In a football world stuffed with technology and big data, it can be hard to untangle from the impressive amount of stats, indexes, and algorithms that nowadays represent the basis for the study and analysis of games, training, and players. In fact, in the last years – also thanks to tools like Wyscout – managers, players, coaching staffs, referees, talent scouts, agents, journalists and many other professionals of the Beautiful Game integrated Advanced Metrics and Big Data in their daily work, starting what we could define Football 2.0.
The process didn’t happen overnight, and there still is some confusion about which metrics and indexes are the most relevant to the analysis of the game and what they actually represent. Just considering the Wyscout platform, there are more than 150 stats, both for clubs and single players, and many of them can be considered as Advanced Metrics, that are deeper stats that analyze every part of the game in detail.
To make everything clearer, we decided to inaugurate the “Stats Corner” column here on Wyscout Blog, presenting some of our Advanced Metrics on a regular basis, explaining how they are calculated and what they mean. We decided to start with the Expected Goals – also known as xG – a statistical index that evaluate the quality of a player or a team’s goal-scoring opportunities. Here’s the definition from the Wyscout platform:
Expected Goals is a metric that assigns to every shot a probability to be converted into a goal. This probability is calculated from the correlation between the chance to score according to specific characteristics of the shot – the position, the type of the shot, the game situation, the type of assist, the distance to the goal – and the respective stats archive on Wyscout.
Pretty clear, right? In case you need further explanation, here you have. First of all, Expected Goals is an index; that means is not a single stat but a group of stats that are correlated and evaluated within a specific algorithm. So, in this case, the “Expected” word can be a false friend. In fact, xG is not a provisional data but an index based on a concluded event. Here are a couple of practical examples to make everything clearer.
Low Expected Goals Index
Let’s start with a situation with a very low xG index. That means, statistically speaking, a shot that has been converted into a goal that – at least according to our stats archive – shouldn’t have had a single possibility to become so.
Let’s take a look at Manchester City’s second goal in the recent clash against Maurizio Sarri’s Chelsea at the Etihad Stadium.
At the 13th minute of the first half, Citizens’ striker Sergio Agüero shots towards the goal from far beyond the box, and also from an edged position. Even if shooting with his strong foot – the right one – statistically Agüero has very few chances to score that goal. How few? According to Wyscout’s algorithm, Kun’s goal had a 0.03 xG value. Almost impossible.
As we were saying, this value is calculated by considering the characteristics of the shot – position, distance to the goal, the play’s build-up, the shooting foot, etc – and correlating them with the stats archive about similar shots. Basically, grounding on existing statistical data, Agüero’s shot had the 3% of chances to become a goal.
High Expected Goals index
Switching to an opposite example, let’s take a look at a goal with a higher chance of conversion. It’s November 2018 and, at the Estadio Ramón Sánchez Pizjuán, Sevilla hosts Liverpool in the UEFA Champions League group stage. At the 30th minute, with the Reds already leading 1-0, this happens.
Following a lateral drive by Sadio Mané, which conclusion is blocked by Sevilla’s goalkeeper, Roberto “Bobby” Firmino finds himself alone in front of an uncovered goal, easily scoring after controlling the ball and also indulging himself with a no-look finishing. But the Brazilian’s confidence was more than justified. In fact, the xG index for that shot was 0.85. Almost impossible… to miss (always in correlation with the stats archive about similar shots).
Of course, these are just two extreme shades among the hundreds of variables that data like Expected Goals comprehends. Just think about the (almost) infinite shot possibilities that can happen during a football game – from penalties to impossible bicycle kicks, from lucky shots to unthinkable trajectories.
So, what are Expected Goals for? Ultimately, is a metric calculated in order to analyze the result of the game with more objectivity, removing – as much as possible – its most casual part. If you follow Wyscout official Instagram profile, for example, you may have seen our “Expected Results of the Weekend” column, where we take some results from the previous football weekend and we compare them with the “Expected Results”, that would be the statistical results according to our xG index.
If you take the 3-3 draw between Atalanta and AS Roma – 21st Serie A gameday – it can acquire a very different meaning if compared to the Expected Goal value: 2.57 for Atalanta, 1.55 for Roma. Other than the rough-and-ready analysis – “Atalanta deserved the win” – xG gives the result a more precise value, both in terms of goal scoring opportunities created, with the Nerazzurri way more active in front of the opponents’ goal, and in terms of concreteness, with the Giallorossi capable of scoring the same goals with fewer occasions to do so.
Then no, the Expected Goal index won’t tell you which team really deserved the win, as that’s always the team that actually won. But it will help you – whether you are a coach, a match analyst, a scout or any other professional in football – to better analyze the goal scoring opportunities generated by a player or a team during a match. That’s rather good, isn’t it?