A lot of this is intriguing, but I'm confused by the focus on preventing fraud through making it easier to deny claims -
We're incentivizing the community be on "the winning side" of the vote so, unless that bias exists collectively, the voting process will pick the right outcome in the long run. "Guaranteeing" a claim is a risky proposition as that would leave us open to fraud - that would also require a central authority to basically say "no, we override the community" which kind of kills the decentralized concept in the first place. The community will be able to have an open discussion on claims as they're voting to compile/share evidence. Worst case scenario as a policy holder is you get your money back.
Also, to be fair, anytime you're trying to extract money out of someone for a cause, you have to provide whatever they claim is sufficient evidence. The UMC community will be a reflection of all of our policyholders.
As a community member with a stake in the float, don't I actually just want to deny every single claim including the good ones just to keep my own stake higher? If anything I would expect there to be a sizable group of people who try to be as stingy as possible. Related to this, we all know that the (real or not) hot girl in trouble is going to get 3x the votes a man in the same situation would, another sizable chunk of the community will be pretty biased against everyone who doesn't look like them and so on. Decentralizing oracles work for prediction markets where there is a clear right answer that anybody can look up, but I don't see a good way to make this work for something with this much subjectivity. Am I missing something obvious? How do you plan on incentivizing the 'right' side as opposed to simply the side with the bigger bias?
We're incentivizing the community be on "the winning side" of the vote so, unless that bias exists collectively, the voting process will pick the right outcome in the long run. "Guaranteeing" a claim is a risky proposition as that would leave us open to fraud - that would also require a central authority to basically say "no, we override the community" which kind of kills the decentralized concept in the first place. The community will be able to have an open discussion on claims as they're voting to compile/share evidence. Worst case scenario as a policy holder is you get your money back.
Also, to be fair, anytime you're trying to extract money out of someone for a cause, you have to provide whatever they claim is sufficient evidence. The UMC community will be a reflection of all of our policyholders.
As a community member with a stake in the float, don't I actually just want to deny every single claim including the good ones just to keep my own stake higher? If anything I would expect there to be a sizable group of people who try to be as stingy as possible. Related to this, we all know that the (real or not) hot girl in trouble is going to get 3x the votes a man in the same situation would, another sizable chunk of the community will be pretty biased against everyone who doesn't look like them and so on. Decentralizing oracles work for prediction markets where there is a clear right answer that anybody can look up, but I don't see a good way to make this work for something with this much subjectivity. Am I missing something obvious? How do you plan on incentivizing the 'right' side as opposed to simply the side with the bigger bias?
Thanks for bringing that up. We've had very long discussions amongst our team members on both opposite scenarios- policyholders intentionally voting No to prevent the pools amount to be used for other agents, as well as colluding on Yes votes by making false claims and splitting payouts amongst pools. The first well address here since you ask, it is a very fine example of the tragedy of the commons. Ideally, wed want everyone to know were trying everyone on our end to keep solvent, so denying someone elses legitimate claim doesnt help their claim in any way. In no way are they entitled to "a bigger share" by denying a claim. We're not an investment fund and the most you can possibly own without making a claim is what you paid in.
However, we understand that might not be sufficient, and we need to infuse some incentives and regulations. Therefore, we intend to provide some incentives in the form of reduced cooling period, increased maximum amount attainable for good citizens of our democracy who execute their civic duty in a timely manner and with good results. For every user, wed be storing their voting history and would tally it against the outcome of the vote in smart contract itself. If theres a sharp discrepancy in the alignment of results over a sustained period, well notify the user with a warning. We aren't publishing user identity and we're not publishing vote results until the vote has concluded. Were also working on alternate ways of reputation in community, but might work on it in upcoming iterations.
To answer your bias vs correctness problem, we do intend on having a claims adjuster on the team to help identify the "right side," at least initially, while the community is built up. This will not necessarily swing votes, but it will give us better insight on who is more in-line with sound analysis. We're trying to build a community built on trust. People trust their insurance assessor even though they can carry the exact same biases as a group collective. We need to help get the community off the ground to build this trust.
Thanks for your questions!
