Content pfp
Content
@
https://warpcast.com/~/channel/quotient
0 reply
0 recast
0 reaction

Jordan pfp
Jordan
@ruminations
In the very early stages of building an interactive + configurable Farcaster leaderboard, as a first step towards @quotient's vision of building a Palantir-esque client for builders/creators growing communities on Farcaster. One early insight - our algorithm optimizes for relationships representing "acknowledged value" over simple reach/centrality, giving a significant boost to Farcaster dev creators. This approach identifies @darrylyeo, @stevedylandev.eth and @proxystudio.eth as the three most influential Farcasters in April. It also makes a lot of sense, if you think a reputation metric for a platform should align with that platform's strategic priorities.
4 replies
1 recast
17 reactions

Darryl Yeo đŸ› ïž pfp
Darryl Yeo đŸ› ïž
@darrylyeo
wowow, that’s me! Curious how you measure “acknowledged value” here – are you inferring this from reactions, contents of replies, or something else?
2 replies
0 recast
7 reactions

artlu đŸŽ© pfp
artlu đŸŽ©
@artlu
use of wowow should be a 10x score boost
2 replies
1 recast
1 reaction

Jordan pfp
Jordan
@ruminations
Great question. At a high level - incoming relationships (like, reply, recast, follows) are acknowledgements of value. Those are weighted by type (i.e. a recast has a higher weight than a like), frequency (i.e. 5 likes between User A and User B during the time contributes to the weight of the relationship). In our model, the reputation / credibility of the counterparty also matter; 100 likes from 948393!* matter very little, whereas a like from @dwr.eth (or you!) matters a lot. There's a lot of other dimensions to incorporate, but starting with something straightforward -- so anyone can plug the algo into their own community of interest and get a ranked list of members/affiliates, to use for things like rewards.
2 replies
1 recast
5 reactions