Victoria came first. That is the least interesting thing in the data.

In the 2026 NAPLAN cycle, Victorian students recorded the highest average score of any state or territory in ten of the twenty combinations of domain and year level, and placed first or second in sixteen. Every state and territory average below is drawn from the same published file, so the ranking is not an estimate; it is arithmetic on ACARA’s own numbers.

The writing row is the one worth stopping at. Victoria is first at Years 3, 5, 7 and 9 — the only complete sweep by any jurisdiction in any domain. Its nearest rival changes as students get older (New South Wales at Years 3, 5 and 7; the Australian Capital Territory at Year 9), which means Victoria is not simply riding one strong cohort past one weak neighbour.

Alongside every comparison it publishes, ACARA reports two things the coverage almost never quotes: an effect size, which expresses a difference relative to how much students vary anyway, and a nature of difference classification, which is ACARA’s own verdict on how much that difference amounts to. A difference below 0.2 — or one that is not statistically significant — is called close to; 0.2 to 0.5 earns above or below; only beyond 0.5 does ACARA say substantially. Applied to Victoria’s lead, that verdict is blunt.

Put concretely: Victorian Year 9 students average 586.8 in writing against a national 572.0. That gap of 14.8 points sits on a distribution whose standard deviation is around a hundred points. It is a real advantage and it is worth having. It is not the chasm that a league table implies, and Victoria’s own education system should not be congratulated as though it were.

The same file that ranks the states also compares groups of students withineach state, using the same effect-size measure and the same classification. Run Victoria against itself — students whose parents hold a bachelor degree or above, against students whose parents completed Year 11 or below — and ACARA’s verdict inverts completely.

Because both quantities are effect sizes on the same scale, they can be drawn on one track — and that shared scale is the whole argument. The pale bars are what was reported. The red bars are what was in the same spreadsheet.

An analysis that only ever found fault would not be worth trusting, and this one does not. On the measure that matters most, Victoria outperforms the country twice over — it has the higher average and the narrower gap. In writing, the parental-education gap carries effect sizes of 0.95 to 1.06 in Victoria against 1.16 to 1.37 nationally. Whatever Victorian schools are doing, they are not buying their average by leaving their most disadvantaged students further behind.

That genuinely matters, and it deserves to be said plainly rather than buried under the criticism. It also does not rescue the headline. A gap of 0.95 is still classified “substantially above”; it is still roughly six times the state’s advantage over the national average. Victoria is running a smaller version of a very large problem.

And the credit comes with a qualification that should not be left out. Victoria’s advantage is in the average gap. At the edges of the distribution it does slightly worse than the country: comparing students whose parents are senior managers or professionals against those whose parents have not been in paid work for a year, Victorian students in the lowest proficiency level outnumber their advantaged peers by 6.4 to one, against 5.8 to one nationally. A narrower average gap and a sharper concentration at the bottom are both true at once, and a page that quoted only the flattering half would not deserve to be believed on the rest.

The Year 9 writing figures show what that looks like in ordinary units. These are Victorian students, in the same state, in the same year, sitting the same task.

Writing rewards a second look, and not only because this practice teaches it. It is the domain Victoria sweeps, the domain where its advantage over the country is largest, and the domain where its own students pull apart most visibly as they get older.

Follow the proficiency levels rather than the averages. In Victorian Year 3 writing, 70.5 per cent of students sit in Strong and 9.3 per cent in Exceeding: a large, tightly clustered middle. By Year 9, Exceeding has risen to 25.4 per cent while Strong has fallen to 41.0 per cent — and Needs additional support has also risen, from 3.6 to 8.8 per cent. Both tails grow. The middle empties.

This is the pattern worth naming, because it is the one a parent can actually act on. Writing is cumulative in a way arithmetic is not: a student who reaches Year 7 without control of a sentence does not simply score lower, they lose access to every task that assumes that control — the analytical paragraph, the argument, eventually the VCE text response. The divergence in the Year 9 numbers is what six years of that compounding looks like from the outside.

Everything so far is a snapshot: where students stood in one fortnight of 2026. The same file also lets you ask a harder question. Match the Year 3 cohort of 2024 to the Year 5 cohort of 2026 and you are looking at roughly the same children two years on — which shows not where they started, but how far they moved.

On that measure the table rearranges itself. Western Australia, which finishes fourth, fourth and third on the 2026 averages, records the largest average gain of any state at all three of the year-level transitions. Victoria places second across the first, then slips to equal third and fourth across the next two — its standing on growth falls away as its students get older, even as its standing on scores holds.

This does not mean Victorian schools are failing; a state can start high and stay high while still adding real value. But it does mean the league table cannot be read as a scoreboard of teaching. A high average is compatible with a state that inherits advantage and holds it — and the further through school you look, the more Victoria’s numbers resemble that description rather than the other one. It is the same caution the rest of this page has been making, arriving from the opposite direction.

That your child’s state ranking tells you almost nothing about your child. The between-state difference everyone reported is, by the measuring authority’s own classification, merely “close to” the national average. The differences ACARA does call substantial run between groups of students inside Victoria — and no league table of states will tell you which side of them your child sits on.

It also does not follow that a NAPLAN result is a verdict on a child. A single timed piece on an unseen prompt is a thin sample of a young writer, and the practice’s longer argument about what this assessment can and cannot see sits in what the writing test actually asks. The useful response to a surprising result — in either direction — is to look at the writing itself.

What the data does support is a claim about timing. If the gap widens with age in the domain that compounds hardest, then the cheapest intervention is the early one: not a term of test coaching in Year 9, but the habit of writing regularly and having it read closely, years before any of it is scored.