I'm reading Robin Hanson's paper on market scoring rules. I dived into it because I wanted to understand the theoretical basis behind automated liquidity algos in prediction markets.[1] The math is complicated, and I haven't fully grokked it, but many of the claimed properties of prediction markets--such as the market price reflecting true underlying probabilities--seem to rely on fully rational agents who are able to do deep mathematical calculations and bet accordingly, especially for large and conditional event spaces.
Thus, given the Austrians' allergy to mathematical models, I am wondering what Austrians' perspectives are on prediction markets.
I was also motivated by @realBitcoinDog's very public mistake in understanding how Predyx's algo works. Gotta make sure I don't fall into the same trap! ↩
I haven’t seen any deep coverage of it.
My sense is that they’d note many of the same criticisms you’ve brought up and say that they’re an inherently flawed but potentially interesting source of information.
There are lots of econ experiments where subjects can figure out a nearly optimal strategy despite not knowing how to solve for the Nash Equilibria. It’s similar to how little kids can learn to catch a fly ball without knowing differential equations.
Fair. It is encouraging to see that there is at least some solid theoretical foundation for an algorithmic prediction market though (vs. an actual order book of trades)
(Since, if you'll remember, one of my main critiques against prediction markets was lack of volume)
Prediction markets shouldn't have automated liquidity bots, because that just messes with the forecast. daahhhhh
Well I think that's my point. There seems to be some solid theoretical basis for the automated liquidity bot. But only if bettors are assumed to be rational, i.e. they know the rules and they know the optimal strategy for betting within those rules
The math is hard. I think I need to spend some time working out simpler examples. One of my goals is to build a tool that would advise people on how to bet given their beliefs in the probability of an outcome, as well as a tool to advise market creators on how to set up their liquidity bot.
It might never happen though... seems like a lot of effort for a side project for a site I don't even engage with much! ~lol
There’s a decent chance they’d pay you for something that helped users make smarter purchases.
There’s mad money to be made in prediction markets, and liquidity bots are the way to get it. But please, don't call them prediction markets, that’s just a buzzword. A dope name would be LiFe
BETCaSiNo! ~lol