The scoring engine

5.1What the number is

The score is a confidence figure, not a forecast. It is recomputed from scratch on every scan out of the data available at that moment: how much the token is actually trading, whether its market is the right shape and size to trade at all, whether its move is backed by volume, and what the contract and the holder base look like.

It is comparable within a scan. Two cards in the same batch were measured against the same rules seconds apart, so a 78 and a 61 are a genuine ranking. Across days it means less: an 80 on a dead Tuesday and an 80 in a frenzy describe different markets, which is what the Market Pulse bar is there to tell you.

5.2The four groups on the composition bar

The bar under the score splits it into four coloured segments. They are the four questions the scanner asks, and their relative widths tell you what kind of card you are looking at.

Two cards can reach the same score from opposite directions — one on turnover, one on a clean contract and a tidy chart. A high score with a visible red segment is not the same card as a high score with none, and the bar is the fastest way to see which one you have.

5.3Why this score

At full density every card carries a Why this score toggle. Expanding it lists what the scanner recognised on that token, in plain language, each with the value it actually saw. The counter beside Copy CA — 14 rules — is how many matched.

Reading the chips is almost always better than reading the number. They are the difference between “78” and “78, and the reason it is not higher is that ten wallets hold a third of the supply.”

5.4Reading a score in context

The number is never the whole card. Four things sit next to it, and all four matter:

• The delta — ▲ / ▼ against the previous scan. A card climbing to 78 and a card sliding to 78 are different situations.
• The lane — the score decides it first, then the security verdict can hold a card back. See Chapter 2 and Chapter 6.
• The risk strip — four tiles, and a red one ends the conversation whatever the score says. See Chapter 3.
• The calibration line — on cards scoring 70+, how tracked calls in that score band have actually behaved. It is the only line on the card built from outcomes rather than from the current snapshot. See Chapter 8.

5.5What the score is not

• Not a prediction, and not advice. It says the token matched a lot of what the scanner looks for at the moment it was measured. Nothing about what happens next.
• Not live. It is a snapshot from the last scan. Between scans it is history, and the older the batch the less it is worth — the header tells you its age.
• Not personal. Everyone on the same chains sees the same score from the same scan. Your filters change which cards you see, never what they score.
• Not a security verdict. The contract check is a separate gate with its own three-way answer — Chapter 6.
• Not complete on a brand-new token. Some checks need history that does not exist yet, and missing is never the same as clean.