This post contains affiliate links. We may earn a commission at no extra cost to you.
What's the Deal with xG? A Quick Introduction
Most punters judge a team's performance by the final score. That's backwards. A team can lose 2-1 but create far better chances than the opposition. That's where xG — expected goals — comes in. xG quantifies the quality of chances both teams create, independent of whether they actually scored.
If you're serious about finding value in the odds, you need to understand xG. Bookmakers price matches on sentiment and money flow. Your model should price them on actual shot quality. That's the gap where profit lives.
In this guide you'll learn:
- Exactly what xG is and how it's calculated
- Why xG matters more than you think when betting on football
- How to use xG data to spot value in real matches
What is xG? How Does It Actually Work?
Expected goals is a statistical model that assigns a probability to every shot taken in a match. A shot from the penalty spot might have an xG value of 0.79 (79% chance of scoring). A long-range effort from 30 yards might have an xG of 0.02. Every shot gets a score based on historical data: distance from goal, angle, defensive pressure, whether it's a header or a foot, and dozens of other factors.
Add up all the xG from a team's shots, and you get their total expected goals for the match. If Arsenal creates 2.8 xG and Man City creates 1.1 xG, the stats suggest Arsenal had the better match — even if the scoreline says otherwise.
Why does this matter to you as a bettor? Because scorelines lie. Over a season, teams regress to their underlying performance. A side that creates high xG but loses 2-1 will eventually start winning those matches. That's where value appears.
How is xG Calculated?
There's no single xG model — different providers use different methodologies. StatsBomb, Understat, and Opta Sports all calculate xG slightly differently. But the principle's the same: machine learning models trained on millions of historical shots.
The model looks at factors like:
- Distance from goal
- Angle of the shot
- Type of shot (header, left foot, right foot, etc.)
- Number of defenders between shooter and goal
- How much time the shooter had to set up
- Whether it's from open play or a set-piece
A shot taken from inside the box, head-on to goal, with minimal defenders nearby might get 0.35 xG. The same shot from a tighter angle drops to 0.12. Historical data shows which situations lead to goals most often.
Why xG Matters More Than Goals Alone
Goals are random in the short term. A clinical striker puts away 1 in 3 chances. A wasteful one scores 1 in 8. Over 38 Premier League matches, that variation evens out — but over a single weekend fixture, it's chaos.
xG cuts through that noise. If Tottenham creates 2.1 xG and Brighton creates 0.8 xG, Spurs deserved to win. If Brighton actually nicked it 1-0, that's an anomaly, not a pattern. Next time these teams play similar opposition under similar circumstances, the xG team usually comes out on top.
Here's a concrete example: Arsenal at home to Fulham. The bookies price Arsenal at 1.75 to win. Your gut says that's tight. But pull the xG data: in their last five home matches, Arsenal averaged 2.3 xG per game. Fulham's last five away matches show 0.9 xG conceded on average. The stats suggest Arsenal should be shorter than 1.75. That's value.
How Winotips Uses xG in Its AI Model
At Winotips, xG is one of the core inputs in our prediction engine. We don't rely on it alone — that would be naive. Instead, we feed xG data alongside form, head-to-head records, injuries, and player ratings into our Dixon-Coles model, a Bayesian framework that's proven especially effective at predicting football matches.
For every fixture, we run 10,000 Monte Carlo simulations. Each simulation uses the underlying probabilities from xG, team strength, and historical patterns to generate a potential match outcome. After 10,000 runs, we know the probability distribution: what percentage chance of a home win, draw, or away win, and the likely goals distribution.
We then compare those probabilities to the odds offered by the market. If our model says Manchester City have a 68% chance of beating Newcastle, but the odds imply only 54%, that's a value signal. See today's AI predictions on Winotips to watch this process in action.
xG isn't the whole story — a team's underlying xG can shift suddenly due to injury, managerial change, or tactical adjustment. But it's a better starting point than sentiment or betting volume. When you combine it with our broader model, you're comparing probabilities to prices, not gut feel to hype.
How to Use xG in Your Betting
Right, enough theory. How do you actually use this on a Saturday afternoon when you're building an acca?
- Check the underlying xG for both teams. Before you look at the odds, pull up Understat or StatsBomb and see what xG both teams average. If you're backing a team to win, they should be creating more expected goals than their opponent — at least 0.5 xG more if you're being picky.
- Compare xG to the odds. If your data shows Team A should be 55% to win but the odds only give them 48%, that's value. Use BestOdds to shop for the sharpest price, then move.
- Look for high xG in cup ties. In a one-off cup match, variance matters more. The team with higher xG doesn't always win. But in league matches where teams play 38 times? xG is gold. Focus your value hunting on Saturday Premier League fixtures where sample size is largest.
- Don't chase single shots. A team might score 1 but create 0.2 xG (a lucky scrappy goal). The next match, they'll create 2.3 xG but lose 1-0. That regression is predictable. Betting on underlying performance, not results, smooths out that chaos.
- Combine xG with other metrics. xG alone doesn't tell you everything — a team might create high xG but be missing their top striker. Check team news, suspension lists, and recent form. xG is one piece of the puzzle, not the entire picture.
Frequently Asked Questions
What does 2.5 xG actually mean for a football team?
It means that team's shots, when evaluated across all their chances, added up to 2.5 expected goals. On average, based on historical shooting patterns, they'd score 2-3 goals from that performance. They might have scored 0, 1, or 4 — football's unpredictable. But 2.5 xG is a fair reflection of how clinical they were and how lucky they got.
Is high xG always a good sign when betting?
Not always, no. A team can create high xG and still lose due to poor finishing or a goalkeeper playing out of their skin. But over multiple matches, high xG correlates strongly with wins. Use xG as one signal among several — form, injuries, head-to-head — rather than a guarantee. Our model can help identify value, but no model guarantees results. Football is unpredictable.
Should I focus on xG for goals or BTTS betting?
Both teams to score (BTTS) is actually one of the best markets for xG bettors. If both sides are creating decent xG (0.8+), BTTS becomes attractive at shorter odds. Check both teams' attacking xG and defensive xG separately. If one team attacks well but defends poorly, and the other does the reverse, BTTS value appears fast.
Can xG help me find value in midweek cup ties?
Midweek cup matches are trickier because variance is higher — smaller sample size, teams might rotate, fatigue plays a role. xG still helps, but don't weight it as heavily as you would in a Saturday Premier League match. In a one-off cup tie, a lucky goal or dodgy penalty call matters more than underlying quality.
What's the difference between xG and shot-on-target ratio?
Shot-on-target is crude: did the ball hit the goalkeeper? xG is sophisticated: it evaluates the quality of every shot, on target or not. A team might have more shots on target but lower xG if those shots were from poor positions. xG is the more reliable predictive metric for value hunters.
18+ | Please gamble responsibly. Betting should be entertaining, not a way to make money. Free help: BeGambleAware.org | GamStop.co.uk | GamblingTherapy.org
Winotips provides predictions for informational purposes only. We do not guarantee any results. Always bet within your means.