> ## Documentation Index
> Fetch the complete documentation index at: https://docs.qoory.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Scoring, rankings, and trending

> How Qoory separates long-term relevance from short-term market momentum.

Qoory uses several ranking systems for different questions. A score is a
derived ordering signal, not a price target, risk rating, or investment
recommendation.

## Project ranking score

The general project score is designed to surface established, relevant, and
actively maintained projects. It combines log-scaled signals so one extreme
number cannot dominate the entire ranking.

Signal families include:

* Market presence, such as market-cap rank, watchlists, TVL, and TVL momentum.
* Adoption and community, such as public social and GitHub activity.
* Legitimacy and relationships, such as funding and investor coverage.
* Recent activity, such as news, timeline events, hiring, search engagement,
  and outbound research interest.
* Website and documentation signals.

Recent engagement signals require activity from multiple users before they can
boost a project. The final score applies a freshness multiplier so stale records
lose influence over time.

## Hot score and tiers

The hot score is a separate short-window momentum signal. It emphasizes recent
smart-money positions, X mentions, news, TVL movement, GitHub activity, and the
age of the discovered project. Newer projects can receive an age multiplier so
early momentum is not drowned out by older projects with larger histories.

Qoory maps the result to labels such as `viral`, `hot`, `rising`, `early`, and
`watch`. Tier thresholds require both a score and, for some tiers, enough
independent signals to be firing.

## Curated discovery

Trending lists, fund rankings, stock rankings, market movers, and heatmaps use
purpose-specific inputs and bounds. They are not necessarily sorted by the
general project score or hot score.

## Known limitations

* Coverage varies across entities and signal families.
* A missing signal can mean unavailable data, not poor performance.
* Social and engagement signals can be noisy despite spam-resistance gates.
* Scores can change when source coverage, weights, freshness, or methodology
  changes.