Tech, ExplainedGaming, streaming and media
Part

In-play odds

Odds on the next ball, updated live.

INSIDE: SPORTSBOOK SOFTWAREHOW IT WORKS · A BET ON ONE OVER AT LORD'SLive data showsthe 40th overabout to beginThe model gives6 or more runs a40% chance: fairodds of 2.50The bookmakeroffers 2.30, anda bettor stakes£10 to receive £23The first ball goesfor four, and bettingis suspended for afew secondsThe model now gives a75% chance, and theodds reopen at 1.25ALSO TALKS TOSports data↗ DATA PROVIDERSINSIDE: SPORTSBOOK SOFTWAREHOW IT WORKS · A BET ON ONE OVER AT LORD'SLive data shows the 40thover about to beginThe model gives 6 or moreruns a 40% chance: fairodds of 2.50The bookmaker offers2.30, and a bettor stakes£10 to receive £23The first ballgoes for four, andbetting is suspendedfor a few secondsThe model now gives a 75%chance, and the oddsreopen at 1.25ALSO TALKS TOSports data↗ DATA PROVIDERSINSIDE: SPORTSBOOK SOFTWARE · A BET ON ONE OVERAT LORD'SLive data shows the40th over about to begin↓The model gives 6 or more runs a 40%chance: fair odds of 2.50↓The bookmaker offers 2.30, and a bettorstakes £10 to receive £23↓The first ball goes for four, and bettingis suspended for a few seconds↓The model now gives a 75% chance, and theodds reopen at 1.25ALSO TALKS TOSports data↗ DATA PROVIDERS
Inside sportsbook software, live data from a Test match at Lord's moves the odds on 6 or more runs in an over from 2.30 to 1.25 after a four, with betting suspended for a few seconds in between.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

In-play odds are betting odds that a bookmaker keeps updating during a match, using live data, down to the result of the next ball, the next point or the next goal. Betting during a match is called in-play betting, or live betting. The software that sets these odds is the pricing part of sportsbook software.

Odds start from a probability, which is the chance that a result will happen. A pricing model is a program trained on thousands of past matches. It works out the probability of each result from the state of the match. Live sports data gives the model the state of the match, ball by ball, within a second or two of each event. The model turns each probability into odds. Then the bookmaker shortens the odds slightly, so that a winning bet pays a little less than the fair amount. This difference is the bookmaker's margin.

Imagine a Test match at Lord's, in London, at the start of the 40th over. A bookmaker offers a bet on that over: will the batting side score 6 runs or more? The model looks at the score, the batters, the bowler and the age of the ball, and gives this a 40% chance. Fair odds would be 1 ÷ 0.40, which is 2.50. The bookmaker offers 2.30 instead, to keep its margin. A bettor in London stakes £10 at 2.30. If 6 or more runs come in the over, he receives £23.

Then the first ball of the over is hit for four. The model now gives 6 or more runs in the over a 75% chance. While the new data is arriving, the bet is suspended for a few seconds, and no bets are accepted. Otherwise, a person at the ground, who saw the ball before the data arrived, could bet on a result that had already happened. When the bet reopens, the odds have fallen to 1.25. For the same reason, in-play bets are held for a few seconds before they are accepted.

A big match can have hundreds of such bets open at once, all priced by the model. So the speed of the data feed matters more than almost anything else. The bookmaker also limits how much it accepts on each result, through sportsbook risk management.

Fast, repeated betting during a match is a known risk for problem gamblers. So regulators watch in-play betting closely, under the rules of responsible gambling. In-play betting on sport is legal in Britain. In India, all online betting is banned, and the Online Gaming Act 2025 bans every online game played for money.

Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.