Road to Punter Series: What is xG in Football Betting?

Hey, it’s Liz from GoalBible, and today I need to talk about one of those terms that gets thrown around in football Twitter threads like it’s some secret code. You’ve seen it: xG. Expected goals. People love dropping it into conversations to sound clever. But what xG actually is in football, and why should you care? If you’re betting, it might be the most useful number you aren’t using yet.

xG (Expected Goals) Explained

Let’s strip the jargon away. I treat xG as a simple reality check. It counts how many goals a team should have scored from the opportunities they produced, full stop. It’s about what probably should have happened. Every shot gets an xG value between 0 and 1. An xG of 0.6 means you’d expect it to be scored 60% of the time. 

Here's why I care about this for betting: the final scoreline is often a terrible indicator of who actually dominated a match. A team can lose 1-0 while generating an xG of 2.4. That's not a bad team — that's an unlucky one. And unlucky teams tend to correct themselves. Spotting that before the bookmakers adjust the odds is where the value lies.

xG for Different Situations

xG on its own is solid, but the real insight comes from the extended family of metrics. Here's what I actually look at:

xGA (Expected Goals Against) measures the quality of chances a team concedes. A team might have a clean sheet run, but if their xGA is quietly ballooning, their defense is holding on by a thread. That matters going into a tougher fixture.

xG Open Play vs. xG Set Play breaks down where the chances are coming from. Some teams are clinical from open play but basically gift-wrap set pieces to opponents. Knowing this matters a lot for niche markets like first-goal-scorer or corner-related bets.

Non-Penalty xG (NPxG) strips penalties out of the equation. Penalties skew everything — they're high-probability, high-xG events that don't reflect typical match patterns. NPxG gives a cleaner read on a team's actual attacking quality.

Expected Goals on Target (xGOT), also called post-shot xG, goes a step further by factoring in where the shot actually landed. A low-xG attempt fired into the top corner gets more credit than a point-blank effort comfortably saved by the goalkeeper. This is the metric I use most when evaluating strikers and goalkeepers.

How GoalBible Analyses a Match With xG

You don’t need a calculator or a maths degree to use xG stats. I look at them the same way I’d look at a restaurant menu: scan for what stands out, ignore the fluff. Let me show you a real match example so you can see what I do.

Manchester City hosted Liverpool. Here’s what the xG numbers looked like:

Liverpool posted 0.71 xG to City’s 0.65. Close but clear. City had the edge in Open Play xG (0.58 vs 0.32). But Liverpool owned Set Play xG (0.38 vs 0.07), Expected Assisted Goals (0.50 vs 0.40), Non Penalty xG (0.71 vs 0.65), and xGOT (a staggering 1.38 vs 0.41). That xGOT gap told me Liverpool’s shooting was far more threatening and accurate, while City’s efforts were easier for the keeper to handle.

When one side leads five out of six meaningful categories, I take notice. The match finished 2-0 to Liverpool on December 1, 2024. The xG data pointed directly to a Liverpool win, a clean sheet, and multiple goals. That’s three potential betting angles, all backed by logic, not guesswork.

How Expected Goals Are Calculated

Different data providers use slightly different models, but the smart ones all work from the same idea. Analysts review hundreds of thousands of shots and measure the key variables. Close-range shots get better xG. Tighter angles get worse xG. A shot on the weak foot? That lowers the probability compared to a strong foot hit. Headers, volleys, rebounds, defensive pressure, and keeper positioning: all of it feeds into the number.

The models don’t always agree perfectly. One system might give a penalty of 0.76 xG, another 0.81. As long as you use a consistent, reliable source, the overall picture holds. Think of each value as a percentage, and the whole thing suddenly stops feeling abstract. 

Liz reminder: read the percentages, question the narratives, find the edge.

Team xG vs Player xG

Football expected goals gets split into two buckets. Team xG tells you the overall attacking threat of the side based on the quality of chances they produce collectively. Player xG zooms in on an individual’s opportunities. Both matter, and together they cover a broad range of markets.

If I’m betting on anytime scorers, I care about player xG and recent shot locations. If I’m looking at the match result or total goals, team xG and xGA are my starting point. The combination is what separates a vague hunch from a well-reasoned bet.

Team xG Examples

Now, let’s look at a few clubs to see how team xG paints a fuller picture.

Liverpool's xGD of +34.90 is stratospheric. They're also slightly over-performing in attack (64 goals vs. 59.40 xG), but the gap isn't unsustainable — their underlying quality justifies confidence in Premier League outright markets.

Nottingham Forest is a different story. Their goals (44) are running well ahead of their xG (33.90), which is a significant over-performance. Their defensive numbers look sustainable, but the attacking efficiency is going to regress at some point. Fade them in big price matches.

Manchester United are underperforming their xG in attack and conceding more than their defensive xGA suggests they should. That's a squad that's not converting chances and is getting unlucky against. I'd normally say regression to the mean helps them — but given the underlying structural issues at the club this season, I'd want more evidence before trusting it.

Player xG Examples

Let’s put some names to the numbers. I’ve pulled a snapshot of three top forwards. Here's where it gets interesting. Take this snapshot from the 2024/25 Premier League season:

Salah is outperforming his xG by 5 goals. Isak by nearly 4. Both are on exceptional runs, but my recommendation is to be cautious about backing either to maintain that rate across a full season. Consistent xG over-performance is rare. At some point, the law of averages taps you on the shoulder.

Haaland, on the other hand, is essentially matching his xG. Nineteen goals against an xG of 18.30 is as clean a signal as you'll find. He's not getting lucky, he's just ruthlessly efficient. For me, that's the more dependable profile to back for goal-scorer markets.

Where to Find xG Stats

Knowing how to read xG is one thing. Actually getting your hands on reliable numbers is another. The good news? You don't need access to some secret database or a subscription that costs more than your betting bankroll.

Most serious bettors I know pull xG data from the usual suspects: FBref, Understat, WhoScored, and Sofascore. These sites give you team and player expected goals, xGA, xGOT, and all the offshoots I've banged on about. FBref is my personal go-to when I want granular shot data. Understat keeps things clean if you just need a quick look at xG and xGA for a league match. Sofascore works nicely on mobile when I'm checking something last minute before kick off.

A word of advice: pick one source and stick with it for consistency. Different models calculate xG slightly differently, and hopping between three different sites will just confuse you.

Once you've got your stats sorted, the other half of the equation is having a decent place to actually put your bets. Our bookmaker list on GoalBible breaks down what each operator actually offers, so you can match the right platform to how you bet. If you want fast crypto payouts, Roobet and BC.Game are two names worth checking. Both handle withdrawals quickly and keep things simple if you're betting with digital currency. For broader market coverage and a more traditional setup, 1Win and BetOnGame are solid shouts. 1Win covers a huge range of leagues if you're digging into xG data from less obvious competitions, while BetOnGame keeps things straightforward for casual bettors who don't want a cluttered interface.

Finding an edge with xG is great, but it's pointless if you're stuck with an operator offering poor prices or taking ages to process your withdrawal. Stats in one tab, a sharp bookmaker in the other. That's the setup I recommend.

GoalBible's Final Say: From xG Nerd to Sharper Bettor

xG won't turn you into a betting genius overnight. But if you're serious about finding value instead of guessing, expected goals belong in your toolkit. Start small: pick one league, compare xG to actual results each matchday, and note who keeps overperforming or underperforming. Patterns appear fast. Combine what you find with a sharp operator from our GoalBible bookmaker reviews, grab a promotion that actually adds something, and you're already doing more than most punters who skip the homework entirely.

FAQs

1. Is a higher xG always better?

Generally, yes. A higher xG means a team or player created better-quality chances. But xG is most useful when compared against actual goals scored. 

2. Does xG work for all football leagues, or just the top ones?

xG models are generally most accurate in top-tier leagues like the Premier League, La Liga, or Bundesliga, where the largest shot datasets exist. For lower leagues, the sample sizes are smaller, and model accuracy can dip.

3. Does xG account for goalkeeper quality?

Standard xG does not; it evaluates the chance itself, not the keeper facing it. That's where xGOT (Expected Goals on Target) comes in. If a goalkeeper consistently saves shots that xGOT says should go in, that's a measurable sign of above-average performance, not just luck.

4. Can xG be used for in-play betting?

Yes, and it's one of the more underrated applications. If a team is trailing 1-0 but is generating significantly higher xG in real time, the live odds may not fully reflect how the match is actually playing out. That gap between perception and probability is where sharp in-play bettors look.

5. Is xG useful for predicting draws?

Indirectly, yes. When two teams show very similar xG figures heading into a match — and both have histories of close xG margins — the draw deserves more consideration than the odds might suggest.