A price you see on a Saturday morning has already had a busy week. It started as a number in a model, was shaped by traders, had a margin added, and has been pushed around by other people’s money ever since. Understanding that journey tells you when a price is likely to be wrong.
Step 1: the model
Every large bookmaker runs statistical models. For football, the classic approach rates each team’s attack and defence, adjusts for home advantage, and uses those ratings to estimate how many goals each side should score. From that, a computer works out the chance of every scoreline — and from the scorelines, the chance of a home win, a draw, an away win, over 2.5 goals, both teams to score and so on.
These models are fed with results, shot data, expected goals, player availability and more. They are good. But they are estimates, and they are built on averages. A model does not know that the goalkeeper has flu, or that the away side has nothing to play for.
Step 2: the traders
Human odds compilers review the model’s prices. They add what the model cannot see — injuries, motivation, weather, a new manager — and they make sure the prices line up with the rest of the market. No bookmaker wants to be the only one offering 3.00 when everyone else is at 2.50. Standing out means the sharpest bettors in the world will find you within minutes.
Step 3: the margin goes on
The “true” prices now have the bookmaker’s margin added, so the implied probabilities add up to more than 100%. If the model says a home win is 50% likely, a fair price is 2.00, but the bookmaker might offer 1.90. That gap, across every outcome, is their profit. (Our article on how odds work shows how to measure it.)
Step 4: the market takes over
Once a price is live, money moves it. If lots of people back the home team, the bookmaker shortens its price and lengthens the away price. This is partly about balancing their book, but mostly it is about information. Some bettors are consistently right. When their money arrives, bookmakers treat it as news and move quickly.
This is why prices at the biggest, most liquid bookmakers and exchanges are so accurate by kick-off. Thousands of people, models and opinions have all pushed on them. The final price before the start — the closing line — is usually the best estimate anyone has of what will happen. Beating it consistently is the clearest sign that a bettor has real skill.
Where exchanges fit in
On a betting exchange such as Betfair there is no odds compiler at all. The prices are set by users offering to back and lay each other. The “margin” is just the small gap between the best back price and the best lay price, plus the exchange’s commission on winnings. Because nobody is adding a built-in profit to every price, exchange odds are often better than a bookmaker’s — and they react to news even faster.
Where the mistakes are
If the market is this well informed, how does anyone win? Three places, usually:
- Early prices. When a market first opens, fewer people have looked at it. The price reflects the model and the trader, not yet the crowd.
- Small markets. Lower leagues, minor sports and obscure bet types get less attention and less money. Prices there are lazier.
- Fast-changing situations. Team news, red cards, weather. Whoever updates their view first wins.
The common thread: you win by knowing something the price does not yet reflect, and by proving it over many bets rather than trusting a feeling. That is what backtesting is for.