hours, or 25 minutes. (I say she because 3 mph is the average walking speed for women.) So, let’s let her walk slowly and get there at 0:30. At that point, you’ve got your first 500K people, right? At 1:00, you’re up to a million, and by 3:00, you’re up to three million. Now, let’s say they were walking especially slowly and that there wasn’t a constantly solid 500K there. What do we bump it down to? A million five? Two mil?

Now, let’s go to the video.

It looks to your Volgi like people are milling around there for the first 16 seconds or so. It’s not until 21 seconds that you see the whole length of Pennsylvania covered (let’s say that’s our first 500K). People then keep the street filled until the tail of the crowd appears at 36 seconds.

So, if the 40-second video is an accurate time-lapse of the 210 minutes from 8:00-11:30, then the 21-36 second march portion actually represents the 78.75 minutes from 9:50:15 to 11:09:00—much less than three hours. Every second represents five and a quarter minutes. If our 500K/30 minutes is right, then there are 87,500 people leaving every second of the video. Which, by the end of the thirty-sixth second means that 1,812,500 people have left. We’re in territory close to the Daily Mail’s number.

Now your Volgi is smart enough to know that he’s not knowledgeable enough to conclude that his assumptions are right. (He didn’t get to be Œcumenical Volgi by overestimating his own abilities. 其言之不怍, 則為之也難!) So he ran it by the Czar, who came up with two interesting points. First, this photo:

[UPDATE: Turns out this is of a 1997 Promise Keepers’ rally. The Czar was right to note the missing metadata. Apologies.]

If we treat the USA Today chart as authoritative, this shows well in excess of LBJ’s 1.2 million people. Doing a quick-and-dirty estimate of the amount of ground the crowds around the Washington Monument are occupying gives another couple hundred thousand, putting the total around 1.5 million.

But, thus spake the Czar:

We, by the grace of God, Emperor and Autocrat of all the Russias, of Moscow, Kiev, Vladimir, Novgorod, Tsar of Kazan, Tsar of Astrakhan, Tsar of Poland, Tsar of Siberia, Tsar of Tauric Chersonesos, Tsar of Georgia, Lord of Pskov, and Grand Duke of Smolensk, Lithuania, Volhynia, Podolia, and Finland, and so forth, and so forth, and so forth, took that photograph and put it on our left, jewel-encrusted Fabergé monitor.

We went to Google maps and went to map view to zoom into the exact area of the mall shown in the picture. We then switched to satellite view, and took a screen capture of it.

We opened up AutoCAD on our right-hand monitor. We pasted in the satellite image. Using a polyline tool, we traced over the areas that people are standing in, including the crowds that spilled over onto the streets, around the curved pavilion behind the Castle, and even drew around the trees because no one was in the thickets. We came up with four distinct crowds: two at the Washington Monument, one between the Monument and the Mall (a small one), and of course the Mall all the way up to the Reflecting Pool.

We deleted the photo, leaving the polylines. Knowing from the photo that 7/8″ actual was equivalent to 500 feet, we scaled up the polylines to full size. AutoCAD computes areas of enclosed polylines, and the grand total of all four crowd shapes was 2,350,069 square feet.

Now. Here’s what happens. If you use the usual method of computing massed crowds (the one the journalists use), you divide the square footage by two… and you get 1,175,035. That’s where the 1.2 million number comes from.

But to be honest, that “divide by 2” trick is only used in extremely packed conditions…like how many guys can you stuff into a locker. A more reasonable number for a loose crowd—like at a concert in the park—is 4.5 sq. ft. per person. If that’s a better description for what you see in the photo, then there are only about 522,238 people in that photo. We don’t think this crowd is tightly packed, especially if it was walking and converging. We would say the crowd is tighter closest to the Capitol, and fading back to about 9 square feet per person. We boldly suggest there are between 600,000-750,000 people in this picture!*

We noticed that USA Today is dividing by 2.5 sq. ft./person We think that’s outrageously tight. That’s like sardines! We have ordered the execution of their statician. Using that figure, we trace an even larger area than they did, and we still wind up with 940,028 people. So not only do we conclude 2.5 sq. ft./person is wrong, we think they’re wrong in their computation of the Mall’s area!


*Caveat: the photo is cropped to the West, and more importantly, the Czar cannot see around buildings to know precisely how big the spillover is. (We have received no reports from the Okhrana on this point. The responsible parties have been shot.) Further, we do NOT know if that was the crowd at its height. Frankly, who knows what that photo was from? Unfortunately, the photo was run through Photoshop or something first, because the camera information and creation data is missing from the file. If it had been the original photo from the camera, we’d be able to determine the exact moment in time the picture was taken to know if that was really the full extent of the crowd.

We do not think the picture was edited to increase numbers: The Czar sees a little repetition of colors—usually the sign of doctoring by cloning one part of the picture to another—but those are small blotches in the center of the mall. That’s the least likely place to doctor a picture…usually you do that at the fringes or sides to increase the crowds furthest from the center. Conclusion: real photo. The shadow lengths suggest afternoon, as the Washington Monument shadow is long but pointing from the southwest. If it were evening, the shadow would point East, and if earlier, the shadow of the Monument would be much shorter. So we suspect the photo is legit, and from the later afternoon.

Still, even at our worst case of 522,000 souls, that’s way bigger than what was reported. And like the Czar said, this photo does not cover everything going on.

Ergo, your unofficial Gormogons’ crowd estimate is, oh, let’s say 750,000 or so. (Incidentally, recalculating the Volgi’s 1.8M above with the Czar‘s 4.5 sq.ft./person density gives a hair above an even million people in the video.) But, given the math of previous crowd estimates, our 750,000 people probably outnumber those at some events put at a million or more.

Just to show you how much of a shot in a dark all these crowd estimates are: the “official” figure of 1.8M for Barack Obama’s inauguration is, upon examination, quite sketchy. The Park Service (probably having MMM flashbacks [or MMMM, as ’Puter would say]) chose not to produce a number after “1.8M” got out there and simply adopted it as a record. Meanwhile, an Arizona State University professor looked at a satellite image and came up with 800K. Another analyst looked at the same photo and came up with almost twice that. The key issue, as the professor mentions in the second article is the crowd-density number.

Meanwhile, for comparison, here’s the Obama inauguration crowd (at 11:19 a.m.). Note the dense clumps of people in dark clothing. It was cold that day and there were video screens set up for people to watch. So they cluster around them. To the Volgi’s untutored eye, it looks like a roughly comparable—or perhaps even smaller—crowd. What do you think?