We’ve all watched soccer matches where our team has dominated the opposition, creating chance after chance, only to end up on the losing side. We’ve all bemoaned our side’s bad luck, but we haven’t had the tools to show that our team actually deserved to win. Until now, thanks to Expected Goals advanced statistics.
The Expected Goals metric is becoming increasingly popular as a means of assessing the performance of teams and players.
Football is an incredibly hard sport to analyze. The rarity with which goals are scored means that luck plays an enormous part in success. This means that the final score is often unreflective of the performance levels of the two sides.
So, how can we use xG to overcome this problem? And how can bettors use xG to give them a better understanding of what’s happening out on the pitch?
2022 World Cup Expected Goals: xG, xGA, xGD
Here are the actual goals data vs. Expected Goals for each of the 16 nations that advanced out of group play to reach the Knockout Stage of the 2022 World Cup in Qatar.
|Team||Games Played||Goals Scored||Goals Allowed||Goal Difference||2022 World Cup xG||2022 World Cup xGA||2022 World Cup xGD|
What is the Expected Goals method (xG)?
The Expected Goals method offers, by far, the most advanced, profound and accurate gauge of footballing performance. It allows us to quantify exactly how well teams or players are performing when luck is stripped from the equation.
Before we study how we can use xG to win in the betting markets, it might be worth explaining how the Expected Goals method rose to prominence.
The Expected Goals method first started appearing in niche online forums in 2013. These football analysts realized that a great deal of insight could be gained from looking at the quality of attacking situations that a team was creating (and conceding), as well as the quantity of them.
Before long, bettors had caught on to this and began to use Expected Goals to take on the bookmakers. Professional gamblers used xG to develop more accurate match forecasts than the bookies, and in turn were able to turn over millions of pounds.
Soon enough, professional football clubs realized the power of xG. Brentford have been the leaders in this field, revolutionizing their recruitment set-up and using Expected Goals to identify and sign undervalued talents. Their “moneyball” style of scouting has taken the small West London side from League One to the Premier League, and has allowed them to turn over more than £130m of transfer revenue since 2015.
The media has finally begun to adopt xG to provide more profound insight into their analysis. The BBC started showing Expected Goals stats on Match of the Day in 2017, while Sky Sports also now regularly shares xG data across their various platforms.
Slowly but surely, xG is becoming more mainstream. An increasing number of people have realized the revolutionary impact that Expected Goals can have on football analysis.
How can Expected Goals be used to analyze football?
The Expected Goals method offers a statistical measure of the quality of chances. In essence, every shot which occurs in a match is given a Shot Probability – the percentage chance that a shot from that location will result in a goal.
For instance, a speculative shot from 35 yards will have an extremely low xG – let’s say, 0.02(xG). On the other hand, a tap in from very close range will have a much larger Shot Probability – in the range of 0.50-0.80(xG). Most shots fall somewhere between 0.05-0.25(xG).
There are several parameters that are used to measure the difficulty of a shot. The location of the shot is obviously incredibly important. How far out was the shot taken from? Was it taken from a tight angle? But other factors also play a part. Was the shot taken on the player’s weaker foot? Was the ball on the ground or was it a volley or header? Were there any defenders between the shooter and the goal? Was the goalkeeper well positioned? And so on.
This information is obviously very useful when it comes to analyzing team performance. Take Tottenham’s one-nil victory over Man City on the opening weekend of the 2021-22 season. Although Spurs emerged triumphant, Man City clearly had the far better scoring chances. This is reflected in the xG data, which read Tottenham (1.06) 1-0 (2.24) Man City, when all of the Shot Probabilities of each team are added up (source: @xGPhilosophy).
Expected Goals allows us to clearly measure how many goals you would have expected either team to score, given the chances that they created.
Expected Goals can also be used to highlight the over-performance or under-performance of individual players. For instance, Son Heung-Min scored 17 goals last season, despite amassing just 11.01(xG).
There are two ways of analyzing this data. First, you might applaud Son’s finishing ability as he’s clearly scored more goals than you would have expected him to. But on the other hand, you might expect him to score fewer goals this season unless he can sustain his exceptional finishing.
What does xG mean for sports betting?
The Expected Goals method can help us discern the true ability of teams; therefore, it is an incredibly useful tool for betting.
Consider a team who has been consistently dominating their opponents, but who have ended up on the wrong side of the scoreline more often than not. The bookies might look at their results and undervalue this team’s chances of winning their next game, meaning that there is value in the market for punters.
Throughout the season, we will study the xG data of all 20 Premier League teams each week in a bid to find this value.
Note: It’s likely that the majority of the bets we suggest will lose, given that we often find value with the underdogs who have 10-30% chance of winning. However, the aim is to make enough money back on those winning bets to end up in profit.
The key to using xG to bet is finding value in the betting market, and hopefully over the long run this will prove to be the case.
By James Tippett
Author of The Expected Goals Philosophy, available to buy on Amazon here.