Research methodology

What the research covers

Narrative research connects source material to candidate tokens. Early Radar evaluates token observations against evidence and profile gates. Wallet-flow research investigates observed activity from wallets that satisfy separate qualification rules.

These are distinct research surfaces. A discovered token is not an endorsed opportunity. A seed wallet or reviewed address is not automatically research-qualified. Live coverage depends on provider availability, deployment readiness and the evidence actually collected.

Sources, identity and freshness

Market observations record provider provenance against the exact stored metric. Current flow price admission requires DexScreener provenance. Wallet activity is collected through Helius-backed infrastructure when that collection is configured and working. A provider name is not an endorsement or a guarantee of completeness.

Token identity uses exact mint addresses rather than ticker similarity. Early Radar publication requires prices no older than 15 minutes and holder observations no older than 60 minutes, alongside its other gates. Missing, stale, future-dated or unverified observations must not silently become fresh evidence.

Discovery is not qualification

Wallet research admission checks approved and active status, a recent profile, complete uncapped fresh 30-day history, at least 20 closed legs across five tokens and seven observed days, no unmatched or excess sells, and positive result measures alongside its score requirements.

Flow publication additionally requires at least two independent wallets or clusters, minimum net activity per entity, fresh trades and price, liquidity, holders and acceptable risk evidence. These are unvalidated research admission rules, not proof that a wallet is skilled or a token will perform. Review approval cannot override missing evidence.

Frozen observations and unknown outcomes

Research entries retain the evidence available at publication. Later horizons are measured against the frozen observation rather than a convenient retrospective entry. Missing measurements become unknown instead of being filled with an invented price.

Reported research returns are gross token-price changes, not wallet profit or executable trading returns. A SELL flow outcome still describes the token return; it is not a short-sale profit. Fees, slippage and execution limitations must be considered separately.

Coverage and sample limitations

Collection is bounded. The market cycle uses a 150-token price budget and an 18-info-call budget; social priority is allocated within those limits. Admin coverage diagnostics are bounded samples, not a census of every Solana token or wallet.

Repeated scan counters are not unique-token conversion rates. Small cohorts, correlated wallets, unavailable history and survivor selection can distort apparent performance. Narrax does not claim a validated predictive edge or guaranteed returns.

Publication and corrections

Public guides are educational explanations maintained by the Narrax team, not live token recommendations. Their claims should be read alongside the evidence limits described here. We do not publish fabricated ratings, customer counts, performance statistics or audit certifications.

If a guide conflicts with current recorded evidence, treat the evidence as unresolved and raise the discrepancy through the available product support channel. Changes to research rules require a deliberate review; a marketing page never overrides application admission or publication gates.

Primary technical references

These references explain the underlying data and protocol concepts; they do not endorse Narrax or any token.

• Solana documentation: accounts
• Solana documentation: transactions
• Solana documentation: token authorities
• DEX Screener API reference