Head-to-head against settled outcomes

Each night we take the pump.fun mints our radar called the day before, draw up to 300 of them evenly over the day, read RugCheck's free public report for each, and score both against what happened to the mint: rugged within 300 seconds of launch, graduated. Window: 3 of the last 7 UTC days on record (2026-09-26 to 2026-09-28). We are one of the two sides, and the outcome labels are ours (see the caveats). Nothing here is ranked.

When each side was asked. Our call is the radar's first call on the mint, made at a median age of 12.2 s, before the outcome existed. RugCheck was asked when the job ran: a median 45.1 hours after launch, when the outcome had already happened. A report read a day later is not a call made at launch, so the two columns answer different questions and are not the same test.

Each side over the mints it answered for (3 of the last 7 UTC days on record (2026-09-26 to 2026-09-28))

Our flag: rug odds of at least 50% at the first call (odds of 30% and 70% are in the JSON). RugCheck flag: the report lists at least one risk at level danger (second column) or at level warn or danger (third). Intervals are Wilson 95%.

loophole tape radar, odds of at least 50%RugCheck: any danger riskRugCheck: any warn or danger risk
mints sampled900900900
no answer (share of sampled)14.6% (131 of 900, 95% interval 12.4 to 17.0%)0.1% (1 of 900, 95% interval 0.0 to 0.6%)0.1% (1 of 900, 95% interval 0.0 to 0.6%)
flagged (share of answered)43.3% (333 of 769, 95% interval 39.8 to 46.8%)5.8% (52 of 899, 95% interval 4.4 to 7.5%)7.7% (69 of 899, 95% interval 6.1 to 9.6%)
answered with a settled rug label765892892
rugged within 300 s (share of answered, settled)46.5% (356 of 765, 95% interval 43.0 to 50.1%)46.9% (418 of 892, 95% interval 43.6 to 50.1%)46.9% (418 of 892, 95% interval 43.6 to 50.1%)
precision on rugs (flagged that rugged)74.2% (247 of 333, 95% interval 69.2 to 78.6%)5.8% (3 of 52, 95% interval 2.0 to 15.6%)20.3% (14 of 69, 95% interval 12.5 to 31.2%)
recall on rugs (rugged that were flagged)69.4% (247 of 356, 95% interval 64.4 to 73.9%)0.7% (3 of 418, 95% interval 0.2 to 2.1%)3.4% (14 of 418, 95% interval 2.0 to 5.5%)
counts behind precision and recallflagged and rugged 247, flagged not rugged 86, unflagged rugged 109, unflagged not rugged 323flagged and rugged 3, flagged not rugged 49, unflagged rugged 415, unflagged not rugged 425flagged and rugged 14, flagged not rugged 55, unflagged rugged 404, unflagged not rugged 419
graduated, among flagged1.2% (4 of 333, 95% interval 0.5 to 3.0%)44.2% (23 of 52, 95% interval 31.6 to 57.7%)34.8% (24 of 69, 95% interval 24.6 to 46.6%)
graduated, among unflagged4.4% (19 of 432, 95% interval 2.8 to 6.8%)0.6% (5 of 840, 95% interval 0.2 to 1.4%)0.5% (4 of 823, 95% interval 0.2 to 1.2%)

Same mints for both sides (768 mints both answered for)

loophole tape radarRugCheck: any danger riskRugCheck: any warn or danger risk
flagged (share of the common mints)43.4% (333 of 768, 95% interval 39.9 to 46.9%)5.9% (45 of 768, 95% interval 4.4 to 7.8%)7.8% (60 of 768, 95% interval 6.1 to 9.9%)
rugged within 300 s (settled)46.6% (356 of 764, 95% interval 43.1 to 50.1%)46.6% (356 of 764, 95% interval 43.1 to 50.1%)46.6% (356 of 764, 95% interval 43.1 to 50.1%)
precision on rugs74.2% (247 of 333, 95% interval 69.2 to 78.6%)6.7% (3 of 45, 95% interval 2.3 to 17.9%)21.7% (13 of 60, 95% interval 13.1 to 33.6%)
recall on rugs69.4% (247 of 356, 95% interval 64.4 to 73.9%)0.8% (3 of 356, 95% interval 0.3 to 2.5%)3.6% (13 of 356, 95% interval 2.1 to 6.2%)

Per UTC day

daysampledrug label settledrugged within 300 sours flaggedRugCheck danger flaggedours no answerRugCheck no answer
2026-09-2630029912841.2% (115 of 279, 95% interval 35.6 to 47.1%)3.0% (9 of 300, 95% interval 1.6 to 5.6%)7.0% (21 of 300, 95% interval 4.6 to 10.5%)0.0% (0 of 300, 95% interval 0.0 to 1.3%)
2026-09-2730029714646.5% (114 of 245, 95% interval 40.4 to 52.8%)6.0% (18 of 299, 95% interval 3.8 to 9.3%)18.3% (55 of 300, 95% interval 14.4 to 23.1%)0.3% (1 of 300, 95% interval 0.1 to 1.9%)
2026-09-2830029714442.4% (104 of 245, 95% interval 36.4 to 48.7%)8.3% (25 of 300, 95% interval 5.7 to 12.0%)18.3% (55 of 300, 95% interval 14.4 to 23.1%)0.0% (0 of 300, 95% interval 0.0 to 1.3%)

Method

RugCheck requests by result: answered: 899, http_400: 1.

Risk names on RugCheck reports that carried any warn or danger risk (mints per name): Creator history of rugged tokens 26, Low Liquidity 25, Copycat token 15, High holder correlation 4, Name Mismatch 2, Symbol Mismatch 2.

Caveats

The same figures as JSON (free): /v1/compare. Rows: /v1/compare/rows. Our own dated scorecard: /track-record.