Data Methodology

Current data sources

Providers are named on the page where their data appears, with a link to the source, so any figure can be checked independently.

How the data is fetched

Requests are made from our server, not from your browser, and the result is cached and shared across visitors. One upstream request therefore serves many page views. We normalise every provider response into our own internal format, so a provider can be replaced without changing what you see or how it is labelled.

Freshness labels

• Updated [time] ago — the reading is within its normal refresh window.
• Delayed — the last refresh attempt failed and we are showing the most recent good reading, with its actual age.
• Temporarily unavailable — no recent reading exists. We show a notice instead of a number.

We never describe cached figures as "live" or "real-time", and we never substitute an estimate for missing data.

Rounding and precision

Fee rates are shown as the provider supplies them. Derived figures — mempool backlog in blocks, difficulty change, transaction cost illustrations — are rounded, and the rounding is deliberately coarse where extra decimal places would imply precision the underlying estimate does not have.

Calculators

Our calculators use no external data at all. They work from the values you enter, each page states the formula applied, and nothing you type is transmitted to us or stored. That is also why they continue to work when a data provider is down.

Limitations

• Fee estimates are probabilistic; confirm the rate in your own wallet.
• Sentiment indices summarise attention and volatility, not value.
• Third-party data can be wrong; we display it, we do not audit the provider.
• Nothing on our data pages is financial advice or a prediction.

Found a discrepancy? Our corrections policy explains how to report it, and our fact-checking policy covers claims outside these data feeds.