The popular image of an odds compiler is a man with a cigarette and a form book, scribbling prices on a whiteboard. That world existed, roughly, until the late 1990s. The modern equivalent is a trading floor at a firm such as bet365 in Stoke-on-Trent or Flutter’s operations in Dublin and Leeds, where quantitative models generate thousands of prices per second and human traders intervene only when something looks wrong.
Where an opening price comes from
Every market starts with an estimate of true probability. For a Premier League match, the base model typically combines long-run team ratings, expected-goals data from suppliers such as Stats Perform, squad news and home advantage. The output is a set of percentages: perhaps 48 per cent home win, 27 per cent draw, 25 per cent away win. Converted to decimal odds, those figures would be 2.08, 3.70 and 4.00. No bookmaker publishes them at that level. A margin is layered on top, shortening every price so that the book adds up to more than 100 per cent, a mechanism explained in detail in the piece on betting fundamentals.
Not every firm originates its own numbers. Price origination is expensive, so many smaller operators licence odds feeds from suppliers such as Sportradar or Kambi and simply apply their own margin. The industry distinguishes between price makers and price takers, and the takers vastly outnumber the makers.
The market as a correction machine
An opening price is a hypothesis. The moment money arrives, it gets tested. Low-margin bookmakers such as Pinnacle, founded in Curaçao in 1998, built their entire model on this idea: welcome sharp bettors, let their money reshape the line, and treat the resulting price as the best available estimate of the true probability. Academic studies, including work published in the Journal of Prediction Markets, have repeatedly found that closing prices at sharp books are more accurate than opening prices almost everywhere.
Betting exchanges accelerated the process. Betfair, launched in London in June 2000, lets customers bet against each other, and its prices respond to information within seconds. Many trading rooms watch exchange movement as a live input. If Betfair’s price on a tennis player drifts sharply during a rain delay, fixed-odds books will usually follow before their own customers can exploit the gap.
Liability, not prediction
A common misconception is that traders are trying to predict results. Mostly they are managing liability. If a book takes £400,000 on one side of a market and £50,000 on the other, an unbalanced result could be expensive, so the price on the popular side shortens to slow the flow and the other side lengthens to attract it. On enormous markets such as the Grand National or a World Cup final, the recreational money is so heavily skewed towards favourites and famous names that books frequently carry large positions rather than balancing perfectly.
Individual customers are part of the calculation too. Trading rooms profile accounts by the sharpness of their bets, measured against the closing line. An account that consistently beats the close will find its stakes restricted, a practice that drew scrutiny from the UK Gambling Commission and from a 2021 report by the House of Lords select committee on gambling, though it remains legal in Britain.
In-play: the hardest desk in the room
Live markets multiplied the workload. A single football match now carries hundreds of in-play markets, repriced after every corner and card. The models run on official data feeds; Genius Sports paid a reported £120 million in 2019 for rights to official Football DataCo data partly because a few seconds of latency is the difference between a fair price and a losable one. Human traders still suspend markets manually around penalties, VAR checks and red cards, moments when the model’s assumptions break down.
The result of all this machinery is a price that looks simple on screen and is anything but. It encodes a statistical model, a margin, a liability position, the behaviour of the sharpest customers in the market and, increasingly, the movement of every rival book. Readers new to the mechanics may want to start with the sports betting basics guide before digging into how those prices are shaded, shifted and occasionally got badly wrong.