KARMA · A PROJECT BY KOMBUCHO
YOU WERE NEVER EARLY. YOU WERE THEIR EXIT LIQUIDITY. ✦ 45% OF CALLED COINS ROUND-TRIP TO ZERO. THAT IS THE BASELINE, NOT THE CRIME. ✦ RECEIPTS, NOT VIBES. ✦YOU WERE NEVER EARLY. YOU WERE THEIR EXIT LIQUIDITY. ✦ 45% OF CALLED COINS ROUND-TRIP TO ZERO. THAT IS THE BASELINE, NOT THE CRIME. ✦ RECEIPTS, NOT VIBES. ✦

▶ HOW KARMA WORKS

THE MESSAGE

every other tracker ranks callers by profit. that rewards the one thing that wrecks you.

the fastest way to top a pnl board is to sell into the people who copied you. so a profit board is a leaderboard of the callers most willing to dump on their own followers.

karma scores the opposite thing: whether following someone tends to hurt you.

pump.fun's callout rewards now pay callers when their followers profit. a payout can't show the other half: whether the caller sold into you. karma shows it call by call, each one tied to the transaction.

HOW IT WORKS

1

read the chain

pull a caller's real on-chain history, every single call, no self-reported wins.

2

two questions, kept apart

did they sell into you? timed from their own trades against the moment they called. and what did following them do to your money? neither answer rescues the other.

3

a grade you can check

one letter, S to F, backed by the receipts. we grade ourselves too and publish the misses.

THE MARKET WE AUDITED

20%
of calls, the caller sells into their own followers within 30 minutes
74%
of calls cost the follower money, or went nowhere
25%
of calls ever offer a clean 2×

measured across 533 callers and 11,779 settled calls. read that and you want to call everyone a scammer. don't. this is the asset class, not the crime. that's why we score against the market, never against zero.

HOW IT SCORES — TWO QUESTIONS, NEITHER RESCUES THE OTHER

SOLD INTO YOU

callers buy a minute or two before they post. we read their own trades: if half their bag is gone within 30 minutes of the call, at a profit, the people buying the call were their exit. 20% of calls look like that. it is the most consistent thing a caller does: the ones who do it keep doing it.

honesty — weighted heaviest, and it floors the letter on its own

YOUR MONEY

a call is bad for you if you were down a day later, it collapsed with no chance to get out, or it went quiet: nobody traded after the call. 74% of calls are bad, 25% offer a 2×. a caller who rode a rug down with you lands here, as a bad picker, not a dumper.

money — bad calls against 2× shots

FLOORS

selling into followers at 1.4× the usual rate caps the letter at D, at 2× at F. no 2× in 6+ calls caps it at C. an S takes 15 settled calls and an A takes 10: a short hot streak isn't a record.

caps, never boosts

THE FAIRNESS RULE

dying isn't the crime — almost everything here dies. a caller who rode it down with you is graded as a bad picker, never as a dumper. a coin that ran and gave you a real window to sell at a profit before it died doesn't count against the caller at all.

every rate is pulled toward the usual rate in proportion to how thin the record is, so a five-call streak can't mint a star, and the band on the score widens while the sample is thin. a caller at every usual rate scores 50, a B. the typical settled return is on the card and is not part of the letter: the middle call in this corpus is down 48%.

THE GRADE SCALE

SChadrarely sells into you, and followers did clearly better than usual (15+ settled calls)
ASolidfollowers did better than usual (10+ settled calls)
BFairabout as good as a typical caller for the people who follow
CCoinflipbelow typical for followers, or none of 6+ calls could have doubled
DExit Liquiditysells into their followers more than most callers
FLarpersells into their followers twice as often as most, or worse

BEYOND CALLERS

DEVS

a dev is graded on two things across every pump.fun launch: did they dump on the buyers (sold their bag in profit and the coin now sits 80%+ under its high), and did other wallets walk out of the creation block with 20%+ of supply. the average dev lands mid-C; an S takes a quarter of the norm's rates on both.

DEV KARMA · AGAINST THE DEV NORM, NOT THE CALLER ONE

TRADERS

a wallet with no callouts is graded on its own buys, priced the same way. on apps like Fomo every buy is pushed to followers who copy it, so a buy really is a call: those are tagged Fomo calls, read from Fomo's gas relayer on-chain. undoxxed wallets stay undoxxed.

NO PUBLIC CALLS · THEIR BUYS ARE THE CALLS

MEMORY

every launch block, every caller's calls and every coin opened is written down per wallet. a crew is wallets that bought 3+ launches' creation blocks together and cashed out in seconds. a side wallet is named only past a hard bar: 10+ of one caller's calls over 2+ weeks, bots filtered, proven when there's a direct transfer between the two.

SEPARATE FROM THE GRADE · EVERY ROW LINKS TO THE CHAIN

COINS

past day one, the risk is who sits on the supply today. every top holder is checked against everything above, the book is summed into a share held by bad records, clean ones and unknowns, top holders' last-24h selling is read from the trades, and an hourly snapshot shows who's arriving and who's leaving.

WHO HOLDS IT NOW BEATS WHO BOUGHT AT LAUNCH

we grade ourselves too, and publish where we're wrong →

LIKE THE PROJECT? BUY $KARMA →