The grading algorithm is a score based system which looks at the technical implementations of a token at the time of mint–what we call “Mint Conditions.” Our current grade focuses on decentralization and NFT permanence with future updates adding additional categories, including tokenomics, community, and market integrity.
A Scorecard contains the resulting scores and aggregate grade of an NFT. A Scorecard is first generated for an NFT’s associated Collection, then generated for an individual NFT, which is part of the underlying Collection.
Collections are projects or tokens using the same overlying contract. The unique NFT’s Scorecard inherits all scores of the underlying Collection and scores generated in real-time specific to the individual NFT. Examples of this include an individual NFT’s Image latency, whether or not the image is accessible, etc.
Scorecards are based on rubrics for specific types of NFTs. The only currently supported rubric is the visual-media rubric which accounts for most collectibles and PFPs. This rubric defines the outcomes of each Scorecard when applied to a visual media-based NFT.
To calculate the overall grade of an NFT, each score (0, 1, or null for n/a) is summed then divided by the total available scores which are applicable to the corresponding NFT. Because there is a wide-variety in the way that an NFT can be created, this system allows for the scoring to be flexible based on how the asset is created. For example: NFTs which contain on-chain assets skip scores which are directly related to linked-assets (metadata/image gateways, latency, etc.) but may score lower on other aspects such as contributing to Blockchain bloat.
Weight System
Each datapoint used in grade calculation has a weight depending on the importance of the item. For example, the image and metadata being available in the first place will weigh much higher than the energy consumed to mint it.
If information in a Scorecard is deemed to be inaccurate by a third-party, a dispute can be submitted with an alternative grade and an explanation of the score modifications. Eventually these disputes will have some form of incentive to encourage accurate token scores.