The honest odds
What the data says happens to most coins and most traders, why the game is stacked, and what the people who last do differently.
This lesson is short on jokes. Before you put money in, you should know what usually happens to it.
What the data says
- The analytics firm Solidus Labs studied more than 7 million pump.fun tokens launched from January 2024 to March 2025. It found that 98.6% showed signs of a rug pull or a pump-and-dump. Fewer than 100,000 kept more than $1,000 in their trading pool.
- From November 2025 to October 2026, about one in 58 pump.fun launches graduated to a full trading pool, and 79% traded only on launch day (Bitquery, October 2026).
- In January 2025 a Dune data analyst found that only about 0.4% of pump.fun wallets, 55,012 out of more than 13.4 million, had made $10,000 or more in profit. Pump.fun's co-founder argued the real number could be ten times higher. Even then it would be about 4%.
- Bitquery sampled one launch week a month from November 2025 to August 2026. Of about 2.9 million ordinary wallets that bought new Solana memecoins, 74% had less SOL than they put in two weeks later. The median buyer lost $13. That count includes network and pump.fun fees but not tips or trading app fees, so the real picture is a little worse.
- The famous coins tell the same story: about two-thirds of TRUMP buyers were in loss by mid-2026, and 86% of LIBRA traders lost money (Nansen).
Why the odds are this bad
- Speed. Bots buy in the first second. You are later than you think.
- Information. Devs and insiders know when, what and how much before you do.
- Costs. Every trade pays fees, win or lose.
- Supply. Tens of thousands of new coins a day compete for the same buyers.
- Emotion. FOMO buys tops. Panic sells bottoms.
Why one big winner usually isn't enough
Made-up numbers to show the math. You put $20 into each of 50 new coins, $1,000 in total.
- Step 1: 40 coins die. You manage to sell each one at -90%, so each $20 becomes $2. That's $80.
- Step 2: 9 coins drop by half. Each $20 becomes $10. That's $90.
- Step 3: 1 coin does a 10x. That $20 becomes $200.
- Step 4: Total: $80 + $90 + $200 = $370. You hit a 10x and still lost 63%.
- Step 5: To break even, the one winner would need to turn $20 into $830, a 41.5x, and that's before fees.
What the people who last do differently
- They size small. Each coin gets a small slice of their trading money, so no single coin can hurt them much.
- They decide the exit before the entry: where they take profit, and where they cut the loss.
- They take profits on the way up. Selling half at a 2x gets the stake back.
- They accept that most trades lose. The aim is small losses and a few bigger wins.
- They keep a journal. Memory lies. Notes don't.
- They practise with paper money first: trades at real prices with no real money at stake. On Trenchpilot, pick paper when you start a bot, because a new bot trades real money otherwise. A notebook of the trades you would have made works too.
If you still want to play after all this, good. You'll play with your eyes open. The next tracks show how launchpads work, how to read a coin, and how to spot the traps before they spot you.
Check what you learned
0 of 4 answered
Finished reading? Tick it off and your progress is kept in this browser.
This guide is for learning. It is not financial advice. Most memecoins go to zero and most people who trade them lose money. Only ever trade what you can afford to lose.