Occupy Data!
My friend Suresh and I are thinking of starting a new working group for Occupy Wall Street, a kind of data science quant group.
One of his ideas: creating a value-added model for police in New York, just to demonstrate how dumb whole idea is. How many arrests above expected did each cop perform? [Related: you will be arrested if you bring a sharpie to the Bank of America Shareholder’s meeting in Charlotte!]
It’s going to be tough considering the fact that the crime reports are not publicly available. I guess we’d have to do it using Question, Stop, and Frisk data somehow. Maybe it could actually be useful and could highlight the most dangerous places to walk in the city.
Or I was thinking we could create a value-added model for campaign contributions, something like trying to measure how much influence you actually buy with your money (beyond the expected). It answers the age-old question, which super-PACs give you the best bang for your buck?
Please tell me if you can think of other good modeling ideas! Feel free to include non-ironic modeling ideas.
mathbabe 
How about trying to figure out how much value Tim Tebow adds to the Jets?
(Or Billy Beane to the A’s, etc etc.)
LikeLike
Strange, I have become a green-legged purple bug.
LikeLike
A better VAM analogue for police would be: what happened to rates of various crimes in their precinct, compared to what overall trends would predict?
LikeLike
Great idea! How about if you hooked up with some of the folks trying to create the social economy? For ex, I’m sure that most worker coops can’t afford to do serious modeling work and could use the help. I bet a group like the Democracy Collaborative could find folks who need help — maybe even Cleveland’s Evergreen Cooperative group, which is doing cutting edge work on creating green worker coops in poor Cleveland neighborhoods near the University. They’re trying to figure out how to build an approach thats financially sustainable and could be replicated elsewhere.
Or if you’re in NYC, you could hook up with some of the OWS-friendly, cash-strapped groups that work in poor neighborhoods of color that probably desperately need modeling help. They might not be sure where to start, but then again helping them figure out where modeling makes sense to do would be a good first step — and a way of empowering them.
LikeLike
Cathy, I remeber you clamouring for open-source alternative sovereign credit ratings a few months ago. Did you take a hand in this? http://ftalphaville.ft.com/blog/2012/05/02/983041/monte-carlo-simulated-sovereign-credit/#comments
LikeLike
No, but I know Marc Joffe! Pretty cool!!
LikeLike