This post contains affiliate links. We may earn a commission at no extra cost to you.
Why xG Matters More Than You Think
Most UK punters lose money because they judge matches by final scorelines, not by what actually happened. A team can dominate and lose 1-0. Another can scrape a 2-1 win with three shots on target. Which team was actually better? xG tells you.
xG (expected goals) is a statistic that measures the quality of chances created in a match. It's not about luck or final scores — it's about underlying performance. For betting, that matters enormously.
If you're building Saturday accas or hunting value in midweek fixtures, understanding xG gives you an edge bookmakers don't always price correctly. You'll spot matches where the underdog actually played better. You'll catch overpriced favourites who won despite being outplayed.
In this guide you'll learn:
- How xG is calculated and what the numbers actually mean
- Why bookmakers sometimes get xG wrong (and how you can profit)
- Three practical ways to use xG in your betting strategy
How Does xG Work in Football?
xG assigns a probability to every shot taken in a match. A penalty kick might have an xG value of 0.79 — meaning penalties go in roughly 79% of the time. A long-range effort from 30 yards might be 0.02. A tap-in from three yards? Close to 1.0.
The model considers dozens of factors: shot distance, angle, defensive pressure, whether it's on the left or right foot, if it's a header or with the boot, how much space the player had. All of this gets processed to produce a single number: the probability that shot should've resulted in a goal.
Add up every shot in a match, and you get the team's total xG. If Manchester City played Arsenal and finished with 2.4 xG versus 0.8, City created better chances. They outplayed Arsenal. The actual score might've been 1-1, but the underlying story is clear.
The Difference Between xG and Actual Goals
Here's where betting value lives. Over one match, xG and actual goals can wildly disagree. Over 10 matches, they converge. Over a season, they're virtually identical.
Think of it like coin flips. One flip might give you heads five times in a row. Flip 1,000 times and you'll get roughly 500 heads. Same principle.
A striker who scores two goals from 0.4 xG is extremely lucky — and unlikely to repeat it. A striker who scores four goals from 5.2 xG is performing as expected. When you analyse xG, you're looking past one match's randomness to see what's actually sustainable.
Why Bookmakers Miss xG Value
Bookmakers price based on historical outcomes and betting patterns. They don't always weight xG heavily, especially for midweek fixtures or lower-league matches where public betting volume is lower. That creates gaps.
Example: Brighton travel to Chelsea in a midweek Premier League fixture. Brighton create 1.8 xG, Chelsea create 0.9, but Chelsea win 1-0. The following Saturday, Brighton play at home to a mid-table side. Bookmakers see "Brighton just lost" and price them tighter than usual.
Our model sees "Brighton actually outplayed Chelsea and were unlucky" — and that underlying performance tends to repeat. Brighton's odds at 1.95 to win at home now look underpriced. That's a value opportunity a punter using xG would spot before the general market.
How Winotips Uses xG in Its AI Model
Winotips analyses xG data using the Dixon-Coles model, a Bayesian framework that factors in team strength, home advantage, and recent form — plus xG trends. We don't rely on one metric; we combine it with defensive solidity, possession patterns, and shot conversion rates.
For every Premier League match, Championship game, or European fixture, we run 10,000 Monte Carlo simulations. Each simulation varies slightly based on the probability distributions we've learned from historical data. The result isn't a single predicted scoreline — it's a probability distribution showing the most likely outcomes.
xG feeds into that model as a core signal. A team with high xG differential (good chances minus bad chances conceded) gets a boost in our win probability estimate. A team fluking wins despite low xG gets a downward adjustment, because we expect that pattern to reverse.
Check today's picks on Winotips and compare odds at BestOdds. You'll see how our model rates each match and where value sits relative to bookmaker pricing.
How to Use xG in Your Betting
1. Compare xG to odds before building your Saturday acca. If a favourite has xG of 1.2 in their last three matches but is priced at 1.50 to win, they're probably overpriced. Their actual goal-scoring ability is weaker than the odds suggest. Look for that mismatch.
2. Hunt xG trends in cup ties. One-off cup matches are perfect for xG spotting because there's no "regression to the mean" yet. A team that creates 2.5 xG in a cup tie but loses to a side that created 0.8 has genuine value in the rematch or their next league fixture. Bookmakers sometimes overeact to cup results without checking the underlying data.
3. Use xG to filter out lucky scorelines in midweek games. Midweek fixtures attract less betting volume and more mispricing. A team that won 2-1 but had 0.6 xG is unlikely to repeat that. Look for that discount and fade them. A team that lost 0-1 but created 2.2 xG is due a correction — their odds are artificially high.
4. Stack xG with other stats for midweek accas. Don't use xG alone. Combine it with clean sheet probability (defensive xG allowed), possession, and pressing intensity. A team with high xG but weak defending is a BTTS candidate, not a tidy 1-0 win.
5. Compare your model's xG projections with bookmaker odds at specialist sites. Once you understand xG, use BestOdds to check live pricing across UK bookmakers. You'll quickly spot where the gap between xG-based probability and bookmaker odds creates value.
Frequently Asked Questions
Is xG ever wrong?
xG models vary. Some are more generous (StatsBomb's model, for example, tends to assign higher xG to long-range shots than Understat). No model is perfect — they're probabilistic estimates, not predictions of exactly what'll happen. Over a season, xG predicts goal outcomes accurately. Over one match, variance is real. That's why we use it as one signal among many, not as gospel.
Can I use xG to find value in BTTS bets?
Absolutely can. If Team A has high xG (1.8+) and Team B has high xG allowed (1.5+ per match), and you're looking at BTTS odds, compare them to the underlying probabilities. BTTS is often underpriced when one strong attacking team plays a weak defence. xG highlights those matchups.
What's a good xG number for a team to win a match?
Generally: above 1.5 xG is strong attacking play; 1.0-1.5 is decent; below 0.8 is poor. But context matters. A team winning 3-0 with 0.9 xG is clinical. A team losing 0-1 with 2.2 xG was unlucky. There's no single "winning xG number" — it's about the xG differential (your xG minus opponent's xG) and whether that difference is reflected in the odds.
Do bookmakers use xG when setting odds?
Some bookmakers employ data scientists who factor xG into their models. Bigger operators like Bet365 and Sky Bet almost certainly do. Smaller or older operators might rely more on historical results and betting volume. This gap is where value appears. Premium or specialist betting sites often price xG-based mismatches tighter because they're run by sharper operators.
Can xG predict injuries or red cards?
No. xG only measures shot quality from completed actions. It can't predict match-changing events like injuries or sendings-off. That's why xG is powerful but incomplete. Combine it with team news, suspension lists, and head-to-head history for a fuller picture.
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.