Before you play, check what usually happens to tickets like yours: who launched it, how fast its curve filled, and what came 10 minutes after graduation. Measured on thousands of graduations, updated every hour. Odds, not advice: most tickets lose.
Peak figures leave out memes whose highest price is the very first one we read after graduation: that reading is often stale.
Bought at T+30 s after graduation, with 4% fees and slippage. "Higher at 5 min" = share of tickets above their T+30 s price five minutes later. A ticket that loses its price counts as −100%.
| Kind of ticket | Tickets | Higher at 5 min | Median at 5 min | Median at 10 min |
|---|
Tickets can fit several rows (a first launch can also be a flash). Medians, not averages.
The bonding curve fills up and the meme moves to PumpSwap, its first real trading pool.
For about 5 minutes, pump.fun automatically buys back and burns the token (around 17.6 SOL). The price goes up mechanically.
When the buyback ends, only real buyers are left. Most of the time there aren't enough of them, and the price falls back.
So a meme "flying" right after graduation tells you very little about the meme itself. It is the cycle, not the project.
Fees and slippage of 4% included. Observed frequencies on past graduations, not predictions.
Access is for holders of APO, the Aporius token. APO is launching soon: until then, Guardian access works as on RugWatch (connect your wallet once, no transaction, nothing is spent).
Price relative to T+30 s after graduation (1.0 = no change). The line is the median meme; the shaded band is the middle half (one meme in four is above it, one in four below).
Memes whose price disappears are left out of this curve (they would only drag it lower). The outcome table below counts them as −100 %.
A meme that climbs its curve slowly is easy to spot before graduation. It is also the profile that falls hardest after the Boost. Bought at T+30 s, with 4 % fees and slippage; a meme that loses its price counts as −100 %.
| Profile | Memes | Higher at 5 min | Median at 5 min | Median at 10 min |
|---|
Medians, not averages: a few broken price feeds can show +1,000,000 % and would make any average meaningless.
"Slow climb" mixes curves that took two minutes with curves that took a day. They behave very differently: the fast-but-not-flash ones fall hardest.
| Curve filled in | Memes | Higher at 5 min | Median at 5 min | Median at 10 min |
|---|
Everything above is measured after the fact, on displayed prices. To know whether anything holds for real, a paper test follows one flash-graduation profile we found in the data, using real Jupiter quotes: what you would actually get, slippage included. No wallet, no money, every trade logged, losses included. We publish the result whatever it is.
"Human pace" is what a person can realistically do after an alert. "Bot pace" is faster than any human and shown for comparison only.
Tell us what you want to see first: this page is built with its first Guardians.