I've got something!
This really doesn't have anything to do with videogames, but last summer I recruited some folks from Qt3 to see if people wanted to participate in an ongoing experiment for the company I work for, Unanimous A. I.
We have a platform that does something called swarm intelligence. It's sort of like a whole "Wisdom of Crowds" thing, only out platform is proving to be smarter than that. It's real time and involves working together as a group (as opposed to focus groups and surveys, which isolate you as an individual and then aggregate your data.)
We're going to try to predict the outcomes of baseball games every week. Tuesdays, we pick them during the afternoon, but to accommodate more people, on Thursdays we're going to run this test during evening or prime time hours, at least on the east coast.
The process is pretty simple. You take a quick survey where you pick the winners of tomorrow's baseball games. Then you join a group of 30-40 other users for about 20 minutes and pick the same games with them.
If your individual picks on the survey are the best of the 25-40 people participating, we send you a $25 Amazon Gift Card for being awesome.
What we don't get out of this: personal data. at all.
What we do get out of this: Scientific data that we can study. We learn how to frame questions, we learn which of the under-the-hood metrics we have that show confidence in a prediction are most valuable. We learn how to identify participants who are likely to be accurate, and those who likely to be inaccurate.
Plus, it's really fun.
I don't want to clutter this thread with too many links about this. We've created an online facebook group where we'll post links for upcoming groups so people can join. That's also the first place we'll post our results and track them across the summer to see how we do.
Easiest way to join us and have some fun, make science, and maybe earn some of our cash: https://www.facebook.com/groups/1469745843088161