is the 95% the accuracy? as in: if the person does die that month, then it's a yes, and if they don’t die that month, it’s a no, and the Yes must be 95% of all predictions?

Also, how far in advance are we talking? currently most death markets probably resolve with 99% accuracy, because someone hears reports of someone’s death, and they bet a lot before the market is resolved. But that’s not really “predicting” anything.

@Nat @Santiago, Resolves if by 2030 there exists a death prediction model that is so efficient, that if you were to put someone's name/details in it, it will tell you the correct day they would die +-30 days, 95% of the time.

Sort of like shimigami eyes from deathnote

A model? it said a market.

model or market. I'd imagine if you had good enough model(s), it would then calibrate the markets to also work

Could you clarify what the resolution criteria is? And specifically, what you mean by / how you're determining the > 95%