Let’s imagine that people once predicted football results without the warming embrace of the thrusting venture science of looking at football-related numbers.
Honest guvnor it happened just like that, we didn’t have no venture science to fall back on, you can trust me on that guvnor.
People like me Nan predicted football results for forty years trying to strike it rich on the pools, even your humble narrator and his equally humble friends predicted football results on a fixed odds coupon from time to time, and none of us had no venture science.
Back then everybody predicted them results without no venture science.
It’s mad ain’t it guvnor, we still gave it a go even without them numbers of science behind us.
You couldn’t keep us down guvnor, we was happy go lucky we was. We was all in the gutter but some of us was looking at the stars
Consider two things that happened in the last few weeks.
The first was the production of this tweet; “Arsenal Expected Goals 5.01 Man Utd Expected Goals 1.82″. I didn’t discover the tweet after the match but I still had the following thought “That’s uncanny!!! The final score was Arsenal 5.01 Manchester United 1.82. If only I could go back in time and put £100 on the result”. If only I could go back in time!
The second was the resumption of the social media debate between the proper football Neanderthals and the thrusting venture scientists that look at football related numbers. It was all Jeff Stelling’s fault because he had the unmitigated gall, if not sheer audacity, to be one of them typical proper football Neanderthal types. He doubted the efficacy of the concept/phrase called “Expected Goals” in a particular context.
Naturally like all other proper football Neanderthals I was aware that a concept/phrase called “Expected Goals” existed. I naturally presumed, like all other proper football Neanderthals, that it referred to the number of goals that one expected to see in a particular match. Presumptions get you nowhere so I fired up google’s banter engine.
It seems that “Expected goals“ is……
“…a metric which assesses every chance, essentially answering the question of whether a player should have scored from a certain opportunity.
Put simply, it is a way of assigning a “quality” value (xG) to every attempt based on what we know about it. The higher the xG – with 1 being the maximum – the more likelihood of the opportunity being taken.
So if a chance is 0.5xG, it should be scored 50% of the time.”
Unless I have grasped the wrong end of the metric stick “Expected Goals” appears to be more than just a method of determining the number of goals that one expects to see in a particular match, it also involves the scientific knowledge of knowing whether a player should score when they find themselves in a position that appears to be a goalscoring position. What a breakthrough!
For a century and a half we proper football Neanderthals have been lumbering around in the dark distinctly unable to tell whether a chance should, or should not, have been scored. Now it seems that we had better pipe down because society finally has a scientific way of determining whether someone should have scored or not from a particular goalscoring position.
Mate, mate, mate, let that sink in, proper football Neanderthals have been quite literally wandering around in the literal darkness of literal football ignorance. Yeah mate let that fact literally sink in.
AND LO they did tweet “THAT ESCALATED QUICKLY” and other social media phrases. AND LO, it did sink in.
For the last century and a half us proper football Neanderthals have allowed players to blithely run around without the correct knowledge, we simply have not been able to tell whether a goal should have been scored or not in a given situation.
Well begone thy darkness, begone!
For now we can see the light!! Rejoice!! Rejoice!!
If I weren’t a proper football Neanderthal I would be really, really angry at those bloody proper football Neanderthals for subjecting us all to the terror of their deeply unscientific football. If only they had been furnished with simple understanding, if only they had been able to assess whether a player has the ability to score goals or not.
I think I need to let that sink in a bit more mate, all those wasted years mate, all those years, wasted! WASTED! I definitely need to let that sink in a bit more mate.
I am telepathic so I know what you’re thinking
“Mate mate mate what do you expect from your proper football Neanderthals? Decisions have never been part of their job descriptions. They’ve literally NEVER had to make decisions about players based on a comparison.
They’ve literally never had to look into the eyes of young hopeful and literally say “I’m sorry Son but you just haven’t quite got it. We wish you all the best for the future.”
You start to ask yourself how could they literally be like that, then you literally remember that they are proper football Neanderthals and they literally can’t tell which players literally haven’t quite got the skills required for a career as a professional footballer.””
Do you know something mate? You’d be correct. Your proper football Neanderthal can’t make decisions based on judgement because they’ve only gained a lifetime’s craft knowledge in the specialised employment sector called association football.
Praise Be!
It’s only thanks to the scientists from the thrusting venture science of looking at football related numbers that we can now predict football results with any level of scientific rigour.
I look back at the time before scientific rigour and shudder. How did we have the audacity to try and predict results? How did we have the audacity to enter betting shops with just the hope of future riches to guide us? It fair makes my blood run cold. How did we cope with that darkest of dark places?
Predicting football outcomes!!! What a breakthrough!
Let us gaze at the glorious idea of “Expected Goals” and hail the breakthrough!
Let us hail all of them science type people mate!!
LET US ALL LET THAT SINK IN TOGETHER!!
AND LO they did tweet “THAT ESCALATED QUICKLY” and other social media phrases. AND LO, it did sink in.
Having said all that, knowledge in this area would be extremely useful for coaching staff, as the BBC article says
“To the naked eye they were struggling, but xG was identifying a team that would improve soon,” said Duncan Alexander, Opta’s chief data analyst.
“Lo and behold, around matchday 12, their fortunes changed, with the team starting to score at – and above – the expected rate. This led to them actually performing better than their xG said they should have in their eventual stroll to the title.”
The concept of “Expected Goals” could help player development. For example it could provide the pictorial tools to support coaching like “Have you thought about standing here instead of there?”. It’s an obvious thing to say but performance in a particular match situation could be improved by thinking about that particular match situation, especially when you may be able to call on data.
I can see how the data may inform tactical thinking. It is extremely difficult to control every aspect of a match but individual incidents in specific regions of the pitch, i.e. the penalty area, can help to determine a result. Relevant data about those regions of the pitch may provide insights.
While data is useful within the work environment of football but I am not sure it is much use outside that environment. Coaches have to analyse options and possibilities dispassionately but fans only need to hope for the best, in other words coaches can directly affect match situations but fans cannot.
Concepts such as “Expected Goals” have been sold as part of an analytical approach that will provide layman with a more refined understanding of football. On a fundamental level this idea has potential. The understanding gained through study allows human society to evolve, everybody is capable of understanding society through study, therefore everybody is capable of understanding football by studying it.
The academic study of football would probably require a level of time and effort that most people would be unable to devote but fans wouldn’t actually need to produce data because the media would probably digest and present research.
Therefore in principle the process of study would allow fans to understand more about football. I can see the potential in debunking commonly held attitudes about typical terrace scapegoats. Before we move on too quickly it’s worth considering whether the information provided data concepts actually adds anything worthwhile to a layman’s understanding of football, can fans benefit from data concepts?
Imagine your side lost 2-0. It won’t matter how much detail the stats convey, or how many times you look at them, your side would still have lost 2-0. The scoreline, and perhaps a brief match report, will tell you all that you need to know. There’s a choice to make; accept the score or howl at the moon as an appeal to the deity of football fairness on the basis of favourable match stats. Football has never been concerned about the fairness of outcome, somebody has to lose after all (unless there’s a draw).
Concepts like “Expected Goals” gild a lily that doesn’t need to be there. Aside from providing a slightly clearer view for betting purposes fans cannot do anything practical with the statistical knowledge.
Knowledge about what could, or even should, happen becomes pointless when a match ends with a different scoreline. The fact that scorelines often differ from predictions immediately calls the utility of a metric such as “Expected Goals” into question.
If the concept of “Expected Goals” had never been conceived it would not matter in the slightest, most fans do not need extravagant methodology to prove what they can recall from memory. Let us look at the following quote from the BBC article;
“So if your team is performing above or below expectations at the start of the new season, a look at their expected goals difference could tell you whether that run is likely to last.
And it may also be something worth thinking about the next time your striker misses what you always thought of as an easy chance.”
Fans will already know when their team plays well, whether a season has been below expectations, which players make a difference and which sides are the strongest in their division. These ideas are hardly new, they’ve been around since the 19th century foundation of association football. It’s not hard to pick theses insights up, you just have to follow football,
OOOOH SCIENTIFIC “Expected Goals”!!!!!! Get you, with your magical powers!!!!!
I’ve managed to notice that fans are also able to recognise what happens in individual situations, I haven’t done any research like, I have just noticed by looking.
Let’s imagine that a particular striker is standing in their typical sort of position, a similar position from which you’ve seen him score many times. You know that when the ball is crossed you could be milliseconds away from seeing a goal. You can picture him striking the ball cleanly and the ball flying past the keeper. The football dopamine factory is in full effect during the milliseconds of anticipation between the crosser’s leg moving backwards and the ball flying through the air.
Now imagine that just before the perfect cross lands on the proverbial pre-decimal coin the striker slightly loses his footing in the slippery conditions, so when he connects with the ball he scuffs it well wide of the goal. You don’t need an expert venture scientist to explain why the striker missed because you saw the striker slip. QED or something.
The naked eye can see the effect of a momentary loss of concentration, or when crosses are slightly too fast, or defenders are marking too tightly, or when strikers stumble slightly. Surely we don’t need a complicated theory to explain what we already know. Like I said earlier I haven’t done any research, make of that what you will.
LET THAT SINK IN MATE.
I appear to have been light years ahead of the curve, I already knew that van Nistelrooy would score more than Gary Neville and Ian Rush would score more than Alan Hansen and Ian Wright would score more than Steve Bould. If I wasn’t so humble I’d ask the Nobel committee to consider my thesis entitled “Expected Passes That Set Up Goals” for next year’s Nobel prize in the thrusting venture science that looks at football related numbers.
I don’t want to blow my own trumpet here but I would be a shoe-in for the award for I doth have a PhD in thinking up stuff what is bleeding obvious. As part of my submission I will tell the Nobel committee about the people of Llandudno; they often crowd around me in order to touch the hem of my garments and tell me that they would love to have the level of intelligence and sophistication like what I have got.
Greater knowledge can lead to greater enjoyment but bare statistics only explain so much. A misplaced pass has the same statistical worth whether it happens just outside a penalty area or near the halfway line. Stats can tell us that a side with 75.4% possession lost a match by conceding injury time goals but they can’t tell us why did that happened, which is the most interesting bit.
The concept of “Expected Goals” appears to have more explanatory power than bare statistics but it still lacks utility for fans.
There are two main reasons, firstly it’s pointless to expect something in football because uncertainty lies at the heart of the sport and secondly, people are attracted to football by emotion not logic.
You can roughly predict general outcomes but it would take a genius with preternatural abilities to continually make correct predictions. A friend that regularly bets on horse racing once told me that football’s inherent uncertainty is the reason he didn’t bet on football results, there are just too many things that can go wrong.
The flow of matches is dictated by the fluid interaction of single events that can alter the course of a match in a split second. A player could slip, someone could pass the ball to the wrong player, a player could stand in the wrong place, a goalkeeper might fumble a cross, a defender’s concentration may wander momentarily. Football’s nature means that similar mistakes don’t always lead to goals.
Uncertainty sounds messy and random but the uncertainty created by football’s inherent drama is the thing that draws people to football, can you imagine how boring it would be if all matches ended in the predicted score?
The proponents of ideas like “Expected Goals” imply that greater enjoyment will come from greater enlightenment but an attachment to football’s vibrant culture is something that people feel on an emotional level and I doubt whether it is possible to explain an emotional attachment with data.
We know what we feel when we see flowing moves, superb skill, precision shooting, great comebacks, unexpected goals. Fans can get behind their side and roar them on, or make the opposition wilt, and we don’t need a detailed scientific outlook to do that.
We all know what it feels like to long for a result, or to taste the glory, or to feel mesmerised by history, or to feel a sense of belonging, or to enjoy the camaraderie, or to hear the stories. We know that the match will probably be dull but there’s always the hope that we could see something fantastic. A recording of a memorable match can enable you to relive the feelings, looking at data won’t elicit the same response.
Let us consider Wales in Euro 2016.
Let’s start with the data perspective. Ladies and gentlemen may I present the minutes in which goals were scored in Wales’ matches between September 2014 and July 2016.
Euro 2016 Qualifiers
Wales – 13, 22, 23, 26, 45, 50, 50, 77, 81, 82, 86
Opponents – 6, 36, 71, 90
Euro 2016
Wales – 10, 11, 20, 31, 42, 56, 67, 75, 81, 86
Opponents – 13, 50, 53, 56, 61, 90
Now let us turn to idea of an average score, the “Expected Goals” if you will, from those matches;
- In the qualifiers the average score was Wales 1.1, Opponent 0.4 (10 matches, 11 goals for, 4 goals against).
- In Euro 2016 the average score was Wales 1.6, Opponents 1 (6 matches, 10 goals for, 6 goals against).
According to the Venture Science of Looking At Football Related Numbers we should have been able to take the knowledge indicated by the above information and prepare ourselves for the unexpected joy dans la belle France, the average scorelines have a degree of similarity do they not?
The fact that we obviously didn’t know what was going to happen shows the limitations of concepts like “Expected Goals”.
Even a cursory analysis of the data tells us that most predictions would have been wrong; during the tournament Wales were not only more potent, playing four fewer matches but only scoring one goal fewer, they also conceded more second half goals. This impression is also reflected in the average scorelines.
All data has limitations, in football it is more useful as an indication of what happened rather than as a signpost of what will happen. But then football results of the past have never determined future football results.
In terms of Wales at Euro 2016 bare statistics cannot even begin to explain what happened because they cannot begin to explain what it felt like to witness the events. Watching Wales play in Euro 2016 was the sort of fantastically bewildering experience that any football fan would enjoy.
The utter joy of qualification did not transfer into expectation. Not even the most positive Welsh fan would have even been able to conceptualise what actually happened in France, and that includes those that bought follow my team tickets.
Data was useless for mental preparation. Wales’ pre-tournament form guide told us that they lost their last four matches, and one of those was a week before the first match yet they reached the semi-final. Hopes, dreams and general football knowledge would have been as reliable an indicator as the data. The only thing we could have expressed with any degree of certainty is the idea that Wales might score and they might concede.
Not even the simple knowledge what happened when Wales play against the same sides helped. The qualifiers against Belgium yielded an aggregate score Wales 1 Belgium 0, when the sides met in the Quarter Final Wales beat Belgium 3-1 in one of the most memorable matches Wales has ever played. The first two matches are unable to explain the third. I still feel warm and fuzzy when I picture Hal’s Cruyff turn.
Data can’t explain what it felt like to see Ashley Williams’ equaliser, or Hal Robson-Kanu’s skilful finish, or Sam Vokes’ late header and it certainly can’t explain the once-in-a-lifetime experience for those in Lille or the open mouthed joy for those that watched on TV.
To underline the point further let’s consider qualifiers for the 2018 World Cup as well. From Euro 2016 to Euro 2016 to World Cup 2018 qualifiers the average score only changed slightly; from Wales 1.1, Opponent 0.4 to Wales 1.6, Opponents 1 to Wales 1.3, Opponents 0.6.
The similar average scores suggest a similar experience yet the three stages caused distinctly different emotions. The first stage had the glory of an historical qualification, the second stage had the glory of an unprecedented semi-final, the third stage had a very irritating defeat caused by Ireland in the most important match.
In other words the knowledge from the Euro 2016 qualifiers didn’t prepare us for didn’t prepare us for Euro 2016 and the knowledge from Euro 2016 didn’t prepare us for the World Cup 2018 Qualifiers.
I may believe that facts and stats don’t really add to a fan’s enjoyment of football but in our gossamer skinned times people actually enjoy using stats on social media to prolong pointless arguments. I quote verbatim;
“Mate mate mate, you can’t challenge me for I have the right to free speech. THAT’S “ME”, YES ME. I HAVE THE RIGHT TO FREE SPEECH. I’m entitled to my opinion mate and it happens to be better than your opinion mate. You know sod all about football. Get over it SNOWFLAKE. LET THAT LITERALLY SINK IN MATE”
Verbatim and ad nauseum, the reduction of football to a never-ending struggle to finally prove which judgement based opinion is the bestest ever mate. They’ll use anything in these social media spats; Net Spend, Shots, Possession. Look at me, Look at me, Look at me.
It’s like some people have forgotten that data does not predetermine the future. A player can play against the same club twice and play well in one match but play poorly in the other, or they can eradicate the mistakes from the first match by the time they play the second, confidence can be regained in the intervening period.
Anybody with a semblance of football knowledge can make rough predictions, fans don’t need data. Knowing that a side has error-prone keeper from watching Match Of The Day highlights is as enlightening as a set of data. Emotions and experience prepare fans for football outcomes rather than data.
If somebody wants to interact with football on the basis of a statistical relationship that’s fine and dandy, it’s between them, their conscience and their free time, it matters not a jot. I just wish they’d keep it to themselves. The thrusting venture science of looking at football numbers has become yet another example of football’s imperialistic tendencies.
Football should be a hugely enjoyable diverting pastime that’s reasonably important to those that are interested but it seems to have developed an extreme sense of self-importance.
I try to ignore the tedious flotsam and jetsam but other people won’t allow me and football continually encroaches too far into my peace of mind. Take last weekend, I was getting ready to go to Rotherham when I caught the short BBC Breakfast News report about the previous night’s Arsenal 3 Liverpool 3 match. The reporter told us about culpability, mistakes and pressure rather than a “clearly exciting and pleasant way to spend a couple of hours”. It was as if somebody needed to be held accountable and punished
As Thom Yorke once sang, I’m a reasonable man get off my case.
I read “Expected Goals” and foresee a new stick to beat a team with.
“I know mate, won 2-0 but our xG was 4.2. That’s not good enough mate! He’s gotta go!”
I foresee the whimpering of data acolytes.
“Mate, Mate, Mate, It’s not my fault the cretin scored in the 83th minute instead of the 9th minute. He’s a proper football man at heart and didn’t know that most of his own goals come from the 8.4-11.7 minute window within the hidden trapezium of max. goal scoritude.
Stats don’t lie mate. Stats don’t lie. He’s just a proper football Neanderthal mate, he’s basically a performing monkey, a lab rat.”
Mate mate mate football doesn’t really matter.
Isn’t it enough to know that football runs by the internal logic of its own nature? It’s like the man on the When Saturday Comes message board implied, you don’t need to know the serial number of pantones to be affected by the beauty of art.
LET THAT SINK IN m7.
You must be logged in to post a comment.