Anatomy of the 2026/27 "No VAR Errors" Table: When Opinion Wears the Shirt of Data

A goal ruled out on matchday three. A vote opened. Another table was born.

Nottingham Forest scored. The referee disallowed it. VAR upheld the call. The Premier League issued a same-day statement citing handball. Most fans did not read the statement through — they voted instead. And the aggregate of those votes became a column of numbers in the "no VAR errors" table Squawka published for the 2026/27 season, where Tottenham sit bottom on zero points, Manchester United land inside the top ten, and Brentford climb to third.

I rewatched that incident four times in my office before sitting down to write this. Not to judge the referee. To answer a different question: are we reading a verification, or a poll dressed up as a league table?

The biggest mistake is not choosing wrongly — it is choosing without enough data. Here, the input data is opinion: not a calibrated offside line, not a semi-automated model, not the league's independent key match incidents panel. That distinction caps the confidence of every downstream conclusion, including the ones that look right.

Before the data, the context. VAR controversy in the Premier League is not new. It has accumulated since 2026/20 and now runs on its own media cycle: a disputed decision appears, two pundits demand abolition, a data platform builds a composite table, and that table gets screenshotted and shared faster than the official report. In the current cycle, Roy Keane and Wayne Rooney have publicly called for abolition. That is the peak of a mature attention economy, not a new summit.

Squawka's table wears the visual form of a real league table: ranks, points, a top and a bottom. But its mechanism is a poll. Only disputed incidents go to a vote. Non-disputed correct decisions are excluded from the sample by construction. That is classic selection bias, and it has a very specific technical consequence: a "corrected" table systematically inflates controversy, because it only sees the visible tip.

One point must be stated plainly: the dataset behind the original piece contains several details outside the verified competitive landscape — Coventry City and Hull City in the top flight, Roberto De Zerbi managing Tottenham, a single-window spend above £300m. I file all of these under "data to be verified" and read them throughout as either a forward-dated simulation or an editorially constructed entertainment piece.

That does not void the exercise. It only forces the reader to redistribute the weight.

---

Start with the hardest signal, because in a dataset where the causal layer is entirely absent, the only hard signal is a negative one.

A side that spends more than £300m in one window and then fails to score across three openers is, on outcome evidence alone, showing a conversion failure from build-up to the final third. That is the classic symptom of a brand-new attacking unit that has not yet established shared movement patterns — the signature of a gelling phase, not a talent deficit.

But stop at that boundary. The original piece supplies intent, not mechanism. It calls Tottenham "dismal" but offers no formation, no pressing detail, no structure. Any claim about why the attack has stalled is speculation. The honest position is that the causal layer is missing.

De Zerbi's profile implies a control-oriented, possession-heavy family of systems. Those systems typically carry a higher-than-average integration cost early in a project cycle because they demand positional discipline and rehearsed patterns. Failing to score over three matches is consistent with a system still being installed — but does not prove it.

Here a technical blind spot deserves naming. There is no expected-goals figure anywhere in the dataset. That means we cannot distinguish "not creating" from "not converting." These carry opposite remedies: the first is a structural and creative problem, the second a finishing and variance problem. Misread either, and the conclusion tilts entirely.

Space cannot be bought with money, but it can be created with thinking. That is the line worth pinning up when discussing a £300m outlay.

Financially, the only quantifiable fact is the single-window spend. On standard accounting, that figure amortises across contract lengths, so the annual P&L charge is a fraction of the headline — but the headline still carries real cash flow and a future amortisation tail. Without wage or revenue data, the compliance-headroom question cannot be answered.

More notable is the sporting-failure multiplier. A squad built at that price that fails to score across three matches — and, in the piece's own counterfactual, sits bottom on zero points — creates exactly the conditions under which asset-impairment and deadweight-contract risk begin to form: players bought at peak market prices on long deals whose book value outruns their performance-implied value.

But be clear: three matches is a statistically negligible sample for judging a transfer portfolio. Log it as a watch item, not a verdict.

The structural point is that clubs concentrating capital in a single window create a compressed integration-risk window. Every early negative result amplifies both media pressure and the risk that the market re-prices the incoming assets.

---

Now the most interesting methodological part: the top of the table.

This is the most robust result in the entire original piece, for a simple reason — it rests on results, not opinions. Manchester City and Arsenal won all three, and neither was party to any disputed decision. This is the one section where the corrected and real tables must coincide, and they do.

In other words, at this stage the elite tier's results are being generated independently of officiating variance. That is a useful control group sitting inside the piece's own method.

Every system has a blind spot. Where it errs is the bigger question.

The blind spot here is at the bottom. Nottingham Forest gaining two points and five places (via Neco Williams' disallowed goal) is the exact mirror image of a single fan-voted decision — meaning the entire bottom-of-table reconstruction rests on one point of support. That is a single-point-of-failure dataset. One incident. One vote. Five places.

Remember the context: at matchday three, the whole league is compressed into a few points. That makes any "corrected" decision reshuffle multiple positions at once. Forest's five-place jump is a structural artefact of early-season table density, not evidence Forest are materially underrated. The corrected-table format is maximally dramatic and minimally predictive at precisely this moment.

Manchester United's case is subtler still. They sit bottom-half in the real table but land top ten in the corrected one — without gaining a single point. They improve purely because others drop. That is an important methodological quirk: a corrected table can lift a team's position without improving its results.

And Brentford's rise to third comes entirely from a counterfactual penalty award. In a three-match sample, the league landscape is that sensitive to single incidents.

At this point I must separate a concept the original piece conflates.

The Williams incident is a threshold dispute, not a factual error. The league's stated basis was handball. The objection was that the footage was inconclusive and not clear and obvious. These are two different rule layers — factual determination versus intervention threshold. Fusing them is a decision-grade flaw in the whole methodology.

A VAR error in the factual sense — a miscalibrated offside line, a wrongly identified player — is objectively checkable. A VAR error in the threshold sense — should VAR have intervened — is a judgement call that crowd voting is structurally unsuited to adjudicate. If any "corrected" outcome in the piece rests on threshold disputes rather than factual errors, that table is not a correction — it is a second opinion, weighted by supporter sentiment.

And a documented bias applies to vote-based adjudication: fans of the disadvantaged club vote in greater numbers and with more motivation than neutrals. No weighting mechanism or neutral panel is described. No vote counts, no participation thresholds, no demographic breakdown. In other words, the statistical power of the corrected table is unknown.

---

This is the counterintuitive point I want to dwell on longest. The piece's real error is not in the VAR calls. It is that it lets a sentiment dataset wear the visual and rhetorical form of an objective league table — and in an information ecosystem where such tables are screenshotted and shared faster than official documents, the risk of misinformation exceeds the risk of any sporting outcome described.

But fairness is owed here. The very table, through its structure, accidentally exposes something valuable: at matchday three, the league is compressed enough that one decision can redraw multiple fates. That is not VAR's fault. It is a property of the early-season table.

Collapse never comes from a single shock, but from the misalignment of spatial layers. Here, the misaligned layers are not on the pitch. They sit between the real table, the hypothetical table, and the fan's vote.

The time context also bears repeating. The VAR meta-narrative is long-lived — it has run since 2026/20. But this specific corrected-table artefact is short-lived: it is superseded by the next matchday. One is a long-term story, the other a fast-consumption product. Blending them is the fastest way to misread both.

On Tottenham, I will close with a governance note. The structural preconditions for dressing-room stress are present — heavy investment plus zero goals. But no behavioural symptom is reported. Conditions for friction are not evidence of friction. An early-season managerial change at a club that just spent £300m would be an extreme governance signal, implying recruitment and coaching decisions were never aligned. But the dataset does not say who led recruitment — and that is the decisive variable.

---

Overall, the risk rating lands at medium-to-high, and the basis for that does not sit on the pitch. Sporting risk is contained: a three-match sample, a stable top. The rating is elevated by epistemic risk in the method.

Before drawing the pass, read the position of the gap. In this table, the gap is that not a single figure tells us how many people voted.

So the question to carry into the next matchday is not whether Tottenham were unfairly placed. It is this: when the table compresses tightly enough that one incident shifts five places, will we keep reading the crowd's vote as though it were an offside line — or start reading it for what it is: a collective opinion with reference value and a very short shelf life.