Right. Let's talk about expected goals, because if you're making football bets without understanding xG, you're essentially flying blind.
For years, people looked at goals and nothing else. Team A scored 3, Team B scored 1, so Team A won. End of story. But that's surface-level nonsense. A team can lose 1-0 and still have dominated the match. Another can win 2-1 despite being utterly outplayed. The real question isn't "how many did they score?" It's "how many should they have scored?" That's where expected goals comes in.
We're seeing this play out right now at the World Cup 2026. Brazil looked shaky in their group stage opener—won 2-1 against Serbia despite generating xG of 1.8. That's not a dominant performance. Meanwhile, Spain's 3-0 win over Mexico came with an xG of 3.2. Two entirely different stories, same result column. Understanding the difference is where your edge is.
What Is Expected Goals (xG), Actually?
Expected goals is a statistical model that assigns a probability to every shot taken. It asks: based on historical data, what's the likelihood this shot becomes a goal?
A penalty kick? That's worth around 0.79 xG (about 79% conversion historically). A tap-in from three yards out? Maybe 0.6. A speculative 30-yard effort from a tight angle? Could be 0.02. Add up all the shots in a match, and you get a team's total xG.
The model factors in distance from goal, angle of approach, defensive pressure, whether it's a header or a foot, and what happened before the shot. Different providers use slightly different algorithms—StatsBomb, Understat, Wyscout—but they're all measuring the same thing: shot quality.
Here's the crucial bit: xG isn't about "luck" or "the team were unlucky." That's lazy analysis. xG tells you whether a team's underlying performance matches their results. If a team's been scoring 2.5 goals with 1.2 xG, they're punching well above their weight. Unsustainable. On the flip side, a team with 2.8 xG generating only 1 goal? They're due a correction—regression to the mean is coming.
Expected Goals (xG) vs. Shots: They're Not the Same Thing
This is where people get confused. A shot is just that—someone kicked the ball toward goal. xG is quality-weighted.
Arsenal might take 18 shots in a match. Liverpool takes 12. On raw shot count, Arsenal looks dominant. But if Liverpool's 12 shots average 0.15 xG each (0.15 × 12 = 1.8 xG) and Arsenal's 18 average 0.08 (0.08 × 18 = 1.44 xG), Liverpool actually created the better chances. Arsenal was wasteful—lots of noise, no quality.
This matters for betting because bookmakers and casual punters still obsess over shot count. They see 15 shots and assume dominance. You see 15 low-quality shots and know to bet the other team. Germany's showing us this right now at the World Cup—they're playing a possession-heavy game but their xG generation has been moderate at best. Teams pressing higher are creating cleaner, quicker chances with better conversion rates.
When you're assessing a team's form, ignore the headline number. Dig into xG. It's the difference between a team that got lucky and a team that's actually good.
Why xG Matters for Your Betting Picks
xG is a predictive metric. That makes it gold for bettors.
If a team's been winning 2-1 with xG of 1.1 vs. 2.2, they're getting beaten in the process. Their results are running hot. Back the opponent next time—the data says regression is coming. Over the next three matches, that team's going to drop points. You're not making a gut call; you're reading the underlying performance.
Conversely, a team losing 0-1 with xG of 2.8 vs. 0.6? They're getting screwed by poor finishing or a goalkeeper in form. They're the side to back going forward. The odds will still have the other team as favourites based on the result, but you know better.
Let's look at France at this World Cup. They scraped through their group stage 2-1 with xG of 2.3. On paper, a winning performance. But they've looked laboured going forward, which the xG reflected. Now they're in the knockouts where margins are tighter. A team that's been generating just above 2 xG per match can't afford to convert at 100%—they won't. France at 3/1 to win their next knockout? That's overpriced. There's money elsewhere.
The practical application: before you place a bet on a match, check both teams' xG from their last 3-5 games. Is one team's score belying strong underlying performance? Has the other team been lucky? That's your angle. Bookmakers price games on recent results. You're pricing them on what actually happened on the pitch.
A specific tip: Any time you see a team with a 3-1 or 4-1 win and their xG is below 2.0, they've fluked it. Fade them next match. The model is brutal about this kind of thing.
The Catch: xG Isn't Perfect
It's not. xG assumes historical conversion rates hold—but they don't always. A world-class finisher like prime Lewandowski will exceed his xG consistently. A dreadful striker will fall short. xG doesn't account for individual quality at the extreme ends.
It also struggles with live, dynamic situations—a goalkeeper's positioning, a defender's reaction speed—things that happen too fast for pre-shot data to capture properly. Different providers weight these factors differently, which is why Understat's xG for a match might be 0.3 higher than StatsBomb's.
Use xG as a framework, not gospel. Combine it with form, team news, head-to-head records, and actual tactical observation. A team with superior xG but a knackered centre-back out injured? That edge disappears. xG is one lens, not the only one.
Where xG Leaves You: The Takeaway
Expected goals strips away the noise of results and shows you what actually happened. A team can win 1-0 on a worldie and an own goal while generating 0.4 xG. Another loses 0-1 but should've scored 3. The xG tells the truth; the scoreline lies.
For betting, that means you're no longer just reacting to what happened last weekend. You're predicting what's going to happen based on underlying performance. Teams regress to their xG. Bad shots don't keep going in. Good teams eventually break through.
Check xG before every bet. Make it part of your routine. The bookies aren't factoring in these nuances as thoroughly as they should—that's where your edge sits.
Responsible Gambling: Betting involves risk. 18+ only. If gambling is affecting you, call the National Gambling Helpline free on 0808 8020 133 or visit BeGambleAware.org.