Six CSSE score tools, tested

Every free CSSE score tool we could find asks a parent to trust a number. None of them will say where that number came from.

Tested 2 September 2026. Sources and method at the foot of this page.

Children across Essex sit the CSSE papers this month. Their raw marks mean very little on their own: a mark becomes a standardised score using figures that describe the whole year group, and those figures change every year. So parents look for a calculator, and there are several.

We tested 6. Not one publishes the cohort figures it calculates with, so nobody using them can check whether the answer is right. Where we could work out what figures were being used, they matched no year CSSE has ever published.

What the 6 disclose

What each of the 6 CSSE score tools tested does and does not disclose.
Of the 6 tools tested	How many
Publish the cohort average and spread they useWithout these, nobody using the tool can check its answer.	0 of 6
Name the year their figures describeCSSE publishes new figures every year. A score without a year means nothing.	0 of 6
Adjust for age in days, as CSSE doesTwo state that the papers are not age-adjusted. They are.	1 of 6
Mention that the two papers are multiplied by 1.5Dropping it understates every total by a third.	3 of 6

One asks the parent to type in the maximum mark for each paper. CSSE has never published that figure. We only established it — 59 for English, 60 for Maths — by reading eight years of raw candidate files, and we had it wrong twice before that.

The figures that match no year

Two of the 6 publish a worked example rather than a calculator, and they publish the same one, built on the same four numbers. Among them is an average Maths mark of 38.349. Neither page says which year that describes. Neither cites a source.

CSSE has published the raw marks of every candidate for eight entry years, and we have parsed all eight. Across all of them the average Maths mark has never once been as high as 38.349. The highest was 31.68, in 2022 entry. The figure in that example is 6.7 marks above the highest year on record.

the 8 years CSSE has publishedthe figure in the shared example

Every dot is one entry year, computed from CSSE’s own published candidate files. The shaded band is the entire range the average Maths mark has occupied since 2019 entry.

What that does to a child

Take a child who scores 36 in English and 28 in Maths — ordinary marks, close to the middle of the year group. Using the figures CSSE published for 2026 entry, that is 304.2. The qualifying score was 303. The child is over the line.

Using the figures in that shared example, the same child scores 285.6 and reads it as a comfortable fail.

Four sets of ordinary marks, scored with CSSE’s published figures and with the unsourced ones.
English	Maths	CSSE’s figures	The shared example
34	31	303.5	286.7
35	30	304.7	287.2
36	28	304.2	285.6
37	27	305.4	286.1

Across the middle of the year group — both papers within one standard deviation of the 2026 averages, so English 28 to 42 and Maths 15 to 40 — there are 112 mark combinations where CSSE’s own figures put a child above the qualifying score and that example puts them below it. The gap runs from 11.2 to 25.7 points.

What we are not saying. That is a count of mark combinations, not of children. Working out how many real children it affects would mean knowing how English and Maths marks move together across the year group, and CSSE has never published that. We are not going to estimate it, because inventing a figure is the thing this page is about.

How we checked

A claim about somebody else’s arithmetic has to survive being checked, so here is ours. CSSE standardises each paper against the year group’s average and spread, adjusts for the child’s age in days, and multiplies the two papers by 1.5.

Fed with the four numbers from that shared example, that formula returns their own printed answer: 111.8165032 for English and 102.3125409 for Maths, against the 111.8165032 and 102.3125409 they publish. So this is their method and their figures, not a version of our own making. Swapping in CSSE’s real 2026 figures is the only thing that changes.

Both sides above are run without an age adjustment, which is what the oldest eligible child scores. The tools being compared apply no usable age adjustment, and including ours would be measuring two different things and calling the difference their mistake.

Our own constants are published, per year, on how we verify the CSSE calculation, along with what happened when we checked our calculator against CSSE’s own published results for all 5,103 children who sat the 2026 test. Our worst disagreement anywhere in the range was just over half a point.

Why we have not named themWe publish a CSSE calculator ourselves, so naming 6 competitors would make this an attack rather than a finding. Everything above is checkable without the names: CSSE’s files are public, our figures are published, and the arithmetic is on this page. Any journalist who wants the list can have it.

What a parent should actually ask

Not “is this calculator accurate”, which every calculator answers yes to. Three questions that can only be answered honestly:

Which year’s figures is this using? A score is meaningless without one. The same 35 marks in English was worth 114.2 in 2020 and 100.0 in 2026 — a difference of 21.3 points on the total.

Where did those figures come from? CSSE publishes them in an annual Standardisation Report. A tool that cannot point at the document is guessing.

Does it ask for my child’s date of birth? CSSE adjusts by age in days. A tool that only asks for marks cannot be doing it.

Not affiliated with CSSE. The 6 tools were those reachable by public search on 2 September 2026; this is not a ranked survey of every tool that exists. Two more were found and could not be read, and are excluded rather than assumed. Every CSSE file this page relies on is listed on how we verify the CSSE calculation. Our own CSSE calculator is here.