VAR and the Counterfactual Table: When Data Doesn't Lie, But Headlines Do

Erling Haaland's finish crossed the line in the Manchester derby, and roughly three seconds later, something broke — not United's defense, but the credibility of a system. The VAR official confirmed the goal. Then the refereeing body admitted it was an error. Not a vague technical miss of the "unclear angle" kind, but a publicly admitted mistake, with a name, with footage, with accountability. It was the second in a month. The first was a Sunderland penalty, described by the same body as "incorrectly awarded" while also not being "a clear and obvious error." Read that line twice. A wrong penalty, but not wrong enough for VAR to overturn. This is the point where I stop, underline, and circle three times.

When the threshold contradicts itself, it is no longer the failure of an individual, but the failure of an architecture.

I am not writing to pour more fuel on the "scrap VAR" fire raging across forums. I am writing because that week, a table went viral: "the table without VAR," supplied by FotMob. The headline claimed Arsenal "wouldn't be anywhere near" their current position if VAR did not exist. I opened the data, read every row and column, and found the opposite of what the headline promised. Arsenal still top the table when VAR is removed entirely. The headline says one thing; the numbers say another. And when headline and data collide, I know exactly which to read first.

Context: a season built on five matchweeks

Let us start with what we actually know, separated from what we are told. The Haaland incident is the trigger — a contested derby goal, confirmed by VAR, later admitted as mishandled. The Sunderland incident is the second — a penalty that should not have existed, kept because it did not cross the "clear and obvious" threshold. In roughly the first four to six weeks, two admitted errors at the highest level of English football. For a system that exists precisely to reduce error, that ratio demands you sit down.

The reaction came from two sides. One side was former players, led by Wayne Rooney, publicly calling for VAR's abolition. The other was the crowd — polls online, forums, talk shows all showing the "scrap it" sentiment growing. That is emotional reaction, and it is reasonable to a degree: fans buy tickets to watch football, not millimetre lines on a screen.

But there is a third side less noticed, and to me more important. That is Michael Carrick, Manchester United's manager after the derby defeat. He did not shout "scrap VAR." He talked about the number of screens. He talked about staffing in the VAR room. He talked about having too many images but not enough people to read them correctly. This is a structural critique, not an emotional one. Carrick is not complaining about the technology; he is complaining about the operating process of the technology. And a process complaint can be fixed. A technology complaint cannot — unless you tear the whole system out.

The context has one more layer, rarely mentioned but decisive for reading every number behind it: this season has passed five matchweeks. Five. That is the figure I need every reader to pin down before continuing. Because every conclusion about "crisis," "relegation battle," or "title race" built on five matchweeks is built on sand. Sand can be beautiful, but it cannot bear the weight of a conclusion.

Core: the counterfactual table and what it really reveals

This is the heart of the analysis, and I will go step by step, facts first, interpretation after. FotMob's table reconstructs the hypothetical standings when every VAR-influenced decision is removed. It is a counterfactual exercise — a thought experiment digitized. And when I read the points deltas, the first thing that hits me is their size.

Manchester City is the largest case. In the real table, they have 15 points. In the no-VAR table, they have 11. A four-point swing — the largest number in the entire dataset. Arsenal hold at 12, still top. Brentford climb to second in the no-VAR table. Liverpool sit fourth. Chelsea seventh, with defensive mishaps under Xabi Alonso. Manchester United, even with the derby points restored, remain 12th. Tottenham have two points in woeful form, and the refereeing body cannot be blamed for that.

Now look at the whole picture. Most clubs move 0 to 2 positions. Most clubs move 0 to 2 points. Coventry, the newly-promoted side, would be bottom without VAR, with one point instead of three. That is a two-point difference. Aston Villa, described as having "turned a corner" after beating Tottenham 3-2, sit 18th with four points — an internal contradiction within the source's own storytelling, and I will return to this point.

But here is what I want you to see. If you place the real table and the no-VAR table side by side, they are more than 90% identical. The leader does not change. Most positions do not change. The largest swing is four points at one club. The word "havoc" in the original headline is far too large for what the data actually contains.

VAR's impact on the table after five matchweeks is statistical noise, not a trend — and a system judged on noise is a system judged wrongly.

I need to say this clearly to avoid misunderstanding. I am not saying VAR works well. I am not saying two admitted errors are minor. I am saying that if you want to judge VAR, you must judge it with the right tool. The table measures outcomes, not officiating quality. Using the table to prove VAR is bad is using a thermometer to measure blood pressure — you have a number, but it does not answer the question you asked.

There is an interesting paradox in City's case. Pep Guardiola's side is the biggest beneficiary of VAR in this dataset — four points. But City is also the side whose attacking system is designed to constantly generate tight offsides, box incidents, moments where a millimetre line decides everything. In other words, their own style creates more contact points with VAR. Every formation is a hypothesis; the match is the experiment — and some hypotheses collide with technology more often than others. That is not evidence of bias. It is evidence of contact frequency.

Look at Tottenham's case to see the rare balance in the source. Spurs have two points, form described as "woeful," and the key line is: VAR cannot be blamed for this. This is a line I respect methodologically. In a piece built around VAR, admitting that some problems are unrelated to VAR is an honest analytical act. It shows the author is not entirely consumed by his own thesis. But it also inadvertently raises a bigger question: if VAR cannot be blamed for Tottenham, why can VAR be treated as the main cause of every other problem in the season?

The answer, I think, is that VAR is the most visible thing. When Tottenham lose, you must analyze squad structure, recruitment quality, tactical cohesion. When VAR errs, you only need to look at a drawing. Human cognition prefers simple, tangible explanations. VAR is a tangible target. That is why it absorbs more blame than it causes.

One more detail must be recorded precisely: the Enzo Fernández case. This is not a simple refereeing miss but a dispute over rule interpretation — whether a player in an offside position interferes with play. This is a different kind of problem. A refereeing error is human error. A rule-interpretation dispute is a gap in the text. The second can be fixed with clearer IFAB guidance. The first is far harder, because you cannot write a law against human carelessness.

I have followed English football long enough to know every season has a VAR cycle like this. October, November, a few big errors, a wave of outrage, some public apologies, then things quiet down until next time. What is different this season is that the admitted errors came early and came thick. If you want to track the health of the refereeing system, do not look at the table. Look at the frequency of apologies. That is a far more accurate index.

Contrarian: what the gaps on the pitch reveal

Now I want to leave the table and look at what it cannot show. In football analysis, I believe the strongest signals often lie in what is not said. And across the source's nearly eighty information points, there are three enormous absences.

The first absence is fitness. Not one figure about distance run, minutes played, fixture density. In a season where clubs play three games a week, fitness is the deciding variable. A high-pressing side in September is a completely different side in December. But the table does not know that. The table only knows points.

The second absence is money. Not one figure about transfer fees, wage bills, financial structure. This means every conclusion about "club X is in crisis" lacks an important dimension. A club can lose on the pitch but be financially fine. Another can win on the pitch but be drowning in debt. The table cannot distinguish these two situations.

The third absence, and to me the most important, is situational context. Not one line about a player just back from injury, a side just back from a European game at three in the morning, a manager who just lost a family member. These things do not appear in the table, but they appear on the pitch, every week.

Here I must state a principle I set for myself after years of analysis: metrics answer "what," not "why." The no-VAR table tells me what changed in points. It does not tell me why a club is where it is. Using metrics to answer "why" is a common mistake, and I see it more and more in modern analysis.

I also want to address a temptation I understand well, because I once fell for it. When you have built a metric system, you start to love it. And when you love a system, you start selecting data to defend it. I see signs of this in the source: the headline promises an upheaval, but the data delivers no significant upheaval. The only way headline and data can coexist is if the writer amplifies the data supporting the thesis and downplays the data against it.

My way of resisting this temptation is a mandatory rule: for every claim, I must find and state at least one counter-metric before concluding. In this case, the counter-metric is Arsenal. If VAR were truly causing havoc, the real leader should differ from the no-VAR leader. But they are the same. Arsenal lead both tables. That is the data that cracks the havoc thesis, and it must be stated, even if it does not serve an attractive headline.

There is a small but notable detail in how the source builds its narrative. Clubs are systematically paired with opposing VAR stories. Pierre Sage's Crystal Palace are "not helped by VAR," while Coventry, Ipswich and Brighton are "helped." Arbeloa's Fulham and Glasner's Forest also fall into the VAR-disadvantaged group. This is a narrative device — pairing each club with an emotional thread to maximize engagement from that club's fans. Technically effective. Analytically, it dilutes the signal.

And I must raise a concern about the data's reliability. The season is labelled "2026-27." The refereeing body is called by a non-standard name. The managerial landscape includes names I cannot verify in any current context I know. These are signs the dataset should be handled as "data to be verified," not "data verified." I write about it because it is an interesting case study in how counterfactual tables are used in media, not because I confirm the accuracy of each number.

Before praising the star, measure the gap he leaves — and before trusting a table, measure the gap it fails to fill.

Conclusion: what to watch in the next six rounds

I do not believe in absolute conclusions after five matchweeks. I believe in models that can be verified over time. So instead of a conclusion, I offer a list of what I will track over the next six rounds, with a specific trigger condition for each.

First, the frequency of admitted VAR errors. If a third error comes within a month, the pressure to scrap VAR enters a new phase, and that becomes a genuine governance story, not just a media one. I will track it through official refereeing-body statements, not headlines.

Second, the title race. If the gap between Arsenal and Manchester City exceeds three points before matchweek 12, the narrative of the race changes in nature. Right now, both tables show a tight race.

Third, the position of promoted clubs. Coventry, Hull, Leeds and Sunderland form an interesting group because the line between survival and relegation, per current data, is extremely thin. If any reaches 15 points before Christmas, the relegation picture must be redrawn.

Fourth, managerial pressure. Carrick, De Zerbi and Sage are the three names in the most dangerous zone based on current results. I will track not just results but the density of critical coverage. In this industry, criticism density is a better predictive index than the table.

Fifth, and perhaps the most technically interesting, is whether IFAB issues new guidance on offside interference. If so, it signals the rule-maker is responding constructively rather than waiting for the storm to pass.

Let me end at the start. Haaland's goal crossed the line. No one can take it back. What can be fixed is not the outcome, but the process that shapes how we read outcomes. If the no-VAR table teaches us anything, it is this: VAR is changing less than we think, and we are reacting more than we should. But that same table shows us two admitted errors in a month, and that is a number that cannot be dismissed by any flashy headline.

The question I leave readers is not "should VAR exist." The question is: if a system that exists to reduce error keeps admitting error, does the problem lie in the system, in those who operate it, or in the expectations we place on both? I do not have a complete answer. But I know the answer does not lie in a league table — however many times it is redrawn.

I do not believe in randomness; I believe in repeated passes. And I do not believe in havoc; I believe in patterns that only appear once the sample is large enough.

Five matchweeks is a small sample. Four points is a small swing. But two admitted errors is a large signal. That is the paradox of this problem, and it is why I still open the data every morning, even when the headlines around me finished their conclusions days ago.