The most reliable way to check a crypto data figure is to confirm that the provider states its method, that you can name the units and timestamps involved, and that you can repeat the calculation yourself. If any of those three links is missing, the number is hard to verify even when it is correct.
This guide turns that idea into a short reader checklist. It uses one published statement of intent as a reference point, and it treats that statement for what it is: a description of principles, not a formula for any specific metric.
Start with what the source actually commits to
The methodology page states that the provider strives to provide accurate, timely, and unbiased data for cryptoassets. Those three words are a useful frame for verification, because each one can be probed with a different question.
Accuracy asks whether the figure matches the underlying records. Timeliness asks when the figure was captured and how often it updates. Unbiasedness asks whether the selection of assets or markets excludes anything without saying so.
The same page also states that the provider's view is that it should always over-provide data, rather than censoring and policing information, and let users draw their own conclusions. In practice, that means a reader should not assume a displayed aggregate is the only reasonable one, or that the provider has filtered out low-quality inputs on your behalf.
The four-step reader checklist
Work through these in order. Stop at the first step that fails, because later steps rarely fix an earlier gap.
- Name the metric and the formula. Write down what the number is meant to represent before you read the value.
- Name the units. Record whether the figure is a price, a volume, a percentage, a count of assets, or something else, and in which currency.
- Name the date and cutoff. Record the exact point in time the figure refers to, not the time you read it.
- Reproduce it. Recalculate from the stated inputs. If you cannot, note why, since that gap is the finding.
A fifth step helps: check whether the provider invites corrections. The methodology page states that the provider actively listens to feedback and requests from its users to enrich the quality of the data on its website. That is a claim about process, not evidence about any single figure, but it tells you where to send a discrepancy you cannot explain.
A worked example using only labeled assumptions
The source does not publish a worked calculation, so the following numbers are invented for illustration only.
Hypothetical: three markets report trading volume of 1,200, 800 and 400 in the same currency over the same window. The simple total is 2,400. If a reader believes one market should be excluded, the total becomes 2,000, a difference of 400.
Calculation: (2,400 minus 2,000) divided by 2,400 equals 16.7 percent, rounded to one decimal place. That is the size of the disagreement created by a single inclusion decision.
The point is not the numbers. It is that the inclusion rule must be stated before the total can be checked. Without an explicit rule, two readers can compute two defensible totals from the same inputs.
Comparison: what each check catches
| Check | Question it answers | What it catches |
|---|---|---|
| Formula | What is being summed or averaged? | Misread definitions |
| Units | Price, volume, count, or percentage? | Currency and scale errors |
| Timestamp | As of when? | Comparing figures from different moments |
| Reproduction | Can I rebuild it? | Unstated filtering or weighting |
This table is a guide to the reasoning, not a description of the source's own review process.
What the source does not say
The methodology page does not establish any of the following, so a reader should not treat them as confirmed.
- It does not give a formula for any specific metric such as price, volume, or market capitalization.
- It does not state how often data updates, or the exact time of any snapshot.
- It does not list which markets or assets are included or excluded, or the criteria behind those choices.
- It does not describe how disputes or corrections are handled once a request is submitted.
- It does not state whether or how personal data submitted with a request is handled, beyond referring readers to a separate notice.
Because these items are missing, the page supports a process claim, not a numerical one. A reader cannot reconstruct a displayed value from it alone.
Handling requests and personal data
The page asks readers to review certain information before submitting a request. It also states that submitting a request means the reader agrees that personal data will be processed for the purpose of responding to that request, in accordance with the separate notice.
For a verification workflow, that has a simple implication. Keep the data you rely on in a form you can re-check, and keep any personal details you submit to the minimum needed for a response.
Applying the checklist to a chart
When you encounter a chart, apply the same four steps in order. Identify the plotted quantity, note the axis units, record the time range, and then try to rebuild a single point. If the rebuild fails, describe the failure precisely. A precise gap, such as an unknown exclusion rule or an unstated cutoff, is more useful than a general doubt.
The method above does not prove any figure is right. It only tells you whether the figure is checkable, and names the exact step where checking stopped.