An Insight to Expected Goals Model (xG) in Football Manager 2021

A Football Manager 2021 Guide to Expected Goals

Expected goals or xG is one feature I’ve had on my wishlist for a few years now. and now the feature is officially included!

Today I’ll like to talk about the expected goals model in Football Manager 2021 and its benefits, how it’s calculated and delve into the wonders of the xG model and other metrics.

Besides providing you with an insight into what the expected goals stats are, I will also look closer at the benefits and drawbacks of the stats, how it can be used to analyze player and team performance(s) to identify over- and underachievers in Football Manager 2021 as well as looking closer to the xG match story otherwise known as expected goals plot.

Read on as I provide you with everything you need to know about expected goals model in Football Manager 2021 and examines how the xG stat might considerably change how you play Football Manager forever!

A Brief Background

For as long as I can remember, football managers have always tried to find an edge over their opponent.

Over the last ten to twenty years, the rapid change in the way we use computer technology in our society has led more and more clubs and managers to emphasize sports science and statistical data to better interpret what they see on the pitch.

The cognitive bias and the difficulty of accurately reflect on what’s actually happening on the pitch have always been a source for making faulty decisions.

Focus on better understanding available statistics to make better assessments of player performance has changed the footballing landscape in the last years. The inclusion of the expected goals model is one important factor to the major (r)evolution going on in football, together with compiling football data to better analyze football matches by breaking up the match into its smallest details.

While the implementation of expected goals model or xG measurements are brand new to Football Manager, the statistical data to calculate the xG score has in fact somehow been part of FM since the arrival of the match analysis tool a few years back.

Since the xG model takes into account the number of shot attempts on target, where they come from, the types of goals or assists made by the overall team and every individual player, as well as when shots and goals happened in the course of the match, it was only a matter of time before Football Manager could provide you with a calculation of expected goals based on the data they already gather.

For years I’ve provided the ‘Goal Locations’ and ‘Assists Locations’ in relationship with my tactics to let you see where goals come from and whether most of the goals come from placed shots outside the penalty box, powerful shots, free-kicks or whatever, and what type of pass leading up to the goals; short passes, through balls or more typical set-piece variants.

What we have been missing for years is to know more accurately the probability of the chances you create to make more distinct goal-scoring opportunities than the opposition. Without thoroughly analysis match after match, shot after shot, goal after goal we have wandered in the dark, not knowing whether the chances we create and the goals we score, comes from pure luck or if our tactical instructions and attacking game model are onto something.

Yes, we have had the ability to get a clear overview of the shot accuracy and where most goals are put. And, based on the number of shot attempts and goals scored you can find out how often a goal is scored or how many chances your team needs to score one goal – describing how effective your tactics are, or how good the player is when entring the final third.

And yes, we have had the terms clear cut chances and half chances to vaguely describe how big of a chance the goalscoring opportunity was, which gives us a brief indication of the quality of our chances, to a certain degree.

But we have been forced to base our judgement on statistics focusing on the number of shots taken, goals scored without knowing exactly the quality of the shot (e.g was it taken with a player’s weaker foot or under intense pressure?). Am I riding a wave of luck by scoring 2 goals in 6 attempts or was it just simply unluck that made me drew against an inferior team despite recording 10x more shots on target than me, or did they actually play to my level?

We have only received half of the picture – and was forced to draw our own conclusion to whether our tactics worked or not and base our judgement on performance on average ratings and other team data that helps to enlighten the picture. But the statistical data provided to us didn’t give a plain insight into whether we overachieved, underperformed or actually managed to come to great goalscoring opportunities time after time, unless you analyzed and watched your matches thoroughly and carefully in full match mode … but who got time for that?!

Through understanding the xG model of FM21 and analyze both your players and team performance you’ll be better at spotting strength and weaknesses both within your tactics and in your squad. Perhaps your €30million striker who has gone 5 matches without a goal isn’t the most clever in his positioning or is simply highly unlucky with his finishing!

As you’ll see by reading further, expected goals delve deeper into the statistical data than just the current data about goals and shots attempts and provide you with a more clear insight to players and team performances.

Since I can’t claim to be nor a mathematical expert to create an algorithm to measure xG in Football Manager or one who embraces statistical data if it’s not given to me in plain numbers, the expected goals model and all its statistical capabilities will be a highly welcome addition to Football Manager for me.

Continue reading as I explain the expected goals model and what xG really is.

Expected Goals; What is it?

Expected goals or xG is a mathematical model which aims to quantify goalscoring opportunities based on the number of shot attempts in relation to the situational context the shot was taken in. Basically it aims to describe more precisely the quality of shot(s) based on a number of factors that increases or decreases the chance of scoring goals from a number of shot attempts.

Simply put, it calculates how likely you should have scored from a shot in a specific event and could provide you with a better representation of whether your tactical strategies paid off (e.g working ball into the box or shoot from distance, cross early or from the byline, specific set-piece strategies), and the performance and positioning of your players to come to distinct goalscoring opportunities, perhaps compared to your opponent.

In reality, the metrics define the likelihood of a specific shot finding the target from a specific position of the field. Whether or not the player actually scores or you win is a different story. But with this metric, you’ll have an additional way to assess player’s and team performances in a clearer way than simply focusing on possession stat and the number of shots on target.

Including expected goals in Football Manager 2021 provides a better statistical foundation to base your judgements on.

Opta defines expected goals as;

Expected goals (xG) measures the quality of a shot based on several variables such as assist type, shot angle and distance from goal, whether it was a headed shot and whether it was defined as a big chance.

Adding up a player or team’s expected goals can give us an indication of how many goals a player or team should have scored on average, given the shots they have taken.
Optasports, source

Optasports, source

In plain terms; the context of the shot provides the probability to score from a chance or shot attempt.

Opta’s definition of expected goals only reveals a small part of what might determine or affect the success of a shot. In fact, the algorithm behind expected goals and the factors which influence the outcome of the shot attempt to be converted into a goal is a bit more comprehensive than that.

Even though it’s not yet 100% sure how many of these factors Football Manager 2021 will take into account, as there are different method of calculating it, I’ll provide some of the factors that may influence the xG calculation below.

Expected Assists (xA)

Relating to the expected goals are expected assists aka xA. It’s somehow similar to the key passes but provides a more accurate value to the pass.

The expected assists will measure the chance a pass leads up to a goal assists – meaning the probability of a pass to turn into an assist based on the types of a pass being made, its length, where the pass came from and where the receiver is located.

Instead of providing a value of the number of goals a player would score, the xA metric indicates how many assists a player should have had at an average rate.

This data provides an insight to which players are consistently creating chances or a player’s ability to set up scoring chances without relying on the actual result of the shot on target.

A player might frequently deliver 10 key passes per match but none of them is turned into great goalscoring opportunities, while another might deliver 2 key passes and provide an assist.

The expected assist metric will take into account all completed passes and provide a score from 0 to 1 where a higher xA value will see an increased probability of that pass leading up to a goal.

This means you’ll get the facts about how much of a chance a specific pass can become an assist -whether it’s a 50% chance or simply 10%.

According to Opta, the metric does not take into account whether a shot was made once the pass was received – meaning the metric will rate the quality of passes on a better scale than what is currently achievable in Football Manager.

The benefits of this metric are that it let you spot players who create more chances than being converted or who aren’t able to create as many chances, whether they perform over or under expectations. It can be used to analyze the real quality of your attacking line, the abilities of the playmakers or whether you create promising chances.

Take this example concerning Jamie Schalketon’s passing game against Derby.

He created 7 key passes, but only one turned into an assist. The quality of 4 out of the other 6 key passes might not in reality have a high probability of becoming a high probability goalscoring opportunity. Most of them are aimed to the sidelines which makes the angle for a shot too narrow to get scored.

The other two key passes might have seen a high xA value as they are aimed to a position of the field that has higher chances of getting scored on, as the pass is made into more dangerous areas of the pitch.

The Influencing Factors of the Expected Goals Model

In order to quantify the quality of the shot, there are at least 5 factors that affect the success of the shot.

The foundation of the xG metric is founded from the five different types of goalscoring opportunities – all with their different difficulty level for shots to find the target and the probability of a chance to be labelled ‘big’ or ‘small’.

1) The Shot Location & Angle

The shot location refers to where on the field the shot was taken from, its distance and angle to the goal.

A shot attempted from the half-way line has a slimmer chance of finding the back of the net than a shot attempted inside the six-yard box, unless the goal is empty, but still it’s a slim chance of the ball finding the target.

Similarly, a shot struck in the centre of the penalty box has a higher probability of getting converted into goal compared to a shot from a narrow angle such as close to the corner flag or struck from the wide channel. In fact, around 75% of all penalties are scored!

It’s not only the angle of the shot that Football Manager 2021 match engine takes into account to gather their xG metric. Sports Interactive’s xG model will also take into account shot speed as well.

2) The Types of Attack

The events leading up to the shot has a major influence on the probability of whether the shot will result in a goal or not, as well.

Put simply, the overall types of attack where goalscoring opportunities can occur are :

• Open Play
• Counter-Attack
• Corners
• Free-Kicks
• Throw-Ins
• Penalties

What’s evident is that the types of attack will massively impact on the type of finish and where the shots are coming from; inside or outside the penalty box, in the centre of the pitch or outside, in addition to the corresponding defensive positioning of the opposition in the different attacking scenarios.

As you might have already interpreted, it’s a quite different scenario whether the type of attack is a corner, counter-attack or from an established possession where the opposite defends with two lines of four compared to being limited to only one to three players between the ball and the goal.

3) Defensive Positioning
The situation of the attack differs greatly relating to the defensive positioning of the opposition and the types of attack we encounter.

StatBomb is a football analytics company who base their calculation of expected goals according to the goalkeeper’s location and how far the opposition defenders are from the shot taker.

The studio director of Sports Interactive, Miles Jacobson, has confirmed that the expected goals model in Football Manager 2021 will include defensive positioning.

This means that the metric will take into account the number of defenders around, or in close relations to the ball carrier, or how organized or disorganized the opposition are. Their distance to the shooter and the positioning of the goalkeeper is highly important to get an accurate metric of expected goals.

Football Manager 2021 will therefore take into account whether the goalkeeper or defenders are out of position (an open goal perhaps), the number of defenders between the shooter and goal, and other finer details that affect the quality of the shot, such as whether the player is being pressurized or not.

Example:
Missing the target on an open goal is quite different than if a player shoots from 30 yards simply because he’s closed down and doesn’t have any other opportunities than to finish off the attack instead of losing the ball in a fragile position of the pitch.

4) Body part (e.g. preferred foot)

Besides the positioning of the player who shoots and how far from the goal he is when taking a shot, which is highly important in regard to the probability of the goal-scoring opportunity as well, comes also which part of the body the finisher uses.

If the finish was struck with the players weaker foot, then it’s a lower chance of a goal than if it was struck with his preferred foot.

Similarly, if the attack is finished off by a header or he’s forced to strike the ball with his heel, knee or any other allowed body parts, the calculation may be affected.

This means that the time available for the player in possession of the ball to shot might affect the metrics as well, as it’s less likely the shot will find the target if he does not have time to get full control of the ball and are forced to try first-time shots or use a body part he wouldn’t ordinarily have been used.

Whether a player finishes off and scores with his right or left foot, head or other body parts, is something that’s currently missing in Football Manager… at least I haven’t found it.

Hopefully, FM21 will see the inclusion to the statistics whether a player scores most of his goals from using his head, left or right foot, meaning you’ll have more information about the player’s tendencies.

5) The Types of Assist

The type of assist leading up to the shot on target is as important as the positioning of the shooter as well. Whether the ball arrives on a players feet or head, from a rebound or an opposition mistake will dramatically alter the probability of finding the target – all depending on where the ball receiver is positioned relating to the opposition defenders.

It’s two completely different scenarios if the through ball comes after a counterattack that puts the attacker in a one-on-one situation with the goalkeeper, compared to a placed shot from outside the penalty area with the opponent in a secure defensive positioning, and where the opponent puts pressure on the ball carrier.

Whether or not a shot or header was immediately preceded from a cross or a free-kick is also a factor that must be taken into account.

Another is the height of the pass.

Let say the ball is delivered above the ground in hips or chest height. It requires more of the receiver to get control of it compared to if the ball was struck at the ground. If he’s forced to finish off the attack with a volley instead of carefully getting control of the ball and spend a millisecond extra to get it in a better position to shot, the success of that shot may be affected.

The difficulty in executing the shot or header from a pass made above ground is something that might be important to take into account to get the best picture of how big of a chance the goalscoring opportunity is.

The pass height can be something FM will have difficulties of interpreting as I’m not sure if the match engine takes into account the height of the ball, whether it’s 30cm above ground or 180cm. The problem for the ME in these instances is to correlate the player’s height with the pass height.

Expected Goals in Football Manager 2021

In Football Manager 2021 the expected goals model is included thanks to the services and partnership of SciSports. The British-based company provides performance analysis and tactical data insight, as well as providing recruitment solutions ahead of the transfer window.

Their ground-breaking football metrics not only relating to player roles, skill index aka SciSkill and assessing players potential are just a few other things they offer additional to their Expected goals metric, which Football Manager 2021 will feature.

Sports Interactive has for FM21 created their own xG system. It’s tailor-made to work with the Football Manager match engine which allows us to go beyond what current expected goals models are capable of in real life.

Detail information about xG stats are presented at half-time reviews and full-time summaries – letting you see the xG match story and how each team’s xG has fluctuated over the course of the game. It will also form part of the post-match analysis compiled by your data team.

Let’s take a look at how the expected goals metric will be calculated?

How are Expected Goals Metrics Calculated?

Every shot on target is rated on a scale from low xG, apparently in a light blue colour according to the screenshot in the FM21 teaser, to high xG rating, coloured red. The rating ranges from 0 to 1 where 1 is a clear goal.

A rating of 0.05 means that the shot had a 5 percentage chance of hitting the back of the net, whilst an 0.8 xG rating means the player most likely will score 8 out of 10 times in similar situations.

In other words, each shot on target is given a percentage chance from 0 to 100 to score a goal.

A high expected goals score means the shot attempt has a higher probability of finding the net in relation to all the external factors that minimize the same factor.

Equally, lower expected goals measurement would rarely find the target no matter the number of chances you’ll get to come in a similar situation.

What we basically are doing is to provide value to the quality of the shot and its likeliness to find the back of the net.

What’s mandatory to determine for this algorithm to work as intended is to quantify areas of the pitch that are most likely to result in a goal if a shot is struck within that area.

The importance of one area of the pitch over another will be the same for all players coming up in a similar situation.

This means that the algorithm must take into account historical data and create a map over the pitch about zones and key areas which in average results in more goals than another.

The model needs to know where the shot has been taken on the pitch and in which context. Here, Football Manager might collate data of tens of thousand matches played each day whilst you are playing the game to perfect their calculation of expected goals.

Or, they can base their algorithm on historical data from the Premier League and other leagues already collected by major football analytics companies.

How the expected goals metrics are actually calculated in FM21 is beyond my comprehension, and is yet to find out. But what I can elaborate on, is the different types of xG metrics you’ll come across in Football Manager 2021.

Shot Map xG Metric

The shot map xG metric is most likely available within the in-match team analysis as well as the individual player analysis. It provides you with an overview of the number of shots taken and from where – providing you with a metric that shows you the percentage chance of the shot being converted on average for all players coming up in a similar situation.

In regard to the shots on target statistics, the new expected goals model will be highly beneficial. It provides a supplementary way to analyze your shots. Was there really a chance for a goal from the shots fired, or weren’t they threatening at all?

In Football Manager 2021 the shot map will feature 5 different types of shot attempts – all with their different symbol and calculated in different ways according to the type of finishing attempt and point-to-end – basically its location.

The symbols within the shot map are:

• Glowing Star: Goal
• Purple Hexagon with black: Shot Saved
• Hexagon with diagonal line through it: Shot Blocked
• Black Hexagon with purple border: Shot hit woodwork
• Hexagon with purple cross: Shot Off target

The shot map will give you an insight to how many shots that were struck with feet or head, whether they came from direct set-pieces in addition to give you an insight to the overall expected goals value of the match.

I will provide deeper analysis to the shot map once FM21 is released.

[Fully updated after FM21 release]

Player xG vs Team xG

With the inclusion of expected goals model in FM21 we can imagine that continually updated player xG ratings are at our fingertips in both player competition stats and within the overview of a player profile stats.

A players xG rating is gathered from data about his overall shot attempts on target, conversion ratio and shot accuracy.

By combining data of the number of major chances the player comes to, a player’s expected goals can give us an indication of how many goals a player should have scored on average, given the shots they have taken.

This means that goalscoring opportunities and goals are related to combining a metric which describes the number of goals the player should most likely score.

A player with a higher xG value will indicate that the player is able to finish off high-quality chances better than one with a lower rating in same player position. But it might also mean the player with low xG is unable to come into great goalscoring opportunities, perhaps due to his role (and duty) or positioning on the field.

By gathering data of shot location for each player, their angle to the goal based on the overall experiences within the match engine, the game can calculate the expected goals a certain player can account for.

Prior to FM21 release, we can only hope the game will distinguish player xG and team xG and include expected points and expected goals for and against for each club to provide an alternative league table.

The team xG is compiled from all the data about the players within the team – their value of expected goals – and provides an indication about how many goals the team should have scored on average from the chances they have created or the shots they have taken.

Here the overall statistics of teams total amount of shots on target, their likeliness of finding the target (accuracy), major scoring chances (clear cut chances) and other data about how effective set-pieces are utilized, provides us with the overall team’s expected goals value.

As information about expected goals is gathered not only for your own team but for the opposition as well it’s easy to calculate expected goals for and against relating to the number of goals the team should have scored or conceded, but also determine the expected points as data are compared with the other teams within your league.

Hence it will be easier to see if there are teams or individual players who do better or worse than expected based on the data we have at our disposal.

xG Match Story or Expected Goals plot

The first look at Football Manager 2021 features also confirms that expected goals plot or xG match Story, as SI call it, are included as well. The visual interpretation of the expected goals plot or xG Match story will be a welcome inclusion to the match narrative.

Expected goals plot or xGplots, aims to not only show the quality or the number of major goalscoring opportunities in relation to the goals scored laid out in a horizontal timeline but also give you a second experience to the match.

Here all shots on target (their xG rating) will be combined and become visually presented throughout the match timeline. The graph will increase the bigger chances you create and the more goals you score. Creating 0 chances will let you see the dreaded 0.0 value. Let me know if it ever happens in FM21!

In fact, it’s a representation of how the match unfolds based on the quality of chances.

It helps to see not only when chances are created but can also help you analyze the effect of your tactical changes and whether or not your approach to the match is working as intended.

The visual graph showing the number of attempts on goal through the course of the match will help you assess whether you dominated the match or if it was a close encounter.

Perhaps you’ll notice that going for a more attacking approach lets the opposition comes to more distinct goalscoring opportunities than yourselves or your vision about the match was in fact quite different to what you momentarily got the impression of.

Was the match narrower than the scoreline reveals or did you create more distinct goalscoring opportunities that would 9 out of 10 times let you win, instead of going home with a poor draw in your bag?

Representing xG in a horizontal timeline of the 90 minutes depicts not only when you create the better chances, or not, but also if the goal scored or conceded ‘came out of nowhere’ or if one team dominated the match – somehow revealing through mathematics the chance of winning the match – or its match odds.

The implementation of expected goals and xGplot can massively improve our understanding and the calculation of, and our knowledge of, match odds. Its visual presentation of not only how many goals your team was expected to score but also the percentage chance of a win, draw or loss will massively help to analyze the game after the match is ended.

The calculation of match odds will not only help us to see our percentage chance of a win, draw or a loss within the match we just played, but combining the expected goals and results from all teams in the league, their form and expected position can help to improve the calculation of match odds provided before the match and give you an indication of how the upcoming match might proceed.

The question is how you determine to use that knowledge to approach the upcoming match in the best manner.

By taking into account the expected points relating to match odds and the expected goals plot from each matches you’ll have an alternative league table with expected points.

Do understand that the statistics totally disregards decisive moments of a football match, such as referee errors, sending offs, injuries to key players that might massively affect the end result.

An Example

Let us take a hypothetical example.

Often we’ve come across gamers who complain about the match engine or says they have been FM’d.

They base their judgement of the match according to the number of shots compared to the opponent and the final score, perhaps in relations to the possession stat without focusing on the quality of their chances.

Let us say you are managing Real Madrid and face Real Betis. The match ended 0-1 to the visitor, despite them only managing 3 shots, only 1 on target, whilst you created 40 shots, 25 on target.

You felt you had full control and you should have scored at least 3 goals. Even the amount of clear cut chances proves it.

With the expected goals model implemented, you will better see the quality of every shot. All shots are then combined to determine the number of goals you probably would have scored if the match was played over and over again.

Perhaps you’ll discover that in the match example above, your team got an xG rating of 2,5 goals meaning you would under normal circumstances have scored at least two goals.

Perhaps the story was quite different coming out of the same match with a value of only 0,5 whilst the opposition had 1.2 xG.

Such a low value indicates you were unable to create major chances. The match would under normal circumstances account for very few goals – most likely a draw or a marginal win.

How Can You Utilize Expected Goals Statistics?

How you can utilize the expected goals statistics to your advantage is a great question. Let’s take a closer look.

There are three important questions that bring to mind when it comes to expected goals and its introduction to FM21.

• 1. What’s the benefit of having expected goals metrics?
• 2. How can xG data be used in Football Manager to assess player or team performances?
• 3. How xG data can be used to analyze match results?

1. What’s the benefit of having expected goals metrics

The calculation of expected goals can be an important tool to assess players or your team’s actual performance. It provides a secondary but neutral opinion about the match – based on facts rather than feelings and cognitive biases.

You’ll get a clear indication of the quality of the chances or shot attempts compared to what’s the likely outcome within a similar situation.

You will be able to see straight on how clinical your players are when it comes to their shot attempts.

Instead of labelling it as a clear cut chance or a half chance, you’ll get a probability score from 0 to 100%, which makes it far easier to determine the quality of the chance.

In this instance, we have to remember the true reason why the expected goals metrics came about.

Instead of ranking players by their average rating which might not fully explain the full picture of their play or base your judgement on the match relating to the end result, xG rates players according to how well they are able to convert chances into goals and come up in promising positions for a shot that might lead to a goal.

And that’s the key when assessing the player’s performances.

Another benefit of having access to the expected goals metric is to get an important insight into your run of form – either assessing the specific player’s form or the overall team performance, according to the exected points or overall expected goals for – giving you a brief insight to whether you overachieve, underperform or are playing at the average level.

What’s intriguing me, is that the implementation of xG can also better spot match engine bugs as it reveals trends throughout the seasons which can more easily be compared with each other.

The expected goals metric compromises data from thousands of shots to provide each shot a value. If shots from a narrow angle are converted more often than a shot from a far more promising position, the data will show that.

This might help Sports Interactive and its match engine testers to gather data in a better way as anything unordinary in the xG metrics will stand out for everyone.

2. How can xG data be used in Football Manager to assess player or team performances?

Let’s start this off by asking another question. What’s a great player performance? Is it his abilities to stop the opposite forward from coming into goalscoring opportunities, continually track runs or block ball paths or make key passes and lots of crosses?

In terms of converting chances to goals which is the ultimate goal in football, a great performance should relate to any matter where the players are able to come into a goalscoring opportunity. Likewise, from a defensive perspective, the task must be to avoid the opposition from getting to distinct goalscoring opportunities.

A player who is able to come into key areas of the pitch that effectively increase our chances of scoring and use his accuracy and conversion rate to strike a goal is worth lots!

What the most clinical strikers have in common besides their technical proficiency at finishing on target (their accuracy) are their abilities to come into promising positions. Finishing off attacks in areas of the pitch that comprises high xG values. Some would call it instinct, others would call it great movements and positioning, whilst others may link it to a mix of technical, tactical and mental capabilities.

What’s certain is that it’s most often not luck or a coincidence for specific players to compromise higher xG values throughout a longer period of time, or through the full season. They are just better at coming up in promising positions to score!

Here the expected goals metric, or the expected assists for that matter, can give you a benefit in your quest to sign new players. Instead of targeting specific attributes or star ratings, you can take a statistical approach to signings and judge players on their actual performance – combining data about their expected goals or expected assists with their average rating, and other statistical data and skills they possess.

Players within the statistics that regularly come into promising goalscoring positioning will be easier to spot with the xG model. So will players who create more dangerous chances than others.

These players are topping the list because they shoot often, and from positions with high probabilities of success.

The expected goals metrics can in this instance become a helpful tool to distinguish the importance of a player – either it’s your own team or the opposition. It will give the ‘true facts’ about a specific players impact to team’s performances.

You can use the access to the statistics by looking at whether players are consistently performing to their expectations as you focus on long term trends before signing the player – meaning you might analyze player performances and overall team performance quite differently than what you’ve been used to.

Looking at the team performances relating to expected goals tally or the expected goals conceded can be an important judgement factor to understand whether your tactics and tactical strategies are able to;

• A) defend key areas of the pitch which increases the probability of securing a point or win matches.
• B) come into favoruable goalscoring opportunities that increases our chances of scoring and therefore winning matches.

This way, it can be a great revelation to ensure aspects and weaknesses of your game is put focus on in training, either for the team in general or for a particular player’s individual training focus.

The drawback with the metric is that gamers might simply look at the high or low xG value and simply relate it to poor luck or overachievement if something performs out of the ordinary, without looking closer at each incidence. They might overlook wrong tactical choices or inabilities to maintain form and overrate their chance of winning simply because the match story or expected goals told them so for a short period of time.

Yes, it’s true that a high xG but low output (points per match) can relate to a run of bad luck and that you might eventually come through it by hard work and continue working on the aspects of play that does work!

But it might also mean the players aren’t good enough in these situations – perhaps due to poor technical, tactical or mental capabilities. Most often it might come down to the last.

Likewise, a low xG with high output (point per match) might mean you’re team is overperforming and that they might drop down to their actual level 1, 3 or 10 matches ahead.

By taking this into account, it will give you an indication of whether your tactics are working as intended. Are the right attacking patterns happening that ensures your best players are coming into more promising shot opportunities, or do you rely on your tactics to long shots and a bit of luck?

It provides an internal sense of direction both from a strategical point of view, both in terms of how you approach the transfer market, but also when it comes to making tactical strategies and match plans or what you decide to focus on in training.

3. How xG data can be used to assess match results

While the ordinary spectator can somehow distinguish major chances from small at key moments of the match, it’s more difficult to remember trends throughout the match or throughout the season. The expected goals model and its algorithm can help you to notice these trends and ultimately address necessary changes further on.

The expected goals do not show the true result of the match. Every football match is based on smaller and major incidents that might tip the favour of the match for or against you. But it can provide more information about the match that you might have normally overlooked.

The xG Match story can give you a better understanding of whether applied tactical changes worked or if you approached the match on the back foot already from the start. Perhaps your defensive strategy means you’re unable to come to promising chances or the players are taking shots from the ‘wrong’ areas of the field.

A lower xG value within the match story over a long time might indicate the opposition either defends well, as they force you to finish off the attacks in areas of the pitch that are less likely to result in a goal, whilst a higher xG but no goal means you’re coming to promising chances, but the technical execution is poor as perhaps the wrong player comes to these shots, or the tempo of the play is too high.

We can’t disregard the mental effect completely in this instance. Coming up in goalscoring opportunities may have a huge effect on a players state of mind. The stress level, composure or concentration may outweigh his technical abilities – meaning despite his in a promising position the finish may be poor.

The implementation of expected goals and match story can massively improve our understanding and the calculation of match odds. Based on the number of shots taken and their quality or goalscoring opportunities, a visual presentation of not only how many goals your team was expected to score, but use that knowledge to identify the match odds.

Was the match narrower than the scoreline revealed, or did you create more distinct goalscoring opportunities that would 9 out of 10 times let you win, instead of going home with a poor draw in your bag?

A visual graph showing the number of attempts on goal through the course of the match will help you assess whether you dominated the match or if it was a close encounter – a match you should feel lucky to have come out victorious from.

The calculation of match odds will not only help us to see our percentage chance of a win, draw or a loss within the match we just played, but combining the expected goals and results from all teams in the league, their form and expected position can help us better address the match odds provided before the match to approach the upcoming match in a better manner.

Another area of the xGplot that few might not think about is that they can use the knowledge of the graph to check the opposition’s true opportunities!

Often we look at our own play and ‘forget’ chances or shots the opposition might have. We overrate our chances and perhaps underrate the oppositions.

Perhaps the reason you lost 1-0 despite creating 10x more chances than the opposition was simply that they were able to create more promising shots than you.

This knowledge might become a great tool to take into account before approaching press conferences or team talks. It can change the way you look at the player’s performance and overall team performance as you can better relate your goalscoring opportunities to the opposition’s.