Content
@
0 reply
0 recast
0 reaction
Dara Khan
@spookyaction-btc
https://warpcast.com/spookyaction-btc/0x70c1e423
1 reply
0 recast
0 reaction
xh3b4sd ↑
@xh3b4sd.eth
I haven't written down what we do at Uvio. The most reliable form of anti-sybil methods in my mind is A) an integrated proof of work system that B) stretches over longer period of time. Uvio rewards users as they use Uvio in a meaningful way. That is then hard to game. For your dApp that means, define the core function that users must facilitate in order to make the product useful for A) the product itself and for B) the users at large. And then incentivize this very behaviour. What you will notice is that there is some kind of "hardness curve" which determines how many malicious actors you take along your own incentive program. The harder, or the more "real" the incentivized user behaviour is, the less fraud do you incur. And conversely, the easier it is to extract value the more people will do it. At Uvio we reward users if they have been honest during the random truth sampling process. That means some users are randomly selected to vote, and if their votes turn out to be honest they get tokens.
1 reply
0 recast
0 reaction
xh3b4sd ↑
@xh3b4sd.eth
On scoring, Uvio collects the raw numbers about 1) how often you voted 2) how often you have been chosen but abstained 3) how often you were right with your stake 4) how often you were honest with your vote Those information allow us to calculate competence and integrity metrics. And those metrics in turn can be used in any algorithm in order to calculate something like "reputation". This is then per wallet. If anyone is interested I am happy to show you around and talk more about this stuff.
0 reply
0 recast
1 reaction
