This post contains affiliate links. We may earn a commission at no extra cost to you.
What is xG and Why Does It Matter for Betting?
Most bettors chase results. Smart ones chase underlying quality. That's where xG comes in. xG — expected goals — is a stat that measures the quality of a shot based on where it came from, the angle, and the defensive pressure. It doesn't care if the shot went in; it only cares if it should have.
Here's why UK punters need to understand this: bookmakers set odds on actual outcomes — goals scored, matches won, clean sheets kept. Your model should be hunting for the gap between what actually happened and what the data suggests should happen. That gap is where value lives.
Think about a match where one team battered the other but lost 1-0. Frustrating to watch, but for betting purposes? That tells you something. The losing team probably had better chances. Their xG was likely higher. Next time they play, the market might still be scared from that loss result, even though the underlying performance suggests they're the stronger team.
In this guide you'll learn:
- How xG is calculated and what it actually measures
- How to spot value using xG data in your Saturday acca or midweek bets
- How Winotips uses xG within its AI model to identify edge
How Does xG Actually Work?
xG assigns a probability to every shot, based on historical data. A penalty from six yards out gets an xG value around 0.79 — meaning that type of shot goes in roughly 79% of the time. A speculative 30-yard effort gets 0.02. Miss the net entirely? You still get an xG value, because the quality of the chance is what matters, not the result.
When you add up all shots in a match, you get team xG totals. A team with 2.3 xG created 10 shots worth roughly 2.3 goals on average — even if they only scored once.
Why xG Beats Just Counting Goals
Goals are noisy. A world-class striker slots three tap-ins; a backup forward squanders two golden chances. Same xG, wildly different quality. Over 30 matches, though, xG and actual goals converge. Teams that consistently underperform their xG get lucky. Teams that overperform it are usually playing above their level.
Manchester City, for example, typically have the highest xG in the Premier League and also the highest goals. But dip into individual matches — City might create 3.1 xG and score four goals. That's overperformance. The market might favour them next week, but the underlying chances weren't quite worth four goals. Value for their opponents might be there.
xG on Defence: Building Clean Sheet Value
xG works both ways. If Arsenal face Newcastle and Newcastle's shots create only 0.8 xG, Arsenal's clean sheet odds might be undervalued. Bookmakers sometimes rely on team reputation ("Newcastle will attack"); the stats say otherwise ("Newcastle didn't create much quality last time").
Cup ties throw this into sharper relief. A Championship team holding a Premier League side to 1.1 xG? That's a potential upset signal. Not guaranteed — football is unpredictable — but the underlying quality suggests the underdog's price might be too long.
How Winotips Uses xG in Its AI Model
Our model consumes xG data alongside team form, head-to-head records, player availability, and home/away splits. We run 10,000 Monte Carlo simulations per match using a Dixon-Coles framework, feeding xG as a key input to predict likely scorelines.
Here's the practical bit: see today's AI predictions on Winotips. We compare our model's xG-informed win probability against bookmaker odds. When we identify a gap — say our model gives a team 58% to win but they're priced at 2.2 (45% implied) — that's a signal. You can then compare odds at BestOdds to find the sharpest price.
xG tells us about process. Odds tell us what the market thinks. By combining both, we're not guessing on individual matches — we're identifying where bookmakers' short-term biases diverge from what consistent chance quality predicts over time.
How to Use xG in Your Betting
You don't need to calculate xG yourself — understat.com and fbref.com publish it freely. Here's how to apply it:
- Check xG for teams you're considering. Building a Saturday acca? Look at each side's xG from their last five matches. If a team's actual goals are way above their xG average, be cautious. Regression is likely.
- Compare team xG to league average. Premier League average is roughly 1.4 xG per team per match. Teams consistently above or below tell you something about their attack and defence quality.
- Look at xG trends, not single matches. One bad xG performance means nothing. Three matches where a team creates 0.6, 0.8, and 0.7 xG? That's a pattern. Their "should be scoring more" narrative is probably false.
- Use xG for underdog accas. If a lower-division side held their last opponent to 1.0 xG, their odds to keep a clean sheet against a similar-level team might be shorter than the underlying chances support. That's potential value in your midweek acca.
- Watch xG on draw markets. Matches where both teams' xG are close often end level. If you fancy a draw at 3.5, check whether both teams have created similar quality shots lately. High xG for one side and low for the other? Draws are less likely than the odds suggest.
Frequently Asked Questions
Does High xG Always Mean a Team Will Score?
No. xG is a quality metric, not a guarantee. A team with 3.5 xG might score four goals, or they might score two. Over 30 matches, though, xG and actual goals align. Short-term variance is real — that's why football is interesting and why odds have value.
Can I Use xG for Live Betting?
You can, but with caution. Live xG updates during matches, but bookmakers adjust odds faster than stats can update. If you're using xG to spot midweek value, do your homework before kickoff, not during it. Our model can help identify value, but no model guarantees results — football is unpredictable.
What's the Difference Between xG and Expected Assists (xA)?
xA measures the quality of chances a player created, not the quality of shots they took. A pass that leads to a 0.5 xG shot registers as 0.5 xA, whether the shot goes in or not. Together, xG and xA tell you about a team's attacking pattern — who's creating, who's finishing.
Should I Only Bet Teams With High xG?
Not exclusively. Teams that overperform xG exist — some strikers are clinical finishers, some goalkeepers are catastrophic. But over a full season or a strong sample, xG reveals truth. Use it as one piece, not the whole puzzle. Combine xG insight with fixture difficulty, player form, and head-to-head patterns.
Is xG the Same Across Different Websites?
No. Different models calculate xG slightly differently. Understat, StatsBomb, and FBRef might give the same shot an xG value of 0.15, 0.18, or 0.13. The differences are small, but they exist. Pick one source and stick with it so you're comparing apples to apples.
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.