What the percentage measures
A qualification boundary is the minimum mark needed for a grade in a particular exam series. We divide it by the qualification maximum and multiply by 100. That makes the numbers easy to put side by side.
But the marks themselves are earned on different assessments. A paper can make individual marks difficult to obtain, with a relatively low percentage reaching a high grade. An accessible paper can need a higher threshold to maintain the grade standard. Ofqual explains that boundaries are decided after marking and take the evidence from that series into account.
Ofqual: what happens after your exams ↗
Why a maths mark and a drama mark are different things
Subjects vary in their questions, mark schemes, coursework, performance assessment and component weightings. Their candidates also differ. Comparing 70% in two subjects does not give the same student an interchangeable grade.
Board comparisons have the same limitation. The qualification code, entry option, tier and maximum all matter. There is no evidence here that a particular board is “giving grades away”. A numerical league table of thresholds is still interesting, provided we keep that distinction visible.
Ofqual’s guide for schools and colleges ↗
What Ofqual means by subject difficulty
Ofqual’s inter-subject comparability work asks a different question: how do students’ grades in one subject compare with their attainment elsewhere? Its analysis uses prior attainment and statistical models, including Rasch modelling, to investigate relative statistical difficulty.
Those models have assumptions. Motivation, teaching, resources, subject choice and differences in what subjects assess can affect the results. Ofqual cautions against treating a statistical difference alone as evidence that a subject is harshly graded.
We have not converted that research into a “hardest subject” score. A defensible chart would need the underlying estimates, their uncertainty and a carefully matched year and grade. Raw percentage rankings cannot substitute for that work.
Read Ofqual’s inter-subject comparability research ↗
Why some years need their own warning
Summer 2020 and 2021 did not have the normal national exam series. Grading in 2022 was transitional. England returned towards pre-pandemic standards in 2023 with grading protection, so a long historical line should not be read as a series of interchangeable experiments.
The volatility tool uses 2024–2026 only. The 2027 forecast requires all three of those years and uses earlier eligible data for retrospective testing where stated. Even then, an unchanged maximum and code cannot guarantee identical assessment conditions or an unchanged syllabus.
Ofqual’s discussion of the grading context ↗
How to use these tools
Use the comparisons to compare published thresholds. Check your actual qualification and tier when calculating marks. Historical boundaries show the range of past results, so keep a margin in any revision target. A 2027 guess is never a guarantee.