Improve the address scoring algorithm

[cedricwaxwing]

• Addresses that have proven personhood through BrightID, Gitcoin Passport, Worldcoin, Proof of Humanity
• Participation in Gardens, 1Hive, Token Engineering Commons
• POAPs from a wider range of events and geographic locations
• DAO multi-sig signers
• Addresses which have received significant funds from DAO multi-sigs
• Addresses which have donated through Gitcoin or Giveth
• Addresses which hold Pooly NFT
• Addresses that have contributed to RossDAO, AssangeDAO, UkraineDAO and others
• Addresses participating in projects that are from marginalized geographic areas
• Addresses of project owners in Gitcoin (some are multisigs but often they're not)
• Users of swaps, for example, total volume the wallet traded.
• Or less bot-prone activities like minting an NFT on OpenSea.
• using the DAO Drops POAP as part of the next system - easiest way, a poap from last round means you have at least x points
• newer action can add more power than older action - older wallets maybe outdated, might not be informed of the market…
• more power if you have experience in that area of projects
• impact badges
• another idea: all addresses active before 2017 have some points
• marco says best solution: use logarithmic normalization after calculating multipliers (round 1 was like one smart contract = 1 point, one POAP = 3 points, max is 100 points")

[magentaceiba]

Current repo's script from Round 1: https://github.com/dOrgTech/DAO-Drops/tree/main/scripts

In Round 1: There's a max # of points each address can have, and a multiplier on each type of data set, for example a smart contract is worth 1 point, a POAP is worth 3 points. The purpose of that is so one person doesn't have one voting point (the minimum # in round one was 10 points), and one person doesn't have 10,000 points.

Question for Round 2: do we grab existing data sets, or run our own chain scraping? One way we can grab datasets is that different projects have done airdrops and aggregated their own sets of informed stakeholders.

[magentaceiba]

To do week of 9/6/23: @LordMarkDev will look into where to find these address sets, and think more on best optimization of an intelligent and dynamic scoring algorithm for who has voting power.

[orishim]

from @LordMarkDev: assign value to each type of action (e.g. 1 point per $100 donated, 1 point per $100 donation received, 5 points per smart contract deployed, 3 points per DAO voted in, 2 point per POAP) then run normalization over the outputted scores

[LordMarkDev]

I made a document about the v2 data sources, "DAO DROPS V2 ADDITIONAL DATA SOURCES":

https://docs.google.com/document/d/1rlSSZCgXQ5GRf4eUTOtgODflliiH_CRERf0t5j-v77k/edit?usp=sharing