When your phone says 30% chance of rain, it’s making a specific claim. The National Weather Service defines probability of precipitation as the chance that at least 0.01 inch falls at a particular point over the forecast period, typically 12 hours. NOAA’s own technical guidance notes the figure doubles as a statement about how much of the forecast area should expect measurable rain. Not a third of the day, and not a third of your city. Most of us have read that number loosely for years, and it still helps us decide about the coat.
Sports odds do the same job in a different notation, and almost nobody gets shown the conversion. Every time you place a sport bet, the price on your screen is already a probability estimate dressed up as a fraction or a plus-minus figure, waiting to be translated back. The American Gaming Association reported that Americans legally wagered $166.94 billion on sport in 2025, and Pew Research Center has found that roughly one in five US adults bet on sport in a given year, most commonly with friends or family. Plenty of us look at prices without ever reading them.
So let’s fix that. The arithmetic first, then how to read a whole board at once, then what happens the moment news breaks.
Every Price Is a Percentage
Take the most common line in American sport: −110 on both sides. That price implies a 52.38% chance for each team, which means the two sides add up to 104.76% rather than 100%.
The extra 4.76% is the listed cost of the forecast, in the way a magazine carries a cover price on the front. You can strip it out in one step: divide each side by the total, and 52.38 becomes a clean 50. People who do this for a living call it devigging, which sounds far more technical than the sum you just did in your head.
Exchange-style platforms skip the step altogether. When a contract trades at 13 cents, that’s a 13% forecast written in plain figures.
Once you can move between the two formats, an odds board stops looking like code and starts looking like what it is: a set of opinions about how likely things are, published by people whose money depends on being roughly right.
The Board Never Adds Up
A full futures market is a probability distribution, and its shape carries information that no league table can express.
Heading into this college football season, Sports Illustrated reported Kalshi’s championship prices with Notre Dame at 13%, Ohio State and Texas both at 12%, Oregon at 11% and Georgia at 10%. Five teams packed inside three percentage points. A preseason poll would rank them one to five and imply a hierarchy the numbers can’t support; the prices tell you the market can barely separate them. Those cent-denominated contracts sit on venues the CFTC has regulated as prediction markets since 2004, which is why the probability is printed for you rather than hidden in a fraction.
The same market laid out Notre Dame’s route as a staircase:
- 86% to reach the playoff
- 65% to reach the quarterfinals
- 50% to reach the semifinals
- 27% to reach the title game
Four coin flips that all have to land, and you can see exactly where the odds thin out.
Zoom out to a whole board and the overround becomes obvious. Using DraftKings’ Super Bowl LXI prices as listed by ESPN on 6 August, the 32 team probabilities sum to around 122.5%. The Rams at +550 imply 15.38% raw, which settles nearer 12.6% once you normalise the board. Arizona at 500-1 comes out around 0.16%.
The price you see quoted and the price you’d pay aren’t identical. That same SI piece describes a $25 position returning $156.23, which implies an execution price closer to 16 cents than 13, and Kalshi’s published fee formula tops out at 1.75% of your outlay around the 50-cent mark. Small gaps, worth knowing about before you commit money.
News Has a Price Tag
Static numbers only tell you where a market sits right now. Watch what happens when information arrives.
After the Rams traded for Myles Garrett, their Super Bowl price moved from +800 to +650, then +600, then +550. In probability terms, the market went from about 11.1% to 15.4% on the strength of one defensive lineman changing address.
Now the more revealing half of that same week. New England’s long-trailed trade for A.J. Brown was completed, and the Patriots’ price didn’t budge from 16-1. Everyone had seen it coming, so it was already in the number. Combat sport shows the same effect from the other direction. A rematch that never gets past the talking stage stays rumour rather than fixture, and a market can only price what looks likely to happen.
That contrast tells you what a price is for. It’s a running tally of everything currently known, updated whenever the known set changes.
Which is also why timing changes reliability. Research on NFL point spreads found closing-line favourites winning 57.19% of the time, with average prediction error falling from 0.0211 at the open to 0.0163 by the close. A separate study across NFL and college football reached the same conclusion on spreads and totals: the closing number is the better forecast.
Meteorologists know this pattern intimately. A verification study in Monthly Weather Review measured same-day rain forecasts at about 36% skill, with performance sliding through the week and effectively vanishing beyond seven days.
So if a number reaches peak reliability an hour before kickoff, what is a July futures price really claiming to know?
A Number That Admits It Might Be Wrong
Calibration is the standard to hold these figures to, not whether any single pick came in. A well-calibrated forecast is one where things priced at 30% happen roughly 30% of the time, and that bar is reachable.
Work on the Iowa Electronic Markets found winner-take-all prices neither over nor under predicting winners across a 100-day horizon, with election-eve forecasts landing within an average of 1.34 percentage points. Weather forecasts hold up well in the middle ranges too, though the same verification study found the low end shakier: a stated 20% chance produced rain only 5.5% of the time on the day. Knowing where a forecast wobbles is part of reading it properly, and it should make you slower to trust a long shot priced at 5%.
More of us will get the chance to practise. Reuters reported roughly $27 billion in Kalshi trading volume across the 2026 World Cup, close to double its projected traffic, and prediction markets took an estimated 27% of US sports-betting volume tied to the tournament, up from around 9% in January. Percentages, quoted directly, no conversion needed.
You already trust a number when a meteorologist hands it to you and tells you it’s uncertain. Give the number beside your team’s name the same reading, and you’ll know what you’re being told.
