Thanks, Christoph, that’s very kind. I think equality of participation is implied by any blockchain due to censorship resistance, but that prediction markets are one of the most obvious cases where that would be useful. Not only are there serious legal concerns with attempting to run one from within a corporation (in fact, it is basically illegal in most of the developed world) but the entire premise of a well-functioning prediction market is that it becomes dramatically more useful the more people (and the more different kinds of people) are using it. There is vastly diminished social benefit in running one that isn’t open to practically everybody because you would inevitably be stopping prices from moving in the ‘right’ direction and biasing in favour of the preconceptions of whatever group you do allow to participate. You want anybody willing to participate to do so.

There is another, more subtle, point that individual markets run by a central body must be conceived of by the central body before they are organised, almost by definition. Therefore very unusual or bizarre events will be unlikely to be hosted, since the centralised organisation will have fixed costs and will need to make money on every market. But again, it is beneficial for there to be markets in anything and everything, because what this does is force those with sincere belief to make that belief known to the whole world. Prediction markets have enormous social utility as aggregators of perceived probabilities that otherwise remain widely distributed solely in people’s minds.

One interesting area I didn’t discuss in the post is that it isn’t really the entirety of the prediction market that would need to ‘run on a blockchain’, but only the payments, provided the token had a viable smart contract system. All that really matters to the integrity of the prediction market is that people who buy securities get paid appropriately. That the UI is buggy or occasionally goes down would be a far less serious and really superficial concern. In fact, it is really more ‘smart contracts’ that are relevant here, not ‘blockchain’ specifically. So you could, for example, have the infrastructure of the market exist as an open source tool that is really facilitating multi-sig timelocked bitcoin transactions. This could actually be a very exciting bitcoin layer 2 utility (or ‘3’ if built on Lightning for whatever reason) as opposed to the likes of Augur, which bundles the entire functionality into a tool on top of Ethereum.

This is all just academic, to me, though, hence my deliberate agnosticism in the post towards specific projects. I am first and foremost a fan of prediction markets and applaud any effort to make them work.