Proposal for "Request Removal" Incentive

Proof of Humanity
1

Greetings,

I would like to clarify before beginning my post that this does not concern the “challenge request” feature for pending registrations. This concerns the “request removal” feature for already registered profiles.

I have noticed a flaw in the incentivization for removal requests of an already registered profile. Removal requests follow the same policies as the pre-acceptance submission challenges, however they differ in the fact that there is no deposit in reserve by default since a registered profile has already had the submission deposit released upon successful registration.

This creates an issue in removal request incentivization because if a request removal is made and the other side of the request does not match the deposit, the good faith requester is financially penalized by having his removal request deposit returned to him minus gas. In this case, the requester would have “won” with the registration being removed and the PoH registry integrity being preserved, but have lost financially.

There are many such cases where a request removal deposit would not be matched by a rational actor, such as an obvious error like an incorrect ETH address in the video that slipped by the pending registration phase and made it to being fully registered. Kleros has always ruled against incorrect ETH addresses, so any rational person confronted with the request removal would not match the deposit, resulting in an ultimate financial loss for the requester because of a default judgment.

I hope that we can address this problem and properly incentivize active registry curation. It is important to the integrity of the registry

I have two different proposals for the community to consider and discuss:

Proposal 1: Deposit is Held Until Registration Expiry

In this case the submission deposit would not be returned to the submitter after registration, serving as a constant incentive for registry curation by good faith challengers, just as is the case in the pending registration stage. The deposit could then be carried over to the next registration period, functioning kind of like an apartment deposit carrying over to the next lease period. It would be returned in the case of a self-removal or unwillingness to renew PoH registration.

Proposal 2: Tax Levied on UBI Drip for Request Removal Rewards Pool

Instead of holding the deposit permanently, I propose that a undefined percentage of each registration’s UBI drip be taxed and then added to a common rewards pool that then serves to fill the default defense deposit for a request removal. This would be attractive to both challengers and defenders, because the challengers are always guaranteed to be rewarded for a valid removal request as determined by Kleros and defenders would not need to worry about putting up the first round deposit for a removal request, serving as insurance against bad faith request removals. This would also allow the submission deposit to continue to be released to the registrant after successful registry.

As a final thought for this proposal, it could be that the tax is not persistent and only levied registry-wide upon the initiation of a removal request, scaled to properly incentivize the challenger after ETH gas fees are considered.

Edit 1: The Taxed UBI in proposal 2 would be released back to the community, just as it was taxed, in the event that the challenge fails.

Thoughts?

3 Likes
2

I think Proposal 2 makes sense. There does seem to be a clear gap in the procedures.

1 Like
3

I like proposal 2, which I guess can be seen as imposing a fine on PoH as a whole for not performing proper screening. Regarding edit 1, you could just generate the necessary UBI when/if the removal is confirmed, which would simplify the logic.

Still, I think proposal 2 would be quite hard to implement since it would certainly require the use of an ETH-UBI oracle and would require tighter integration between UBI and PoH (I think the PoH contract is not aware of the UBI contract currently).

1 Like
4

You raise an interesting point. I support this idea.

5

I like your proposal as a whole. Tho the first idea of keeping the deposit till registry end is inconvenient. A lot of new people would think the project is a pyramid scheme, and also if POH is aimed for people of low resources, keeping 0.157 eth deposited goes against the spirit.

An alternative would be that the person who successfully challenged a registered user gets 15 days of extra UBI, equal to the amount that would have accrued the challenged profile. Thus, you create incentive to challenge regustered users and not punish correct profiles.

1 Like
6

Without an incentive to remove profiles from the registry, we could have some people that shoulden´t be there and nobody makes the first step to remove them. It is not good for a desired curated registry. I agree with the second proposal. Even a method where the one who ask for the removal could keep the UBI token acurred till that moment for the wrong profile.

1 Like
7

Normally the system should work even without removals. If we ends up having a second layers providing anonymous identities, we will actually need to get reed of removals altogether (as profile on 2nd layer would be anonymous and not linked to first layer ones, removing them would be impossible).

Removals are still good in early stage to recover would something fail with POH screening.

8

Great idea, This is what we want so-called “managed by community”. Proposal 2 is a very good move. We need help from experts in the crypto-economy field to work out a scheme and procedure.

9

I appreciate your response and would like to hear more about this. I don’t understand what you’re getting at.

10

Thank you for your reply Laurence. Hopefully we can implement something soon. Gas has already broken the incentive for submission challenges as well.

1 Like
11

You could look at page 7 “Anonymous Sybil Resistant Identities” of this doc: Proof of Humanity - Google Docs

The limitation of this scheme is that profiles could not be removed before expiration (as we’d lose the link between the profile and the submission).
So that removals will probably not be possible in the future if we implement this anonymity feature.

12

I believe I understand now. Correct me if I am wrong but removal would interfere with the batch nature of the traceable ring signature.

I find this future prospect interesting, however something needs to be fixed in the PoH removal system incentivization since any missteps in its infancy hold much larger consequences.

1 Like
13

I agree with PNK_Brahmin, currently it’s needed an incentive to remove wrong profles. At least until the poh users have the knowleadge to challenge other profiles and understand all the mechanisms. Until that moment, where the PoH will be more secure because the first bareer is stronger, a second mechanism of removals is needed, besides the batch ring implementation of the document that @clesaege has posted.

Another options, maybe easier to implement is to lock the profile UBIs at the Removal Request and if the removal is success ONLY the challenger can seize that UBIs. With the current implementation anyone can do that, and even the human of the removal request profile can move their UBis to another wallet and the challenger who put money at risk with the removal request, just lose money.

14

The problem with this is that you limit challenging incentives to registered PoH members only, effectively vastly reducing the universe of people who can do it.