Here:
https://drive.google.com/file/d/1NXpgnvlZgs6Be4Oa9zlVX394BfR4O8UW/view?usp=sharing
is a recent, preliminary draft of some work we have done on analyzing the kinds of anti-pre-revelation suggested in
and how these games might be applied in the context of Kleros. (This is an attempt to address the eventuality that an attacker might try to influence the Schelling point of a case by announcing her votes in advance.) Feel free to comment/ask questions here.
I read the paper briefly. Looks very interesting. Lots to digest.
What I don’t understand is how would you identify who to bet against in Kleros system? Since I don’t know who are the selected Jurors and what is their address.
How would I as a better bet against them?
That’s the thing, if no one pre-reveal, you should not be able to know who to bet against. But if someone reveals his vote prematurely (for example by making a smart contract which would burn some deposit if they vote differently of what they promised in the smart contract see https://github.com/kleros/kleros-attacks/tree/master/contracts/early-reveal-penalization) you could bet against him.
The goal is to prevent people from revealing prematurely.
Interesting, didn’t think of it like that. Will this be implemented in the next version?
No, it’s still research level and is not scheduled for Athena release (next release).