Three of four pump.fun coins never reach a 5,000 dollar market cap. That part we knew. The new part is what happens after. Between 5,000 and 1,000,000 dollars, the ladder stops being a massacre and turns into a coin flip with better odds: each rung passes about two of every three coins that reach it.
Call it the step rate. Of the coins that reach one rung, the share that go on to reach the next one. On today’s record that share sits between 60 and 75 percent at every level, from the bottom of the ladder to the top.
The numbers, from 2,471 launches with a full six hour trace:
| Market cap rung | Reached it | Share of all launches | Step rate from the rung below |
|---|---|---|---|
| $5,000 | 625 | 25% | the first rung |
| $10,000 | 380 | 15% | 61% |
| $25,000 | 239 | 10% | 63% |
| $50,000 | 164 | 7% | 69% |
| $100,000 | 121 | 5% | 74% |
| $500,000 | 73 | 3% | 60% |
| $1,000,000 | 55 | 2% | 75% |
Read the last column top to bottom. It does not shrink. The 500,000 dollar rung, the hardest single step by count of survivors, passes a slightly smaller share than the 10,000 dollar rung does. Sixty percent against 61 percent, close enough to call it the same number.
That is the odd shape of this market. The whole filter sits at the bottom. Getting from a fresh launch to 5,000 dollars is where 1,846 of 2,471 coins die. Once a coin is through, the next wall is not much harder than the last one. A coin at 100,000 dollars is not facing a meaningfully tougher climb than a coin at 10,000. Both are looking at roughly a two in three chance of the next rung.
The counts still collapse, because two thirds is a losing rate repeated. 625 coins make the first rung. Only 55 make the last. Multiply six steps of about two thirds together and you get about one in twelve, and 55 out of 625 is one in eleven. The math is boring and it holds.
What the flat step rate does not mean
It does not mean a 100,000 dollar coin is a good bet. Two thirds of the 121 coins that reached 100,000 went on to half a million at some point in their first six hours, and that is the entire claim. The record says nothing about where those coins sit now, whether the peak lasted one print or an hour, or what the exit cost looked like on the way. A peak is a moment, not a trade you could have taken.
It also does not mean the ladder is fair. These are the same launches where 65 percent peaked inside their first five minutes. Most of the coins that reached a rung at all reached it fast, on thin books, and then stopped. The step rate counts a coin that touched 25,000 dollars for one trade exactly the same as one that traded there for an hour. The table measures arrival, not survival.
And it is an average over 2,471 coins, not a law for any one of them. A coin with a bot-heavy book or a top ten holding half the supply is not the average coin, and the board exists precisely because the average is not the answer to that question. The cards on the board score the book behind each pick, and the checker does the same for any mint you paste in. The ladder tells you what the crowd of launches does. The grade tells you what the individual coin is holding.
Put the two together and the finding gets sharper. The bottom rung is where most coins fail, and the bottom rung is exactly where a launch is too young for any grade to mean much. The evidence a coin has a real book takes time to print. The window where most coins live and die is shorter than that. That gap, more than any single rung, is what the six hour record shows.
The earlier ladder piece covered the first rung: three in four launches never reach 5,000 dollars. The flat six hour record adds the other end. Neither one tells you which coin clears the next rung. Nothing in this data does.
Method
Sample: 4,659 pump.fun launches observed in the tracking window. 2,471 of them have a full six hour trace and form the sample for every number above. Market cap rungs are peak values within that six hour window, so a coin that ran after hour six counts only what it did inside the window. Step rates are ratios of raw counts: the 500,000 dollar step is 73 of 121, no time weighting, no adjustment for how long a coin held a level. Six of the eight numbers in the last column rest on counts of 380 or fewer, and the top two rest on 121 and 73 coins, so treat the 60 to 75 percent range as a band, not a precise rate. Peaks measured on thin pools can overstate a cap, which is why pool depth is scored separately on every card. No coin was named, graded or recommended in this piece.