Most new weather traders on Kalshi gravitate towards YES bets. It's "intuitive" to them: the forecast says 82°F, so you buy YES on the 81–82 bracket. It's also usually the wrong side of the trade. This article explains why the NO side has a built-in structural advantage on temperature bracket markets.
Here's what happens in every temperature market, every single day. Kalshi lists bracket markets: "Will the high be 80–81°F?" "81–82°F?" "82–83°F?" The NWS forecast says 82°F. Traders see that number. They buy YES on the brackets around 82°F. That demand pushes the price of the consensus bracket up. On a typical market, you'll see the peak bracket priced at 40–70¢.
Now think about what that means statistically. Weather forecasts, even good ones, have a typical error of 2–3°F. With that much uncertainty, the probability of landing in any specific bracket tops out around 20–40% on a normal day. Yet the market regularly prices the peak bracket at 35–50%. That gap between the statistical reality and the market price is the YES buyer's tax. The NO side captures that premium.
The forecast anchor is public and unanimous. In weather, everyone sees the same NWS forecast. When participants crowd onto the same bracket it predictably inflates prices above fair value.
Temperature is continuous, brackets are discrete. The difference between 81.7°F and 82.3°F is less than a degree, yet one hits your bracket and one doesn't. YES buyers are constantly exposed to this rounding risk; NO buyers on adjacent brackets benefit from it.
People undervalue dispersion. NO bets on the peak bracket win every time the temperature lands anywhere else — and "anywhere else" covers a lot of territory.
A city has six active brackets. The forecast says 82°F. True probabilities might be: 79–80°F at 5%, 80–81°F at 15%, 81–82°F at 25%, 82–83°F (consensus) at 25%, 83–84°F at 20%, 84–85°F at 10%.
If the market prices the consensus 82–83 bracket at 50¢ when the true probability is 25%, buying YES at 50¢ gives deeply negative expected value. You're paying 50 cents for a 25% chance at a dollar. If you buy NO at 50¢ on that same bracket (which has a 75% true probability of not hitting), your expected value is positive. You're paying 50 cents for a 75% chance at a dollar.
One of the psychological reasons traders resist NO bets is the payout structure. A NO bet at 50¢ risks 50 cents to profit 50 cents. A YES bet at 15¢ on a longshot bracket risks 15 cents to profit 85 cents. The YES side offers lottery-ticket payouts. The NO side offers grinding, consistent returns.
Higher win rate, smaller profit per win, occasional loss: that's the NO profile on overpriced weather brackets. It's the same profile as selling premium in options, laying heavy favorites in sports, or playing a GTO strategy in poker. Over hundreds of bets, the math compounds.
Adjacent brackets at modest prices: If the forecast says 82°F and the 83–84 bracket is priced at 12¢ when the true probability might be 18–20%, YES has positive expected value. Same-day markets where conditions have shifted: If the NWS forecast says 82 but real-time observations show the city running 3°F hot at noon, the 84–85 bracket might still be priced on the old forecast. Model disagreement: When weather models diverge significantly, brackets on either side of the consensus can become underpriced.
Notice the pattern: viable YES bets are almost never on the peak consensus bracket. They're on adjacent or non-consensus brackets at attractive prices.
Kalshi weather markets settle based on the NWS Daily Climate Report. The ASOS station measures in Celsius and converts to Fahrenheit. That conversion introduces rounding steps. All these mechanical factors create settlement noise — the actual reported number can differ from the "true" temperature by up to about a degree.
For the YES buyer on the consensus bracket, settlement noise is an enemy. You need the temperature to land in a specific range. For the NO buyer, any noise that moves the settlement out of the consensus bracket helps you. You win if it moves up, and you win if it moves down.
Buying NO on the consensus bracket isn't contrarian for the sake of being contrarian — it's the mathematically correct response to a market that consistently overprices certainty. The edge isn't huge on any single bet, it's not exciting, and it won't produce stories about "the time I turned $100 into $10K." But over a large enough sample, the math works. And in edge-based trading, the math is all that matters.