A crypto rating is a shortcut. Someone has read a pile of material about an asset and handed you a letter, a score or a position in a ranking so that you do not have to read the pile yourself.
That is useful only if you can check the shortcut. A grade with nothing visible underneath it is a stranger's opinion with a typeface. A grade with its workings published is something you can argue with, and arguing with it is the point.
Here are five questions worth asking of any crypto rating you come across, including ours. Each one has an answer you should be able to find without emailing anybody. Our answers are below, with the links that let you check them.
1. What does it measure, and how much does each part count?
Plenty of ratings will tell you what they look at. The more useful number is how much each part counts, because that is where the judgement actually lives. Two raters can look at exactly the same six things and disagree completely, simply by weighting them differently.
Ours are published, and they add to one hundred:
Tokenomics and value accrual, 28 percent. How the token is issued, who holds it, and whether holding it has any claim on what the project earns.
On-chain usage and economics, 22 percent. Whether anyone is actually using the thing, and what that use produces.
Security, decentralization and durability, 18 percent. Whether the system can be broken, captured or quietly switched off.
Ecosystem and moat, 12 percent. What is built on it, and how hard that is to move elsewhere.
Relative valuation, 12 percent. What you are paying for the above.
Market structure, 8 percent. How the asset actually trades.
You can disagree with those weights. That is a real disagreement about crypto, and you can only have it because the numbers are on the page. The framework they come from is written up at cryptograde.ai/framework.
2. Can anything override the score?
This is the question most people never think to ask, and it changes how you read every grade a provider publishes.
In our framework the weighted score is not the end of it. On top of the six dimensions sit nine pass or fail screens, for conditions such as an anonymous team, unaudited or previously exploited contracts, supply concentrated in insider wallets, or thin and wash traded liquidity. A tripped screen places an asset in our F grade at any score, including a score high enough for a much better band.
Our published bands are 80 and above for A, 60 to just under 80 for B, 40 to just under 60 for C, and below 40 for F, with those screens overriding all four.
Here is what that looks like in our own record. On our October 10, 2026 record, XRP carries a score of 55.8 and a grade of F / Failed, and it does not hold a position in our ranking. A score of 55.8 sits inside our C band. XRP is a substantial asset: in our own description of it, the native asset of the XRP Ledger, a blockchain built for fast payments and cross border settlement, with deep markets and established institutional relationships, alongside a comparatively concentrated validator structure and large holdings controlled through escrow.
Two things are true about that F at once. A low score is not the only road into our F grade, and our published record does not say which road a given asset took, so we do not guess in public. What the pairing shows a reader is the mechanism: in a rating that runs pass or fail screens, the letter is not a label stuck on the score, and reading the score alone would have told you the wrong thing.
3. Is the letter published, or did somebody derive it?
Related, and worth checking separately. If a provider shows you a score and you work the letter out yourself from a band table, you can be wrong in exactly the way question 2 describes. We publish the grade itself as its own field, and we quote that field. We never calculate a letter from a score, and neither should you. Where our record also carries the tier name, we print it beside the letter, as in B / Sound or F / Failed.
4. Who pays the rater?
Ask it bluntly, because the answer shapes everything above it. A rating is worth reading in proportion to how little the rater has riding on the result.
CryptoGrade does not work commercially with the assets or projects it covers, and it does not sell data. Nobody we grade pays us anything. That is the only reason a piece like this one can use our own F grade on a large, widely held asset as the teaching example.
5. Does it tell you what it cannot tell you?
A rating that only ever volunteers its strengths is telling you less than it appears to. So, plainly, here is where our own published record stops.
It carries the score, the letter grade, the ranking position, our research notes and our one line assessment for each asset we rate, each of them stated against the date it was read from. When a position changes, our record can name the other assets that crossed it, which is a complete answer about a position and says nothing about why anyone moved.
What it does not yet attach is the reason a score moved. If that is what you want, log in to your free CryptoGrade account to learn more. The explanation behind a move is work in progress on our side, not something our framework is unable to reach.
And one boundary that is permanent rather than in progress: we publish research, grades and scores, and we never tell you what to do with your money. No reading of your own holdings, no price target, no instruction about how much of anything to hold. What we owe you is the method and the record. What you do with them is yours.
Where to look
Our published ranking is at cryptograde.ai/tokens, ordered by our own score rather than by market capitalization, and the framework behind every grade in it is at cryptograde.ai/framework.
Logged out, the site shows you the top 10. A free account, which costs an email address, opens our ranking beyond those 10 and the research notes behind named graded assets, and it opens Alerts: you pick the crypto assets you care about and ask to be told when our grade on one of them changes.