When I learned that the court had made a list of Madoff’s victims available, I immediately wanted to graph it. I first forwarded the PDF one of my best friends. Within a couple hours he had converted it to text, which he streamed through Yahoo Pipes to generate KML suitable for viewing with Google Maps. The pipes filtered out duplicates, did address discovery and geolocation. Unfortunately, Google Maps can’t display more than 80 markers at once, which is far less than the thousands of people whom Madoff ripped off. He eventually did pull off a single view visualization, but it ran like a dog.
So last night I took a different approach. I extracted out all of the zip codes from the list. (Ignoring all of the international investors robbed by Madoff.) Then I used http://geocoder.us to map each zip code to a latitude and longitude. Next I wrote a short script in NodeBox to draw a projection of the zip codes on a canvas. Each point is color coded to indicate the number of times that zip code occurs in the court document. I was helped a lot by code and ideas in Ben Fry’s Visualizing Data book, which I have access to via Safari Bookshelf. In particular, I stole his function for projecting out the latitude and longitude numbers.
Here are the resulting images. The perfectionist in me would love to tinker with this for hours.




Ugly American said
Take note.
That’s where the ‘educated’ people live.
Mark said
This Google Map has all 10,000 icons on it in an interactive format.
http://www.MadoffMap.com
Dataset of the Day: Madoff Clients Map | Off the Map - Official Blog of FortiusOne said
[...] Here are some other maps folks have made: The New York Times has made a US map as well as this site called Madoffmap.com and this one. [...]
USLaw.com said
Novel presentation. A highly interactive mashup of the Madoff client data is available at:
http://www.uslaw.com/madoff
“Neogéographie” des victimes de Madoff « Monde géonumérique said
[...] autre type de carte se trouve ici. L’auteur qui vient très évidemment de la galaxie Mac a extrait les zip codes de la liste, [...]
dimi said
Hi! Nice graph! One question, how did you turn up pdf to text programmatically? That’s the most interesting….
( for the graph…now you can know where the high-income lies? more taxation? hehehehe )