I applaud the OP for experimenting with new ideas in this space, as every idea is worth trying. And honestly given how bad this space is, I actually wonder if t

I applaud the OP for experimenting with new ideas in this space, as every idea is worth trying. And honestly given how bad this space is, I actually wonder if the only bad ideas are the ones we've already tried. Still, the cynic in me is unconvinced that Twitter could ever be a jumping-off point for valuable debate, no matter how brilliant the landing pad might be. As the operator and lead moderator of a large community it's been my observation that there are three things which most reliably derail challenging conversations: 1. The assumption that words are reliable. It's almost impossible to converse on contentious topics without using words and idioms that become corrupted in transmission because Avery speaks "Avery's English" and Quinn speaks "Quinn's English"—two languages which appear identical on the surface, and seem identical when interrogated but are actually slightly different in the most perniciously subtle of ways. This makes it impossible for Avery to know how to craft sentences which are received as intended. Or even to recognise that corruption occurred; we tend to interpret these corruptions as part of the disagreement, not part of a language barrier. 2. The tendency of people to so identify with their ideas, that criticisms of ideas automatically translate into criticisms of them as a person. This can be mitigated in part with the use of careful, defensive qualifiers, but ultimately this requires acts of consciousness-raising and a deep well of good faith by all parties. I'm not optimistic that deep wells of good faith can be constructed for impersonal, online debates with your ideological opposites. 3. The unwillingness of people to accept retractions of prior statements. All too often I see debate fall apart because Quinn says something inartful and Avery jumps on a misunderstanding of it. This almost always ends one of two ways: either (a) Quinn doubles down out of a tendency to never admit fault, or (b) Quinn attempts to clarify or retract the original statement, but Avery dismisses the clarification/retraction and pursue the original phrasing as "revealing what Quinn really thinks." If I were forced to invent a debate platform, my first experiment would be to require all formal responses be preceded by an interposed round of validation: a mandatory restating of the argument you're about to respond to using different words. Not until your interlocutor is satisfied with your restatement would you be able to post an actual response with counter-arguments. And rather than having spectators vote on arguments, they would vote/comment exclusively on the quality of (and goodwill shown in) the restatements—with bonus points for a successful steelmanning. There would certainly be no attempt to infer a "winner" of any debate by any statistical means. This almost certainly fail as being too infuriating for anyone used to the "submit" button having immediate results.