We check our own labels
Track record
The first time we check a token we record our risk label. After 1 hour, 24 hours and 7 days we look at what happened. Below: how often tokens in each group lost 90% of price or liquidity, or left the market entirely.
Collecting live data: 24-hour results appear once each group has at least 20 tokens.
6
tokens tracked
live forecasts since Sep 24, 2026
Live forecasts
The label was recorded before the outcome was known. This is the honest test.
| Label | 1 hour | 24 hours | 7 days |
|---|---|---|---|
| High risk | too few · n=1 | — | — |
| Elevated risk | too few · n=1 | — | — |
| Low detected risk | — | — | — |
| Insufficient data | — | — | — |
Including reconstructed baselines
For tokens checked before tracking started, the starting point was rebuilt from stored snapshots and the label may have been computed after the drop. These numbers overstate accuracy — use them for scale only.
| Label | 1 hour | 24 hours | 7 days |
|---|---|---|---|
| High risk | too few · n=1 | too few · n=2 | — |
| Elevated risk | too few · n=1 | too few · n=1 | — |
| Low detected risk | — | — | — |
| Insufficient data | — | — | — |
How we measure
- “Gone or −90%”: price or liquidity fell 90% or more from our first check, or no market is left (pool removed or delisted).
- Late measurements are excluded: over 30 minutes late for 1 hour, over 3 hours for 24 hours, over 12 hours for 7 days.
- A rate is shown only when a group has at least 20 tokens. Numbers refresh every 15 minutes.
- This is statistics, not a promise: low-risk tokens fail too, and high-risk ones sometimes survive. Not investment advice.