PredictIt had a limit of $850 on the platform, and even it's in the process of being shut down by the CFTC. I see it as a chicken-and-egg problem because having actual stakes even when miniscule means that smarter decisions are made with allocation; but I feel a lot of users would grow bored or not even use the platform if they could only use play money. My personal favorite example was ran in Popular Science magazine[0] for a few years where users could buy/sell imaginary shares based on the % probability outcome of certain events/discoveries being made by a specified date.
I have tried to get people to use Manifold, but with no actual chance of a payout most people don't see any reason to use it. And most people don't really like the mental thought of "gambling with charity money".
yes, each market has an agreed upon resolver, and you just don't use the market if you don't like it and it never attracts liquidity. its more complicated than that, check it out.
Arbitrary predictions markets sound good in theory until you realize that, as the market grows, eventually you reach a point where there is an irresistible financial incentive for the market participants to intervene in the outcome of the event in order to claim their reward. The most egregious case is "will so-and-so be dead by such-and-such date", which is plainly just an assassination market in disguise.
There are defi prediction markets. As other commenter said big issue is that it leads to fixing. One of the markets was something like "number of tweets this celebrity will have" and eventually someone paid them to start deleting tweets.
