This post contains affiliate links. We may earn a commission at no extra cost to you.
What Does xG Mean? A Practical Introduction for Bettors
Most punters have no idea why a team lost despite dominating a match. They see 15 shots, only 2 on target, and wonder where it all went wrong. xG — expected goals — answers that question.
xG is a stat that measures the quality of chances a team created or faced. Every shot gets assigned a probability of becoming a goal based on factors like distance, angle, defensive pressure, and whether the player was one-on-one with the keeper. Add those probabilities together and you get expected goals. If a team's xG is 2.3, the stat is saying "based on the quality of chances, this team should have scored about 2-3 goals."
Why should you care? Because bookmakers don't always price matches according to xG. That gap between what the odds suggest and what the underlying data shows is where value lives. A 1.90 favourite that only created 0.8 xG might be overpriced. A 3.50 underdog with 1.6 xG might be the opposite.
In this guide you'll learn:
- How xG is calculated and what it actually measures
- Why bookmakers sometimes miss xG when setting odds
- How to spot value using expected goals data
How Does xG Work in Football Betting?
xG isn't complicated once you understand the core idea. Every shot has a probability of going in. That probability depends on the specifics of where the shot came from.
The Basic Calculation
Imagine a player takes a shot from 15 yards out, central, no defenders blocking the keeper's view. Historical data shows that shots from this exact position go in roughly 18% of the time. That shot gets an xG value of 0.18.
Now imagine a different shot: same player, 25 yards out, at an angle, with a defender's leg in the way. History says shots like this score about 2% of the time. That's 0.02 xG.
A team's xG in a match is the sum of all their shots. If they took five shots worth 0.18, 0.02, 0.24, 0.11, and 0.15 — that's a total xG of 0.70. They created chances worth 0.70 goals, even if they only scored once or not at all.
Why This Matters for Betting
Here's where it gets interesting. Let's say Manchester City hosts Newcastle United. City wins 2-1, and the headlines read "City's dominance pays off." But look deeper: City had 1.8 xG and Newcastle had 1.6 xG. The scoreline flattered City. In a 100-match sample where City had the same xG advantage, they'd win fewer than 100 times.
Conversely, imagine Nottingham Forest plays Arsenal away and loses 1-2 despite creating 2.2 xG to Arsenal's 0.9 xG. The 1-2 scoreline suggests Arsenal were the better team. The xG data contradicts that. Over time, Forest would expect to win more matches where they have a 2.2 to 0.9 xG advantage.
Bookmakers set odds based on a team's recent form, league position, and public perception. They don't always account for underlying performance. A team on a poor run but generating high xG is often underpriced. That's value.
Real Example: Spotting Value with xG
Let's say you're building a Saturday acca. Brighton is hosting Tottenham. The odds on a Spurs win are 1.90.
You check the xG model. Spurs' average xG in their last five away games: 1.3. Brighton's average xG at home: 1.1. The model suggests Spurs have a 52% win probability — that's only a 1.92 fair price. At 1.90, the odds don't offer value; they're roughly fair.
But then you see that Brighton's last two matches saw xG of 2.8 and 2.4 at home, despite losing both. Their underlying performance improved. Spurs' recent xG data shows 0.8 and 0.9 in their last two away games — below their five-game average. The model now suggests 48% win probability for Spurs, which makes 1.90 slightly overpriced.
You skip that bet or look for the Brighton draw at 3.40, which now looks more attractive. That's xG in action.
How Winotips Uses xG in Its AI Model
At Winotips, xG data is one of the core inputs into our AI prediction engine. We use the Dixon-Coles model — a statistical approach that accounts for home advantage, team strength, and the correlation between goals — to forecast match outcomes.
Here's how it works: we feed xG data from recent matches into our model alongside team strength ratings and historical performance. We then run a Monte Carlo simulation 10,000 times per match, generating thousands of possible scorelines. From those simulations, we calculate the probability of a win, draw, loss, and other markets like BTTS and over/under goals.
xG isn't our only input — we also weight fixture difficulty, injuries, and form trends — but it's fundamental. A team generating consistently high xG but getting poor results is flagged as undervalued. The opposite applies to teams riding lucky winning streaks on low xG.
See today's AI predictions on Winotips and compare the odds we identify as value with BestOdds.
How to Use xG in Your Betting
You don't need to become an xG expert to profit from it. Here's a practical approach for UK punters:
Step 1: Find your matches. Look at Saturday's fixture list or midweek games. Pick a match you're interested in.
Step 2: Check recent xG data. Use sites like Understat or Wyscout to see both teams' xG in their last 5-10 matches. Is one team consistently overperforming (high goals, low xG) or underperforming (low goals, high xG)?
Step 3: Compare to the odds. If a team is heavily backed but their recent xG is weak, the odds might be inflated by sentiment. If an underdog has strong xG trends, they might be underpriced.
Step 4: Build your acca with conviction, not just form. A team in good form but with declining xG is riskier than it looks. A team with rising xG but recent poor results is potentially overdue a turnaround.
Step 5: Track your bets. Over time, you'll notice patterns. Bets based on xG outperformance tend to have better expected value than bets chasing headlines.
Frequently Asked Questions
What's the difference between xG and actual goals?
xG is expected goals — a probability-based measure of chance quality. Actual goals are what happened on the day. Over a small sample (one or two matches), they can differ wildly. Over a large sample (a full season), they tend to converge. Our model uses both to predict future outcomes, but no model guarantees results — football is unpredictable.
Can I use xG to predict every match?
xG is a powerful tool, but it's not foolproof. Cup ties and derbies introduce unpredictability — emotional intensity and tactical adjustments matter more in those contexts. xG works best when you're looking at a sequence of league matches where patterns can stabilise.
Which teams in the Premier League generate the most xG?
Typically, attacking teams like Manchester City, Arsenal, and Liverpool generate the highest xG. But a high xG doesn't guarantee wins if the defence is also leaky. Look at both xG for and xG against to get the full picture.
Is xG better than other betting metrics?
xG is useful on its own, but it's most powerful when combined with other data: form, injuries, head-to-head records, and home advantage. Our model integrates xG alongside these factors to identify value. No single metric tells the full story.
Can bookmakers use xG to set their odds?
Some bookmakers use xG data, but not all weight it equally. Major operators build their own models, but they also factor in betting patterns and perceived public sentiment. That's why gaps between xG-based expectations and actual odds exist — and that's where you find value.
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.