Scroll through any thread about Kalshi's 15-minute Bitcoin markets and someone will say they're rigged. The last-minute swing that flips a sure thing is a real experience, and it hurts. So we went back through our own ledger: 281 live trades placed by our bot between October 8 and 10, 2026, across Bitcoin, Ethereum, Solana, XRP and ten other 15-minute markets.
We can't prove a market is fair from 281 trades, and we're not going to pretend to. What the data does show is where the money went, and it points somewhere much more boring than a rig.
The headline numbers
- 281 trades, 264 wins, 17 losses: a 94.0% win rate
- At the prices paid, break-even after fees was about 92.6%
- Net result after fees: +$34.78
A 94% win rate sounds like a money printer. It earned less than $35. That gap is the whole story.
Why one loss feels like a rug pull
Every Kalshi contract settles at $1 or $0. If you buy the favorite at 93¢, a win pays about 6.5¢ after fees and a loss costs about 93.5¢. One loss wipes out roughly 14 wins. When you've watched the price sit at 93¢ for ten minutes and it flips in the last ninety seconds, it feels engineered. Usually it's just the 6–7% outcome the price was already telling you about.
Where the money actually went: price paid
| Entry price | Trades | Wins | Net after fees |
|---|---|---|---|
| 85–89¢ | 11 | 6 | −$57.20 |
| 90–92¢ | 166 | 156 | +$39.64 |
| 93–95¢ | 104 | 102 | +$52.34 |
Eleven trades bought in the high 80s lost more than everything the 90–92¢ trades earned. Those were the rounds where the market was genuinely uncertain, and the price said so. The fix isn't a conspiracy theory. It's a rule: don't buy a side the market itself rates as shaky unless you have a real reason to.
Where the money went: market by market
| Market | Trades | Wins | Net after fees |
|---|---|---|---|
| Ethereum | 54 | 53 | +$47.30 |
| Bitcoin | 39 | 37 | +$15.89 |
| Zcash | 25 | 24 | +$8.55 |
| XRP | 61 | 57 | +$2.37 |
| NEAR | 20 | 19 | +$2.04 |
| Doge | 5 | 4 | −$10.31 |
| Solana | 56 | 51 | −$21.01 |
| HYPE | 9 | 7 | −$21.75 |
Six smaller markets (BNB, EUR/USD, natural gas, oil, silver, copper) added 12 trades, all wins, about +$11.70 combined. Solana won 91% of the time and still lost money. HYPE lost two of nine. Small samples, but the pattern is the same as the price table: the losses come from a handful of trades, and those few trades decide the week.
So, rigged or not?
Honest answer: 281 trades can't settle that question, and we have no inside view of Kalshi. What they do show is that you don't need a rig to lose money on these markets. Buying favorites at 90¢+ means you're right almost all the time and still one bad round away from giving it all back. Most people who call it rigged are reacting to the math, not to manipulation.
What to do with this
- Know your break-even before you click. The price is roughly the win rate you need, plus fees. Our free Kalshi break-even calculator does it for any price.
- Track wins per loss, not win rate. At 93¢, you need about 14 wins to pay for one loss.
- Be pickiest when the price is lowest. In our data the high-80s trades did the damage.
- Stand aside near settlement unless the setup is clear. The last minutes are where a sure thing turns into a loss.
For the full fee math, read Kalshi fees and the break-even win rate.
Data: LIVE settled trades from the DeskBot developer's own Kalshi account, October 8–10, 2026. Small sample; not typical results. Not financial advice; trading can lose money. DeskBot is independent software, not affiliated with or endorsed by Kalshi. We build and sell DeskBot, the bot that placed these trades.