How to Read Expected Goals (xG) for Smarter Soccer Bets
Expected goals measures shot quality, not just shot quantity. Understanding it changes how you read a game.
I buy a lottery ticket once a week. It costs two dollars. For seventy-two hours, it buys me the right to imagine a different life. The probability is meaningless. The imagination is real. That's the economic transaction.
Soccer betting is similar but inverted. Instead of buying imagination with money, you're selling imagination in exchange for money. The better you understand what you're actually watching, the more accurate your sale price becomes.
Expected goals (xG) is a tool that measures this gap. It quantifies what you see versus what the numbers say.
What Expected Goals Actually Measures
Expected goals assigns a probability to each shot. If a striker is one yard from goal with the goalkeeper out of position, that shot has an xG value of 0.8 or higher. The striker has an 80 percent chance of scoring from that position, historically. If the same striker is twenty yards out, shooting through traffic, the xG might be 0.05. A five percent chance.
Over a full match, xG is the sum of all those probabilities. If a team generates ten shots with an average xG of 0.10 per shot, their total xG is 1.0. That means, on average, they should have scored one goal. If they actually scored two, they got lucky. If they scored zero, they got unlucky.
The gap between xG and actual goals is sometimes luck. Sometimes it's execution. Usually it's both.
Fortunately, the luck smooths out. Over a season, a team's actual goal total will converge with their xG total. That convergence is where betting edge lives.
How xG Shapes Betting Decisions
A team dominates possession and territory. They create chances. But the actual score is 0-1. The losing team has one shot from 20 yards. Their xG is 0.04. The winning team generated eight shots with a combined xG of 1.2.
Betting moneyline on the team that dominated territory but lost? Risky. They're likely the better team. But today, they lost. Betting them to win next week? Reasonable. The underlying performance suggests they're superior.
Betting over/under becomes interesting when xG is involved. If Liverpool generates 2.8 xG against a team and Everton generates 1.1 xG, the xG total is 3.9. That's close to four goals. If the over-under is set at 2.5, the under looks mathematically loose. The xG suggests more goals will be scored.
This works because xG is stable and predictive. A team that consistently generates 1.8 xG per match will, over 38 matches, produce roughly 68 goals. The standard deviation tightens with sample size. After 10 matches, xG is noisy. After 38 matches, it's accurate.
Practical Application: Liverpool vs Tottenham
Liverpool 2, Tottenham 1. Possession: 62 to 38 percent. Shots: Liverpool 18, Tottenham 7. Liverpool's xG was 2.1. Tottenham's was 0.9. On paper, Liverpool should have won. They did win. The xG aligned with the outcome.
Now imagine the same match with a different score. Liverpool 1, Tottenham 2. Everything else unchanged. Liverpool created chances worth 2.1 goals. They scored one. They underperformed. Tottenham's shots were worth 0.9 goals. They scored two. They overperformed.
Betting Liverpool to win next week? Their underlying performance says yes. Betting against them because they lost today? That's extrapolating noise.
Betting total goals market? The combined xG was 3.0. If the total was set at 2.5, the under won, but the underlying probability favored the over. That's a loss made from a correct premise.
Why This Matters
xG is not perfect. Goalkeepers matter. Set-piece execution matters. Some teams score more penalties than their underlying play suggests. Some teams miss more.
But over time, xG is more predictive than gut feeling. A coach watching a game might think the dominant team deserves a win. xG says the dominant team's chances were only worth 1.2 goals and they actually scored none. The coach's intuition is wrong.
The edge in soccer betting comes from identifying when the market's probability estimate (reflected in the odds) diverges from xG's probability estimate. If xG suggests a 65 percent chance of over 2.5 goals and the market prices the over at minus 110 (52 percent probability), you have value on the over.
That value doesn't guarantee a win today. But across a season, betting value at reasonable odds produces profit. The lottery ticket buys imagination. The xG bet buys evidence. I prefer the evidence, even when imagination is more fun.
