Another aspect is dynamics.  Perhaps when the market is created, there is no relevant insider info so all the trading is based on common knowledge.  Once an insider enters, it's almost as if the market is "resolving", in one direction or another, just not with perfect certainty yet.

Is there a particular direction that the overconfident ought to be wrong in? 

If so, that would be something to trade on. If not, it's not a problem, right?

Maybe all the overconfident are offering is incentive for insiders to participate.

Undisciplined

But if the participants are irrationally confident, it conflicts with the notion that the signal is informative.

Ultimately, I'm not sure you can get around the tension between signal accuracy and market participation..... 

?  Unless I'm missing something... or unless an "average of overconfident insiders" is actually informative.  But even so, you'd have to posit some behavioral rationale for overconfidence.

econ

Kalshi Misunderstands the Economics of Prediction Markets (Matt Levine)

denlillaapan

But if the participants are irrationally confident, it conflicts with the notion that the signal is informative.

Ultimately, I'm not sure you can get around the tension between signal accuracy and market participation..... *right*?  Unless I'm missing something... or unless an "average of overconfident insiders" is actually informative.  But even so, you'd have to posit some behavioral rationale for overconfidence.