$100 for a Day: $88.30 Median, Worst in Four
- Median $88.30 (worst in four days) against a mean of $376.86 — 4.3x apart, all from PAIREX
- Excluding PAIREX and CATGPT, the other ten average $72.76, a 27.2% loss
- Golden dogs on day two: four samples, all negative, averaging $67.20
Same arithmetic every day: put $100 into each token on the Robinhood Chain hot board 24 hours ago, and see what it is worth now. Today's answer is a median of $88.30 — the worst in four days ($101.13 on 09-12, $94.68 on 09-11, $98.07 on 09-10), with 5 of 12 up and 7 down.
But the number worth staring at today is not the median. It is the gap between the mean and the median: mean $376.86, median $88.30 — 4.3x apart. All of that gap comes from one token. This piece is about what that does to your position sizing.
Data timestamp: 2026-09-13 00:09 UTC research package, back-calculated from the 24h change. This is a look-back, not a forecast.
Today's twelve $100 bets
| Token | 24h change | $100 becomes | Liquidity | Age |
|---|---|---|---|---|
| $PAIREX | +3,249.96% | $3,349.96 | $95,714 | 18h |
| $CATGPT | +344.78% | $444.78 | $7,435,001 | 24h |
| $ZZZ | +29.27% | $129.27 | $618,767 | 192h |
| $AI | +15.76% | $115.76 | $29,564,675 | 1446h |
| $PONS | +1.92% | $101.92 | $8,783,436 | 1366h |
| $CHUMP | -10.06% | $89.94 | $1,274,613 | 1054h |
| $FLYBRAIN | -13.35% | $86.65 | $703,611 | 51h |
| $CME | -26.23% | $73.77 | $688,986 | 119h |
| $ROBIN | -43.93% | $56.07 | $186,731 | 175h |
| $LPAD | -58.49% | $41.51 | $91,729 | 53h |
| $PONIE | -73.76% | $26.24 | $9,382 | 140h |
| $FRONTIER | -93.53% | $6.47 | $9,290 | 9h |
$1,200 in, $4,522.34 out — up 276.9% on paper. Looks great, doesn't it?
Take it apart: where that +276.9% comes from
| Method | Result | What it means |
|---|---|---|
| Mean of all 12 | $376.86 | Carried entirely by one token |
| Median of all 12 | $88.30 | The typical outcome: -11.7% |
| Mean of 11, excluding PAIREX | $106.58 | Still propped up by CATGPT |
| Mean of 10, excluding PAIREX and CATGPT | $72.76 | -27.2% |
That is what the return distribution in memecoins actually looks like: outcomes concentrate in one or two names, and missing them leaves your book negative. And when you had to decide 24 hours ago, PAIREX was an 18-hour-old token with a $95,714 pool and zero public project information — on the data, it did not look all that different from $FRONTIER, which is down 93.53% today.
So "average return +276.9%" is true and it is harmful. The numbers that should drive sizing are the median ($88.30) and the mean-without-the-two-winners ($72.76).
Then why not just buy the golden dogs?
The obvious response: if returns concentrate in a few names, buy only the daily golden dogs. Here is how the last four days of golden dogs (24h ≥ +100% with liquidity ≥ $20k) did on their second day:
| Listed | Token | Day-one 24h | Day-two 24h | $100 on day two |
|---|---|---|---|---|
| 09-10 | $PONIE | +175.42% | -38.39% | $61.61 |
| 09-11 | $JUGGERNAUT | +379% | -17.12% | $82.88 |
| 09-11 | $FRONG | +166% | -17.2% | $82.80 |
| 09-12 | $LPAD | +530.87% | -58.49% | $41.51 |
Four samples, four negative, averaging $67.20 — a 32.8% loss. Small sample, clear direction: waiting for a token to appear on a golden-dog list means buying the part that already happened. Whether today's two ($CATGPT, $PAIREX) break the pattern is tomorrow's entry.
Four days of medians, in one place
| Date | Median | Best | Worst | Up / down |
|---|---|---|---|---|
| 09-10 | $98.07 | PONIE $275.42 | CRUMBS $7.28 | 6 / 6 |
| 09-11 | $94.68 | JUGGERNAUT $479.29 | SIH $3.39 | 6 / 6 |
| 09-12 | $101.13 | LPAD $630.87 | STOCKTARD $36.60 | 7 / 5 |
| 09-13 | $88.30 | PAIREX $3,349.96 | FRONTIER $6.47 | 5 / 7 |
Four days, four medians between $88 and $102. The best outcome has ranged from 2.8x to 33x and the worst from $3.39 to $36.60 — the tails swing wildly while the middle barely moves. That is the shape of the market you are actually trading: a typical outcome of roughly break-even-to-slightly-down, decided entirely by whether you happened to hold one of the two outliers.
Three things today's numbers teach
- Read the median, not the mean. Whenever someone tells you a sector "averaged +X%", ask for the median. Today those two are 4.3x apart.
- Buying a basket of hot tokens is not defence. Buy all 12 today and the typical outcome is -11.7%; strip out the luck (the two golden dogs) and it is -27.2%. Diversification in memecoins lowers variance; it does not make expectancy positive.
- The losing tail does not fall more slowly than the winners rise. Today's worst, $FRONTIER, is worth $6.47 — a token that is nine hours old, sits in a $9,290 pool and carries five pools charging 80%+ (post-mortem). Every one of those features was visible before you would have clicked buy.
Turn it into a pre-trade check
Over four days, the three numbers most correlated with "still holdable tomorrow" are:
- Turnover (24h volume ÷ liquidity): today CATGPT 1.9x token-wide, PAIREX 53.1x, FRONTIER 187.9x. High turnover is churn — fast up, fast down.
- Market cap ÷ liquidity: CATGPT's 1.5x is the lowest on the board, ZZZ's 72.9x the highest. The higher the multiple, the less of the paper cap a sell wave can hit.
- Unique buyers ÷ unique sellers: below 0.8 is distribution. Today CHUMP 0.67, CATGPT 0.69, AI 0.70.
All three can be computed from any token trade page; today's full board is in the daily hot tokens report.
FAQ
How accurate is this $100 math?
It is an idealised back-calculation from the 24h change, with no slippage, pool fees or token tax. Reality is worse, especially in thin pools — $FRONTIER's $9,290 and $PONIE's $9,382 pools slip visibly on orders of a few hundred dollars.
Does a median of $88.30 mean don't buy today?
It means the expectancy of buying at random is negative — not that nothing is buyable. The only way to push expectancy positive is selection (turnover, market cap ÷ liquidity, head-count ratio, pool fee structure), not breadth. Not investment advice.
Why no holder data?
The upstream holder endpoint has failed for four days running and we do not invent numbers. The substitutes are in No holder data? Read the pool fee fingerprint.
Do launchers always do better than buyers?
Not necessarily, but the launcher side has a more transparent cost structure: on BigPump no initial liquidity is required, the curve phase charges a flat 1% pool fee (95% platform / 5% creator), graduation at 4 ETH creates the pool and burns the LP, and creator tax plus holder dividends only start after that. Launch page and book.bigpump.ai.
BigPump is a Tax NFT launchpad on Robinhood Chain and an independent project — not affiliated with, endorsed or sponsored by Robinhood Markets, Inc. Figures come from the daily 00:09 UTC research packages for 2026-09-10 through 2026-09-13 and are back-calculations. Not investment advice.
Disclaimer: memecoins are extremely volatile and most go to zero. This article is not financial advice. Do your own research and only spend what you can afford to lose.