We (the team at BlockBrain) were amazed at how bad most NFT pricing applications out there were. Especially for new joiners to the NFT market, price setting can be highly opaque - what is my NFT actually worth? Is this price a good deal for the NFT? Is it worth paying extra for a higher-rarity NFT?
So for this hackathon, we set out to develop a better system for estimating a realistic value of your NFTs, taking into account rarity and sales data.
This project consists of both a front-end and a backend. FE (React): https://github.com/theblockbrain/hellonftworld-frontend BE (Python-FAST-API): https://github.com/theblockbrain/HelloNFTWorld
Live demo here (Chrome with HTTP only, as I didn't have time to setup nginx/SSL): http://bb-hackathon-frontend.herokuapp.com/
The Backend is currently using a MongoDB Cluster as Database, but this can be easily switched out for any other DB.
The Backend takes care of getting metadata for a collection, calculating rarity, getting OpenSea Sales via their API & doing regression-analysis to estimate a price-function. While the frontend is currently only using Rarity-Points in relation to total Price, it is also prepared to use alternative metrics, such as rarity rank, and percentage over floor.
In the future, the regression analysis model will have to get validated using mean-error and the likes, but there wasn't enough time for this in this week. I had a blast coding all of this, and I hope you enjoy the project just as much as I enjoyed building it!