Wednesday, September 23, 2009

Speaking of Rigor - Another DC Tea Party March Crowd Estimate

In my ongoing attempt to find crowd estimates for the Sept 12 DC march that appear to be based on something other than direct rectal derivation -- one found me.

Mainly because he has a beef with the others I've found as compared to his, which comes out to about 17% of the minimal justified estimate I'd seen so far. He puts the number at ~ 135,000. Still 2 to three times bigger than I expected. But troublingly far, far off other estimates that appear to have relied on the traffic cam video.

I have not had the time to wade through the whole thing yet, as it is rather extensive and detailed (as a serious analysis based on the limited data would have to be) ... because he does attempt to take into account varying densities in different parts of the crowd.

He has links to various pictures and even some video that I have not seen before.

It's hard to say, really, which is one of the reasons there is so much controversy. The other, of course, being that leftists have an interest in portraying as few people as possible and right-ists have an interest in portraying as many people as possible.

Nobody expected the number of people that showed up to show up, and most people came of their own volition as this was organized on the local level, for the most part. Most of these people wouldn't have thought of saying, "hey, how much for an overhead shot?" I wouldn't have, and I strongly considered going.

At any rate, I'm including this analysis in my list of analyses, because it meets the test of someone having put some thought into it. Actually, a lot of thought into it. It is a valid approach. A daunting one to tackle accurately, but a valid approach, and impressive in its ambition.

As always, remember my mantra. Models are not reality. Models are representations of belief about reality. Some models are better than others at representing. I know how models are created -- they are all rife with assumptions. All of them, even the Park Service model and the traffic cam model. And even the best model is limited by the quality of its input. Sadly, we really don't have high quality input to put in any of these models for this event. Which is why so many disparate methods have been tried to arrive at some sort of estimate by brokers with any honesty in them at all.

I'd still take 135,000 people as a very respectable number. My comfort zone of 1 million, I admit, was largely based on comparative analysis -- that if there were 1 million people at Obama's inauguration, and that if 4 "Pensylvania Avenue Loads" of people marched on the mall, then there were certainly on the order of a million people there.

Perhaps not. Perhaps hundreds of thousands is a better estimate -- and from everything I've heard is where organizers are putting the number. The only critique I have of deathbymedia's analysis would be that I'd have to believe that the crowd from the 912 march around the speaking area was 83% to 86.5% (depending on which inauguration estimate you believe) less dense than the inauguration crowd. I will allow that it was significantly less dense, but not THAT much less dense, and there's the complete lack of people around the reflection pool at the inauguration to add to that.

We do not have the benefit of satellite photos (I wonder if the FBI or CIA do? Well, what do we pay them for anyway??? ;-) ) so we can't see how many are on the rest of the mall or elsewhere. Another thing to consider is that this wasn't an event for which everyone needed to assemble at a particular time or they'd "miss it" -- at no one time could you ever take a snapshot and say, "yup, that's the sum total of all the people who came". I submit that this was a much more dynamic crowd with a significant amount of flux throughout the day in any given area. There was no single event to see or miss. The point was merely to "show up" that day.

In the end, I won't try to directly dispute his numbers, as mine (and anyone's) analysis ultimately rely on certain assumptions. I think the number was higher than his. Maybe not a million. But certainly a very significant crowd.

No comments: