How research and claim scores work

The daily sample contains up to 250 coins ordered by market capitalization from CoinGecko or CoinPaprika. Exclude incomplete data, observations over 24 hours old, prices at or below zero, market caps below $10 million, daily volume below $1 million, and recognized stablecoins. Stablecoin detection is heuristic and may miss pegged assets; check the asset yourself.

A positive 7-day move up to 25% earns 25 points. A positive 30-day move up to 60% earns 25. A 24-hour move between −3% and +8% earns 10. Daily volume divided by market cap between 2% and 30% earns 20. Market cap of at least $100 million earns 10; daily volume of at least $10 million earns 10. Gains beyond any surge limit cap the score at 59.

High: 70–100; Medium: 40–69; Low: 0–39. Established means a top-100 market-cap rank; Emerging means outside that group within the sample. Neither describes coin launch age. Trends and reported turnover do not establish fair value, execution liquidity, or future gains.

A hypothetical coin has a $200 million market cap, $20 million daily volume, +2% over 24 hours, +10% over 7 days, and +20% over 30 days. It earns 25 + 25 + 10 + 20 + 10 + 10 = 100 points. Change its 24-hour gain to +20%: the model caps it at 59 despite other positive factors. This demonstrates the rule, not a tested success rate.

There is no validated probability of profit or backtested win rate. The model omits token unlocks, fundamentals, holder concentration, contract vulnerabilities, exchange depth, and manipulation checks. Do not interpret High as “safe to buy.”

For supported public contract actions, the claim model adds 30 points per strong claim-name match, 8 per action up to three, 12 per familiar read-only probe up to two, and 8 for a native balance; a proxy subtracts 5. Scores are bounded from 5 to 99. High begins at 55, Medium at 30. View/pure functions and common ownership/admin claims are excluded as action leads.

A claim signal describes contract function evidence. It does not prove wallet entitlement, an open claim, project identity, token value, or security. Market scores and claim scores must never be compared as though they measure the same thing.

One successful snapshot per calendar day is stored using Antigua time (UTC−4). It stays unchanged after publication. There are no invented editions for missing days. Provider failures trigger a 15-minute cooldown and keep the latest saved edition clearly dated. Ordinary visits can prepare the edition without duplicating it. The unattended daily schedule is pending verification of an authenticated market-data connection.

Archived editions can be searched by coin and filtered by date, strength, and coverage. Read the source time before using a result. Corrections to explanatory pages are recorded in the changelog; a future model change receives a new version rather than rewriting old scores.