Review Methodology
Last updated August 2026
Every listing carries a score out of ten. This page sets out exactly how that number is produced, what each band means, and when a score changes.
How the score is built
Five criteria are assessed independently and combined by weight: contract and audit quality (25%), token distribution and vesting (20%), evidence the product exists (20%), team visibility and accountability (20%), and community quality (15%).
Each criterion is scored on the same ten-point scale, then multiplied by its weight. There is no discretionary adjustment applied afterwards, and no criterion can be waived because a project is prominent or paying.
How missing information is treated
Unpublished information scores as unpublished, not as neutral. A project that has not released its vesting schedule scores lower on distribution than one that has published an unflattering schedule.
This is deliberate. Rewarding silence would make the score easy to game by simply saying less.
What the bands mean
Scores cluster high across listings because projects failing badly are rejected before publication rather than published with a low score. A 7.5 here is not a mediocre result — it means a project cleared the threshold with a known gap.
When a score changes
- The project publishes or amends an audit.
- Tokenomics, supply or the unlock schedule change.
- The team changes, or previously anonymous members are named.
- A sale stage closes or the sale window ends.
- A credible report reaches us about the project's conduct.
What the score cannot do
It cannot tell you whether a token will rise. It measures how a project presents against verifiable standards, and well-presented projects still fail.
It cannot detect a determined fraud that has produced convincing documentation. We check what can be checked, and that is not the same as certainty.
It is a snapshot. A score set in March describes what was true in March.
The five criteria
Contract and audit quality
25%We look for a published audit from a firm with a track record. More telling than the badge is whether the findings were fixed — an audit carrying four unresolved medium-severity issues says more than any clean summary page. We also check what privileges survive launch: can the deployer still mint, pause, or blacklist?
Token distribution and vesting
20%A round that sells 8% of supply while the team holds 40% behind a three-month cliff is not an early opportunity, it is a queue you are standing in front of. We read the vesting table before the pitch deck. Long team lockups, a public unlock calendar and transparent treasury wallets move a score more than any partnership announcement.
Evidence the product exists
20%A testnet, a repository with commits from more than one contributor, a demo anyone can open. AI and infrastructure projects get extra scrutiny here because the claims are harder to verify and easier to fabricate.
Team visibility and accountability
20%Anonymity is not disqualifying — respected protocols have shipped pseudonymously. But it raises the bar everywhere else. We want either named founders with traceable history, or an anonymous team operating under multisig control with milestone-based fund release.
Community quality, not size
15%Follower counts are cheap. We look at the ratio of substantive discussion to price talk, whether the team answers hard questions in public, and how the channel behaves on a red day. Forty thousand members with moderators deleting criticism scores worse than four thousand people arguing about tokenomics.
Score bands
Audited with findings resolved, transparent distribution with published unlocks, demonstrable product, accountable team. Still a high-risk asset class.
Most criteria met with one meaningful gap — often a thin product record or a team disclosing less than we would like.
Real substance alongside real gaps. Typically unresolved audit findings, heavy team allocation, or a short unlock cliff.
Several criteria unmet. Listed because it is verifiably real, not because we find it convincing.
Falls below the threshold for publication. These are rejected rather than published with a warning.