> For the complete documentation index, see [llms.txt](https://nfa.gitbook.io/nfa-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://nfa.gitbook.io/nfa-docs/live-data-and-adapters/the-data-sources.md). # The data sources ## The data sources The difference between a good forecast and a lucky one is usually the data behind it. NFA maintains a layer of structured data sources that fetch current, authoritative measurements and hand them to the engine as anchors. When one of these sources answers a market, the forecast is tied to a real number instead of the model's memory. ### The principles {% hint style="success" %} The value of this layer is not any single source. It is that the sources are chosen and combined so that anchoring is reliable, reproducible, and honest about its own limits. {% endhint %} * **Authoritative and current.** The layer draws on primary sources: official statistics, central-bank publications, official weather products, regulatory filings, and open data feeds. The aim is always the source closest to how the market will actually resolve. * **Deterministic.** A given market is matched to a given source and returns a given value, repeatably. The same question does not get a different anchor from one run to the next. * **Honest about provenance.** A live reading, a pre-release estimate, and a projection are labelled and framed differently. A forecast is told whether it is standing on a settled fact or an estimate with a range, and never treats one as the other. * **Silence is explicit.** When no source can responsibly answer a market, the engine records that and falls back to open research, rather than stretching an unrelated source to fit. A source never guesses just to have an answer. ### The breadth The layer spans the categories where prediction markets cluster, including: * macroeconomic and monetary data (growth, inflation, and policy rates), * official weather and climate measurements, * company and market data from regulatory filings and public markets, * elections and legislative status, * geopolitical and conflict signals from open-source intelligence, * and media, culture, and science indicators. When a market falls in one of these areas, an anchor is often available. When it does not, the engine forecasts from research and reasoning and says so through the [quality badge](/nfa-docs/the-forecasting-engine/quality-tiers.md). ### Why it earns the anchored badge A forecast anchored by one of these structured sources carries the blue **anchored** badge, because it rests on a measured value rather than recall. A value found by an open web search does not earn that badge. The distinction is deliberate: it keeps the strongest label reserved for the strongest evidence. Two capabilities in this layer are worth their own pages: the [Ukrainian OSINT signal](/nfa-docs/live-data-and-adapters/ukrainian-osint-signal.md) and the [price models](/nfa-docs/live-data-and-adapters/price-models.md) that value short-horizon markets with math instead of debate. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://nfa.gitbook.io/nfa-docs/live-data-and-adapters/the-data-sources.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.