You are viewing a single comment's thread:
RE: Anti-Phishing Proposed Solution
I see no utility in applying stake as a metric. It would make far more sense to me to use rep. There are several stakeholders with far more stake than any reasonable stake threshhold that could maliciously manipulate such a mechanism profitably. Reputation is not so potentially manipulable, and is a more meaningful metric regarding trust. Rep threshhold could be set to, say ~250, which would require at least three high rep accounts to report a link as sketchy, while dozens of accounts possess enough stake to surpass any reasonable threshhold based on stake.
Otherwise this is a great idea, that may well keep people safer on Hive.
Thanks!
0.00124811 BEE
That would actually make Rep useful for something for once!
So u r saying world-travel-pro can have 76 reputation and keep false flagging while having rewards declined so we can't even reduce reputation.
Anyone with high reputation once gone anti hive can stop our ability to reduce their reputation while continuing to exploit phishing flagging feature.
View more
If u r talking about reputation score, anyone can get hacked with 60 or more reputation and it will take alot of downvotes to reduce it to negative score (till then, more accounts would be compromised).
This is a stake based system to protect the nature of decentralization.
If you are talking about few people reporting it, then have high requirements and u can follow proxy (same proxy proposal and witnesses but a sperate proxy feature to not merge with those 2) reports of anyone thus your REPUTABLE reporters can also fit in this current model. Sorry I am not understanding u currently. Do u mean reputation score? If yes, it's not effective in this wide spread airdrop scams and memo transfers. Some users recover account in 5 days at max so we have to negative reputation them which is a very serious punishment.
Stake has utterly failed to decentralize Hive.
Then only whales can trigger the report.
Using reputation score enables longstanding users to trigger the report when at least a couple people report it. Using stake enables 1 whale to do so, potentiating manipulation for financial benefit.
How does that game the reporting system? With the threshold at ~250, they'd have to have 3 accounts to overcome the threshold, even if the accounts were >80 rep.
View more
I see no utility in applying stake as a metricI brainstormed all of edge cases and any system except stake can have some ways to game around the system. This is more edge cases proof.
How can rep be gamed in this usecase?
The purpose of the post above was to spark a debate regarding anti-phishing solutions at the frontend level. If you suggest that these solutions should be reputation-based, I do not have a problem with that ATM. Currently, we need frontends to consider the fact that implementing anti-phishing measures at the frontend level will not kill the spirit of decentralization.
I see some flaws in the reputation system, but I either do not want to debate them or I am not a skilled enough debater regarding that specific point at the moment.
Since both you and I agree that frontends should take action, let us keep pushing this narrative so we can lower the phishing rate.
Thanks for understanding!
@valued-customer
The devil is in the details. You haven't responded when I asked how rep can be gamed in this use case. You have neither addressed how using stake can be gamed in this use case when I point out ways that can happen. What is the real reason you are adamantly insisting on using stake as a metric rather than rep?
You insist on using a metric that will enable more centralization of power regarding phishing attacks on Hive, while trying to claim using rep won't work, that is not supported by facts. You don't lack debating skills, you lack factual basis for this insistence on stake instead of rep. Why?
View more