This is the only one I’m personally aware of, keying off of a commonly known bias humans suffer from: https://arxiv.org/abs/2609.12878

Obviously not compelling on its own though.

> longshots lose 6.3 cents per dollar when each contract is weighted equally but gain 4.1 cents when related contracts are first grouped by parent event; the two-sided pattern is robust in Crypto and Politics but surprisingly absent in Sports.

This seems like a relatively small inefficiency in the markets, no? If the margin of error notable in a prediction markets is <10 cents per dollar, that implies that they are pricing events pretty accurately. Especially post publication of this paper, when sophisticated traders could trade away this inefficiency.