I’m delighted to say that after a hiatus of several years, I’ve returned to my flipwalks.

If you’re not familiar with my flipwalk project, you can click here for the whole “48 Hours From Ground Zero” gallery. The short version is that while living in New York City, I would leave my house and determine a random course for an hour by flipping a quarter. At the end of that hour, I took photos of the block I was on. I picked one representative photo and put it up on this webpage, along with information on where it was taken and and how I got there. (These are probably my three favorite pictures so far.)

I originally intended to do one hundred flipwalks, but moved from New York to California after forty-eight (leaving with a healthy backlog of unposted walks and photos, which I’m now trying to work my way through). I discovered after I began that my project is considered by academics to be part of the psychogeography field.

That’s the thumbnail for the walk; you can see the larger version along with the full report here.

I received an interesting email from my good friend Robert Rossney earlier this week on the math of the project, taking issue with my description of the odds of ending up twice in the same place (walk #21 and walk #1) as “infinitesimally small”:

I don’t know if anyone else pointed this out to you, but the odds of two random-walk traversals of a simple graph terminating at the same node aren’t infinitesmally small. It’s analogous to the birthday paradox: there are 365 days in the year, but get 17 random people in a room and the odds are quite small that there won’t be two among them with the same birthday.

I ran a little Monte Carlo simulation (assuming perfect 4-way-intersection grid and 50 decision points per walk) and in 70-80% of trials I had more than one intersection at which a walk terminated twice or more.

Now, the odds that the last walk would end where the first one did, that’s a little trickier. But I can tell you that when I run 5,000 random walks, between 1,600 and 1,700 end on one of the 100 intersections in the 10×10 square surrounding the origin.

So I’m guessing, not infinitesmally small.

posted 6 June 2008 in Photos, Self-reflexive and tagged Flipwalks. 2 comments