Comment by @paulburke • Hey

Interesting approach, thanks for sharing! I question how effective this type of scoring is when it seems fairly subjectively based on collecting and publis

Stats

Comments

{"profile_id":"0xd8","handle":"paulburke.lens","lens_score":0.77} you are within the top 111 profiles ;-) I agree with pretty much everything you write. I guess - not sure though - that the public bigquery dataset provides every publication on @lensprotocol , independent of the app (id). Even with biometric proof I haven't seen a convincing approach (yet) to connect a wallet 100% with a person (the biometric proof). I'm not sure though if "proving personhood" can and should be the goal of a "lens score". Imo the first and most important question is: what do we actually need to be on a scale from 0-100 sure, that a profile/account is or is not. When it comes to a social graph. Is it likeliness of personhood, is it "spam" (signal to noise), is it recommendations (this profile might be interesting for you with a probability of 96%), ... For me it's primarily "trust" (in the motivation of the person behind the profile) and content/profile discovery. That's probably the reason why I used the calculation and input data as described. (creating sth I see some value in - based on my experiences on lens so far. Besides getting some attention for the discussion, which was of course the main motivation.) Yes, I guess the process behind the maintenance of such a score is massive. Every airdrop (distribution) after the first one is dramatically more challenging :-).