Only remotely related to your question: When I was studying poker, the common wisdom I read everywhere was to never play without real money being involved. The justification for that was that people behave completely differently when no real money is involved, and so skill in play-money games does not translate at all to real poker. Apparently even stakes as low as $0.02 (translating to maximum loss for an entire game being about $5) make a difference.
I think the regulator issues could be easily overcome with bitcoin and TOR. I would be actually surprised if there was no betting exchange or sports betting site there, and those typically allow you to also bet on some non-sports events (elections, etc.).
I would describe Metaculus as a "play-money" prediction market. Why don't you think it's a prediction market? Players/users are rewarded with points (e.g. 'play-money') for making good/better predictions. What's missing?
I'd imagine that all of the previous ones died because it's a 'bad' business. I'm not aware of any trying to require, or even offer, paid subscriptions, but beyond that the only other revenue available to them is ads or maybe donations (or outside funding). And moderating site content generally and adjudicating the resolution of predictions is real work and thus relatively expensive.
Tho, now that you've raised the question, I wonder if this wouldn't work pretty well here on this site. Maybe it could be added on to the existing question/answer feature.
One (very broad) problem – or so it seems to me (and I haven't read any of the relevant academic literature) – with prediction markets, or even just predictions, is that there aren't both clear (let alone obvious) and specific answers available for the most interesting questions. And realistically being able to adjudicate disputes seems to require trusting some kind of proxy, e.g. reported in The New York Times.
I would describe Metaculus as a "play-money" prediction market. Why don't you think it's a prediction market? Players/users are rewarded with points (e.g. 'play-money') for making good/better predictions. What's missing?
What's missing is the ability to sell my shares of "yes" or "no" to other users. A market requires having the ability to exchange commodities. I think it's probably better to describe Metaculus as a prediction aggregator with an incentive system.
Prediction markets that trade in play-money rather than real money seem like a great idea to me: they avoid all the legal issues of real prediction markets, they don't seem to be much less accurate, and it might be fun to participate in them. But I can't find any active ones online.
The only ones I can see are ideosphere.com and alphacast.cultivateforecasts.com, but there's almost no activity in either. Metaculus is very active but isn't a prediction market. I've also heard of a few like NewsFutures which have completely died. Are there any active ones that I'm not aware of? If not, why do all play-money prediction markets die?