Every detected token gets a price + liquidity snapshot at +1h, +6h, +24h, +7d. We tag a token as rugged when its LP drops below 10% of the detection-time depth. The numbers below are the last 30 days. Refreshed every 30 min.
Snapshots
1,564,965
Tokens tracked
501,453
Window
30 d
Source
DexScreener
free public API
+24h
24 hours after launch
First night after launch is when most rugs land. Numbers here separate the calls that held from the ones that didn’t.
Safe (≥75)
Held LP ≥ 10%
100%
1899 of 1900
Rugged
0%
1 of 1900
Tokens we backed. We want low rug rates here.
Risky (40–74)
Held LP ≥ 10%
100%
6745 of 6746
Rugged
0%
1 of 6746
Mixed signals — we expect material rug rates.
Rejected (<40)
Held LP ≥ 10%
100%
25834 of 25835
Rugged
0%
1 of 25835
Tokens we flagged. High rug rates = correct calls.
+7d
7 days after launch
A week in is enough time for sniped peaks to flatten and pump-and-dumps to play out. Survivors here are usually the real thing.
Safe (≥75)
Held LP ≥ 10%
100%
2126 of 2127
Rugged
0%
1 of 2127
Tokens we backed. We want low rug rates here.
Risky (40–74)
Held LP ≥ 10%
100%
7599 of 7600
Rugged
0%
1 of 7600
Mixed signals — we expect material rug rates.
Rejected (<40)
Held LP ≥ 10%
100%
29262 of 29263
Rugged
0%
1 of 29263
Tokens we flagged. High rug rates = correct calls.
Methodology
How we measure "rugged".
Open heuristic. Not perfect, deliberately conservative.
What we count as rugged
· LP at the checkpoint < 10% of the LP at detection.
· The provider returned a snapshot (we don't tag rugged when DexScreener is silent).
· Detection-time LP > 0 (no baseline, no comparison).
What we don't count
· Price drops without LP drops — that's organic dump, not a rug.
· Tokens DexScreener doesn't index — we just leave the row empty.
· Soft rugs (LP slowly drained over weeks) — they may show up at +7d but slip past +24h.
· Honeypots that block sells — separate vector, caught by contract analyzer at scoring time.
Stats refresh every 30 minutes. The rolling 30-day window means early-launch days will look thin until enough tokens have aged through the +7d checkpoint. See /methodology for the full scoring rubric these classifications come from.
The radar is on
Watch the score work in real time.
Register, build a profile, and you'll see post-mortems on every alert in your panel — not just aggregates.
We use a session cookie to keep you logged in (always on) and Google Analytics 4 to measure aggregate usage with anonymised IPs (only with your consent).
Privacy policy →