What actually happened to the wallets you copy

Today's autopsy · 2026-08-06

GROUNDTRUTH

$Dachshund climbed to 84.9 SOL over 88m 23s. It was gone 3s later.

The 12 wallets that bought first had already been early on 1,058 launches. 976 rugged.

Subject: $Dachshund
9VSC...pump (mint)
Selected by: one named coin, start to finish
Corpus: 239,157 of 847,417 launches we followed ended in a rug

A finished case, published free and in full. Drawn from cases that reached a final outcome in the last 30 days (dated by when the outcome was recorded, not when the coin launched). We picked it as a rugged coin, picked for how bad its first buyers' record already was.

groundtruths.xyz · every call frozen before the outcome existed

Share card for 2026-08-06 → · archive

Check a coin yourself Go to this coin

01 · the null result

Deeper early liquidity does not protect you.

Scanners sell one signal: more early liquidity = safer. Across a 12.3× span of real SOL at the 5-second mark, failure stays flat — lowest at band 2, highest at band 4. No gradient. The signal is noise. These bands are pump.fun; Robinhood Chain has no bonding curve to bucket, yet its aggregate failure lands in the same place — 98.67% of 155,364 resolved.

75.06%

AGGREGATE RUG RATE · N=100,248

87–93%

FAILED (RUG+DIED) · ALL 5 BANDS

12.3×

LIQUIDITY SPAN · 0.88→10.90 SOL

the "deeper = safer" curve every scanner implies — illustrative, not data

x-axis: real SOL in the bonding curve at 5s · y-axis untruncated 0–100

Every scanner sells the dotted line. The bars are the data.

NULL RESULT — failure is non-monotonic and flat. Min band 2 (87.26%), max band 4 (92.63%). Higher early liquidity buys no protection.

N = 100,248bonding-curve traded population

bucket edges (real-SOL): 0.8826 | 2.4691 | 5.6 | 10.896 · bars = failed % (rug+died); aggregate rug within this population 75.06% · census 857,796

02 · the other chain

Robinhood Chain: same test, a live gradient.

The identical experiment on Robinhood Chain — five equal-count bands by real ETH on the pool at the 5-second mark, across the 24,220 coins that reached the 0.3 ETH floor. Failure does not stay flat here. It climbs every band, and the rug share climbs with it — 55% to 91%. More early liquidity does not buy safety; it buys a higher chance of being actively rugged.

73.08%

AGGREGATE RUG · N=24,220

86–94%

FAILED (RUG+DIED) ACROSS BANDS

55 → 91%

RUG SHARE · LOWEST TO HIGHEST BAND

rugged — liquidity pulled died — faded out the “deeper = safer” curve scanners imply

x-axis: real ETH on the pool at 5s · equal-count bands · y-axis untruncated 0–100 · red = rug share, faded = quiet death

Same dotted line every scanner sells. On Robinhood Chain the data climbs to meet it.

GRADIENT RESULT — failure is monotonic and the rug share runs 55%→91% across the liquidity span. Higher early ETH is a rug tell, not a shield. Of 135,869 Robinhood Chain launches, 61,185 never reached the 0.3 ETH floor and died on arrival.

N = 24,220early-liquidity traded population

bucket edges (real-ETH): 0.07 | 0.15 | 0.262 | 0.6016 · bars = failed % (rug+died), red = rug share; aggregate rug within this population 73.08% · RH census 174,381

measured, not guessed

What actually happens to a pump.fun token.

Across 857,796 pump.fun launches, each tracked to a recorded outcome. 96.26% rugged or died. 66% never even cleared a 2-SOL peak; every one of those died.

pump.fun · SolanaRobinhood Chain

Of the 294,100 coins with full trade-level labels (the only ones that carry a realized return), the share that ever reached —

Hit +50%47.5%

Hit +100%29.0%

Hit +300%10.5%

Sellable at a profit ⓘ22.6%

On Robinhood Chain, the same replay over the 10,861 launches that recorded pool depth — what a 1 ETH position actually returns at the peak —

Median exit returns16.6%

Got principal back15.9%

On paper the median coin peaked at 1.56x. Sold into the real pool at that same peak, it returned 16.6% — a 9.4x gap between the chart and the till.
n = 10,861 of 99,296 labelled RH launches: the only ones with pool reserves recorded. That subset is 2.2x deeper than the chain at 0.3 ETH and 2.5x more rugged, so it is the actively-traded end — the figures above are a floor, and the chain-wide reality is worse.
Bag size 1 ETH here vs 5 SOL for the pump.fun bars: different size, different currency, not a like-for-like race. RH pools are Uniswap v3; depth is approximated as x·y=k over total reserves, so the fill is indicative rather than exact.

Of all 857,796 launches: 0.155% ever graduated. 84.5% never doubled.

On Robinhood Chain (174,381 launches, 155,364 resolved): 82.84% died · 15.83% rugged · 1.33% survived · 0 graduated — there is no bonding curve to graduate from. 98.67% of every resolved launch failed; 19,017 are still trading.

Two populations, both stated, never blended. The 857,796-launch census carries a recorded outcome for every coin. The bars above are the 294,100 coins with full trade-level labels — the only subset with a realized return — and all four bars, including liquidity, are computed over that same 294,100. "Sellable at a profit" = peak REAL liquidity ≥ 10 SOL: enough depth for a modest exit. Every n is disclosed. This is the layer risk-scanners never measure.

Outcomes labeled 2026-08-06, covering all 857,796 pump.fun launches we have captured. Live scans are real-time.

the difference

Risk scores tell you now. We tell you what happened.

Every scanner grades a token in the present tense. None of them carry the outcome. Forward-labeled history is the asset nobody else publishes.

Point-in-time risk scanners

"This creator has rugged before."

A present-tense flag. Useful, binary, and already free from several funded tools.

creator_risk: HIGH ⚠
mint_authority: revoked
flag: prior_rug = true

GROUNDTRUTH outcome data

A rap sheet, not a flag.

The recorded track record — counts, medians, timings — pulled from outcomes already labeled in the capture.

Every launch, every recorded outcome, the median time-to-rug, and the operator's wallet pool — the full track record, not a binary flag.
GdRSPexhxbQz5H2zFQrNN2BAZUqEjAULBigTPvQ6oDMP — paste it below for its live numbers — or paste a Robinhood Chain 0x address for its ETH-denominated record.

Compared against the attributed-launch baseline of 67.3% — the share of launches we can attribute to a creator that ended in a rug (N=230,062 across 81,212 creators). Not the 27% figure for all 857,796 launches: 66% of those never clear a 2-SOL peak and rug at 0.0% — they cannot rug, they die. We attribute rugs more easily than quiet deaths (a pulled liquidity leaves a signed transaction; a token that fades leaves its dev anonymous), so attributed creator rug rates likely run somewhat high — we flag the bias rather than hide it.

try it

Paste a coin — or a wallet.

GROUNDTRUTH replays what actually happened to a pump.fun or Robinhood Chain coin — not the chart's best case, but what you could really have sold into. Paste a coin for its autopsy, or a wallet for its rap sheet.

Drop a token contract to see who's holding it right now: holder concentration, snipers still in, how many exits before rugss and fast exits recorded wallets are already in front of you — plus the launcher's rug history. Or drop a wallet address to pull its rap sheet.

How to read a card ▾

Realized exit — what you'd actually net selling into real liquidity, not the chart price.

Peak real liquidity — the SOL (or ETH on Robinhood Chain) truly available to sell into at the coin's deepest.

Verdict — the recorded outcome: rugged, bled out, graduated, or survived. Robinhood Chain has no graduation — there, survived is the made-it class.

New here? Start with these ▾

What is a rug? The people who launched a coin take the money out of it. Everyone still holding is left with a token nobody will buy. It is not a crash — it is a withdrawal, and it leaves a signed transaction on the chain with somebody's wallet on it.

Why isn't deep liquidity safe? Liquidity is how much real money sits behind a coin. More of it means you can sell without crushing the price — and it also means there is more worth taking. Depth tells you the size of the pot. It never tells you who is guarding it.

What is a bonding curve? The launch mechanism these coins use. Price rises automatically as people buy and falls as they sell, with no order book and no market maker. It is why a coin can climb and vanish inside a few minutes.

Why look at coins that already ended? Because their story is finished. A coin that is over can be checked against what really happened to the people in it. A live one is still a guess. The settled record is what makes the pattern visible at all.

groundtruth.scan() no scan yet

— —

pump.fun · Solana

◆ GROUNDTRUTH only — not on other scanners

Standard checks

Mint authority—

Freeze authority—

Top-10 hold—

Dev holding—

Market cap—

Liquidity—

Snipers still in—

Bundled buys—

Token age—

Real holders ⓘ—

Curve status—

// paste a Solana or Robinhood Chain (0x) coin or wallet above to run a live scan

— —

pump.fun · Solana rugs recorded fast exits recorded insider-linked

SOL balance—

Token holdings—

Pricing

Look up any wallet or creator — free.

Every lookup is free and unlimited — one address at a time. Paid plans let you search backwards through 850,000+ pump.fun launches (plus 174,381 on Robinhood Chain), screen wallets in bulk, and get alerted the second a wallet with recorded rugs launches again.

Free

always

Unlimited single lookups · one at a time · no signup.

Look up any wallet or creator
Full rug history, real identity, and autopsy
How they stack up against the average creator

Run a scan

Pro

watchlist

$29 /mo

Everything in Free, plus a live watchlist. Cancel anytime · 2 months free on annual.

Watch up to 10 wallets
Live alert the moment a watched wallet enters a coin
Card, SOL, or USDC

Get Pro

Desk Pro

25 seats

$149 /mo

Founders Club — first 25 seats at $149/mo. Everyone after is $299/mo. Monthly, cancel anytime · 2 months free on annual.

Everything in Pro, but watch up to 100 wallets
Bulk / reverse-index screener — search backwards, find everyone tied to a launch
Alerts across all 100 watched wallets

Join the Founders Club

Detailed ReportFree

Pull one wallet or creator, fully worked up — identity, every launch, and the operators behind them — as a clean report you can share or cite.

Get a report

See exactly what you get →

Every number here shows how many launches it’s based on. We rate other people’s safety claims — we don’t make our own. There is no green “safe” badge anywhere on GROUNDTRUTH, and there never will be.

Methodology

The working behind findings that need it. Written so you can disagree with us on the evidence rather than on the conclusion.

Can we match Robinhood Chain creators to their Solana track records?

Short answer: no, not for an individual. We can say something about the population and nothing about any one wallet, and we are not going to pretend otherwise.

Population. The same cut on both chains — creators with at least 20 launches who rugged at least 70% of them. That is 743 creators on Solana and 72 on Robinhood Chain.

Control. We shuffle each behavioural measure independently across the Robinhood creators. Every creator keeps a real value and loses their own combination of values. If real pairings are not markedly tighter than shuffled ones, the pairings are noise. We report the ratio between them.

Result. How someone deploys tokens does not identify them across chains. Who buys their launches does — weakly, and consistently. The buyer-crew measures beat the deploy-only baseline in 16 of 16 independent subsamples. Chosen on one half of the Robinhood creators and scored on the other half, which they were not chosen on, they still beat it: 1.23 against a baseline of 1.01.

What that licenses. Real pairings are roughly a fifth to a third tighter than shuffled ones. That supports a statement about the population. It does not support naming a wallet, and at this effect size it never will. No pairing has been published and none will be.

Why there is no pass or fail here. The baseline's own spread straddles the threshold we would have tested it against. Two runs of the identical configuration nine hours apart landed on opposite sides of it — one noisy measurement, read twice. With 72 creators the threshold framing is not decidable, so we report the effect and its spread instead of a verdict.

A result of our own that is void. An earlier version of this test defined a dirty creator as one whose launches mostly failed rather than mostly rugged, and scored better. On Robinhood Chain that selects launches that never got traction; on Solana it selects rugs. It was comparing two different populations and measuring the gap between them. It is recorded here because the number reached our own working notes before it was caught.

Coverage, which shapes all of the above. Every Robinhood Chain launch in this test carries an outcome label; on Solana roughly a quarter do. So the Solana side is drawn from part of that corpus, not all of it.

Dated on purpose. Robinhood Chain opened on 1 July 2026 and its buyer data is far thinner per launch than Solana’s. A thin population is where a real effect is hardest to see. This is a measurement taken on 6 August 2026, not a permanent finding, and re-running it produces a new measurement rather than an update to this one.

Liquidity isn't safety. It's bait.