Set reports

Market temperature. A signed score from −100 to +100: positive means appreciation, negative means depreciation, and 0 means unchanged. The size of the move and its standing among tracked cards both matter. The percentage return remains visible alongside the score.

Calculation. Score = sign(return) × 100 × √(min(|return| ÷ 20%, 1) × direction rank), rounded to a whole number; nonzero moves retain a minimum score magnitude of 1. Direction rank is the fraction of eligible same-direction cards with an absolute return at or below this card’s return; ties share a rank. A 20% move fills the movement input. A tiny gain cannot earn +100 solely by leading its peers. Frozen: −100 to −70; Cold: −69 to −21; Steady: −20 to +20; Hot: +21 to +69; Blazing: +70 to +100.

Benchmark and limits. Individual cards rank against all eligible tracked cards in the same grade universe, period and endpoint date, including cards hidden by search or filters. This is our tracked coverage, not the entire Pokémon market. Cards without comparable endpoints are Unrated. Cohort and index badges use a movement-only score with direction rank set to 1; they are not ranked against other cohorts. Scores describe past movement, not volatility or future returns. Saved legacy temperature evidence keeps its original 0–100 scale and is marked “legacy”.

Comparable assets. Raw Near Mint, PSA 9 and PSA 10 are separate universes. Charts and changes use dated provider estimates, not executable offers or realized sales. “24H” compares consecutive dated daily observations, not a rolling intraday feed. “6M” uses 180 days. A return requires qualified observations on both endpoint dates; a shorter history is not substituted.

Cohort index. Eligible exact assets form a fixed, equally weighted cohort rebased to 100 at the period start. Exclusions remain visible. Results describe the tracked selection, not the entire Pokémon market. Shared multi-species or multi-type cards can appear in more than one report.

Value Outlook · outlook-v0. Weighted daily log trends (7 days: 20%, 30 days: 50%, 90 days: 30%) with 30-day exponential damping. Requires qualified endpoints over 90 days. Scenario lines reflect alternative recent slopes, not confidence intervals. Parameters are proposed defaults; predictive advantage is unproven.