Almost one in four boys took a top grade in maths. The grade that decides a life is a 4.

Maths did the lifting this morning. Almost one in four boys in England took a grade 7 or above in it, 22.3% of entries reached that mark overall, and the strength of the subject helped push top grades for 16-year-olds to 23.1% — a record outside the years when exams were replaced by teacher assessment. Nearly 700,000 pupils sat them. GCSE results, England, 20 August 2026

It is a good morning, and the cohort earned it. These are the pupils who were in Year 5 when schools closed, and every leader quoted today was right to say so.

But every headline figure above describes the top of the paper. The grade that decides what a 16-year-old does next is not a 9. It is a 4 — and it is the number the headlines skip.

Twenty-three take a 7. Twenty-nine miss a 4.

Below is every grade awarded to 16-year-olds in England, drawn as a hundred. The amber block is the one in the headlines. The dark block is the one that decides who resits.

• 23 — grade 7 to 9
• 48 — grade 4 to 6
• 29 — below a grade 4

More grades fell below a 4 than reached a 7. Both numbers come from the same release, and only one of them is a story. Pepe Di’Iasio of the Association of School and College Leaders made the point in his response: too many pupils fall short of a grade 4 in English and maths at sixteen, fall short again in post-16 resits, and that shortfall is bound up with the number who end up not in education, employment or training.

For a school, this is not a statistic. It is a list of names. In a fortnight they start post-16 courses carrying a resit — and the Year 11 coming up behind them holds the same proportion, already identifiable, still teachable.

The gap that led, and the bigger one that didn’t

This morning’s gender story was maths: 24% of boys at grade 7 and above against 20.5% of girls. It is real. But set the two compulsory subjects next to each other at the grade that matters, and the shape of the day changes.

The largest gender gap in either compulsory subject today is boys in English, and it is five times the size of the maths one. A third of boys did not reach a grade 4 in the subject every sixth form, college and apprenticeship asks for. A school reading only this morning’s coverage would spend the year working on the smaller of its two problems.

The distance between a grade 4 and a grade 7 is usually visible inside one sentence. That is where the teaching has to happen.

Articulate Writers, Confident Schools

The only line that moved backwards

One figure in the whole release went down. Girls reaching a grade 4 in maths fell from 71% to 70.7%, while boys rose to 72.8%.

Be careful with it. Three tenths of a point across one cohort is small, and it moves for reasons that have nothing to do with teaching — entry patterns, a paper that suited one group, ordinary year-to-year noise. One year is not a trend, and anyone presenting it as one is overreaching.

The reason to look at it anyway is that it is the only line moving down while everything else moves up, and it sits on the boundary that decides post-16 options rather than the one that decides a personal statement. That is worth a look at your own data, which is the only place the question can actually be answered. In most schools the pupils on that boundary are also the pupils with the weakest attendance — worth testing before anyone designs a maths intervention for a problem that is not a maths problem.

Three questions for the first department meeting

Results day is a poor time to decide anything. September is not. These are the three questions that separate a department that improves its grade 4 rate from one that reports on it.

1. 01 Who was actually on the boundary? Not the predicted 4s. The pupils whose mock sat within a few marks either side of it, listed by name before the first lesson of September. Curriculum
2. 02 Which misconception produced the error? A score says a pupil got it wrong. It does not say whether the fluency failed or the reasoning did, and those need different lessons. Evidence
3. 03 Would every teacher give the same next step? Ask three teachers what this pupil needs on Monday. If the answers differ, the intervention is a timetable change rather than a teaching one. Standard

Most assessment tells you what a student scored. Very little tells you which misconception produced the error, which is the only information that changes Monday.

Fluent Mathematicians, Confident Schools

Count the twenty-nine

None of this takes anything away from today. A cohort that lost most of Year 5 has just produced the strongest set of higher grades outside the pandemic years, and the schools that got them there should say so loudly.

But the twenty-nine in every hundred are not a national statistic. They are a list of names, and the next twenty-nine are already in your building, in the year group that starts back in a fortnight. The department that knows which of them could not start, rather than which of them were borderline, is the one whose figure moves next August.