Community Moderation
As noted in the Sybil Defense article, computers cannot properly judge quest completions because:
completions can be any kind of content,
completions must implement unpredictable quest requirements,
completions must meet minimum quality criteria that are entirely a matter of human perception,
completions can be published on any available platform on the Internet, most of which fight against automated access to prevent DDoS attacks and content scraping.
In addition to judging content, computers are very bad at risk assessment, especially when it comes to the outside world, since risk assessment in many cases is heavily based on social reputation.
For this reason, Questfall relies entirely on the help of its users to solve such problems, which is implemented in the form of community consensus through voting, or in other words, community moderation. It protects Questfall users from scams and helps to distribute fair rewards to miners according to their efforts, by censoring three types of entities:
In addition, community moderation is the cornerstone of the defense against Sybil attacks used to secure mining. Therefore, the voting itself must be protected from any kind of manipulation, multi-accounts, bots, and blind/random voting. And it is.
The main principle of securing the community moderation voting is that moderators are rewarded or penalized with Silver, depending on how their vote matches the final consensus. Aside from directly burning Gold or QFT, community moderation is the only way to earn Silver by adding value to the entire community through voting in line with the majority.
The incentives are balanced so that, on average, the penalties of random or blind voting outweigh the rewards, making it a losing strategy. The system also generates honeypots to balance the odds of encountering good or bad voting outcomes.
There are other safeguards, such as balancing bypass costs with rewards, and splitting completion moderation into two phases to protect community moderation from any possible abuse.
Last updated