A documented weighted sum over liquidity, lock state, contract control and sellability. Unobserved signals are named, never assumed.
How this calculation works — the formula, its inputs, what it assumes, and what it does not establish.
Score a token on what can actually be observed about it right now: how deep its pool is against its market cap, whether liquidity is locked or burned, whether the contract still carries mint or owner privileges, and whether it can be sold. Every weight is stated, the sum is arithmetic a reader can check by hand, and a signal no provider reported is removed from the weighting rather than counted against the token.
A provider that did not answer has told us nothing, and nothing is not the same as bad. Scoring an unavailable check as zero would punish a token for a provider outage and make two very different situations produce the same number. Unobserved signals leave the weighting, the remainder is renormalised so the score still means out of 100, and the confidence falls to say how much evidence is behind it.
Below half of the intended evidence, no score is produced at all. A number assembled from one signal looks exactly like a number assembled from six, and there is no honest way to show that difference as a figure. This is the same refusal rule the register uses for an incomplete holder set: not measurable is a different statement from low.
It is not a safety rating and no arrangement of these signals establishes that a token is safe to buy. Locked liquidity can still be withdrawn when the lock expires, verified source code can still be hostile, and an absent owner privilege today says nothing about an upgradeable proxy tomorrow. The score describes the evidence available at one moment, with each part of it named.
Score a token on liquidity, lock state and contract control — unobserved signals are named, not assumed. MoonMath labels the data basis for this tool as Observed, Formula, Heuristic.
No. The Token Launch Quality Score is an educational scenario tool, not financial advice, a recommendation, or a forecast. Provider coverage, liquidity, routing, fees and market conditions can make real outcomes materially different.
Because a provider that did not answer has reported nothing, not something bad. The signal is removed from the weighting and the confidence drops instead, so a provider outage cannot be mistaken for evidence against a token.
No. The score reflects what could be observed at one moment about liquidity, lock state, contract control and sellability. A lock can expire, verified code can still be hostile, and no combination of these signals establishes that buying is safe.