Diamond-hand ratings should be explainable, not mystical.
The rating is a heuristic summary of holder quality, conviction, concentration, and sell pressure. It is designed to help users investigate faster, not to replace judgment.
Headline rating
Each token gets a single front-door read so users can decide where to dig deeper. The rating compresses several signals, but the component metrics remain visible next to it.
Holder quality
We look at legitimacy-related fields, real-holder ratio, and whether the current holder base still looks independent after collapsing suspicious grouping patterns.
Conviction and maturity
Diamond-hand percentage matters, but mature holders matter more. A token with a deep 30d+ or 90d+ base should read differently from one with shallow short-term stickiness.
Risk and distribution
We surface cluster concentration, internal transfer activity, and visible sell pressure so the score is not blind to coordinated supply or active distribution.
Freshness and limits
Freshness is part of the product, not a footnote. Scores and insights should always be read against when the underlying snapshot was collected.
Missing or partial data is shown explicitly. The UI should never silently treat unavailable signals as healthy or unhealthy.
Some tokens will always remain ambiguous. The terminal exists to tighten the search space, not to guarantee certainty.