I propose markets on the votes of individual US Supreme Court justices, as follows:
On the website fantasyscotus.lexpredict.com users predict the votes of individual justices with approximately 70% accuracy, without any money involved. If you were to run similar markets on here, you could quantify the added value of people making predictions with money, compared to when people make predictions for free.
To keep it easy for PredictIt, the Rules could be phrased as: "On July 31, 2019 23:59:59, FantasySCOTUS will record the vote of Justice X in case Y as 'Reverse'."
I would certainly be interested in all of these markets, and judging from the users of FantasySCOTUS (no pun intended), so would many others.