Same Marks, Different Board Totals: Comparing 1100 Against 1200 Fairly

Two Cousins And A Running Argument

At every wedding, every Eid dinner, every visit from the elders of the family, the same comparison arrived with the tea: 985 beats 940, the boy did better, case closed. The argument ran, by my count, across eleven gatherings in two years before one division ended it. The two students at the centre of it are cousins. One sat her intermediate exams under a scheme with a total of 1100 and scored 940.

The other sat his under a scheme with a total of 1200 and scored 985. Put both scores through their own totals and the ranking turns over: 940 marks out of 1100 is 85.45 percent, and 985 marks out of 1200 is 82.08 percent, so the cousin with fewer marks holds the stronger result.

The boy himself, to his credit, never claimed the crown. The claim was made for him by adults doing arithmetic the lazy way, comparing the top number of two fractions and ignoring the bottom number entirely. Raw marks feel comparable because they look like the same kind of thing: two scores, both in the nine hundreds, both printed on certificates.

They are not the same kind of thing. A mark is a share of a total, and the totals differed by a hundred marks, which is nearly a whole subject. Comparing 940 and 985 without their totals is comparing two prices without asking which currency either price is in.

Two scores can look like twins and still be strangers. The total underneath each one decides what the score means.

The Conversion That Flipped The Ranking

The honest comparison takes both scores through the same operation: divide by the total, multiply by 100. The table gives the two results with their working shown, because a family argument ends faster when nobody has to trust anybody's mental arithmetic.

The gap is not a rounding whisker. It is 3.37 percentage points, a distance that on an admission merit list can separate dozens of candidates. The raw-marks ranking had the boy ahead by 45 marks; the percentage ranking has the girl ahead by a margin any examiner would call clear.

Both rankings come from the same four numbers. Only one of them respects what the numbers mean. The marks to percentage calculator exists for exactly this moment in a family's life, and the presets on it carry both totals because this argument reached my own table long before it reached this page.

At the next family dinner I wrote both percentages on one sheet of paper and laid it beside the teapot, in the spot where the argument usually started. My uncle picked it up, studied the two lines, and nodded once. A dispute that had survived two years of weddings and Eid dinners ended in under a minute, ended by a sheet of paper that did nothing except refuse to compare a numerator with a numerator.

Five Pairs Of Scores That Flip

One flipped pair could be luck. Five pairs, spread across the range, show a pattern: whenever the 1200-scheme score leads by less than about 9 percent in raw marks, the percentage ranking reverses it. The pairs below pair a score out of 1100 with a higher raw score out of 1200, and in every row the higher raw score holds the lower percentage.

The break-even point explains the pattern in one line. A score out of 1200 must reach about 109 percent of the rival's raw marks just to draw level, because 1200 divided by 1100 is 1.0909. The cousins' raw lead was 45 marks on a base of 940, which is a lead of 4.8 percent, less than half of what the bigger total demands.

Any family can run this test at the table without a calculator: if the bigger-total score leads by less than a tenth of the smaller score, the percentage ranking flips. My family met the same trap in an older costume on the night a result arrived by SMS with no total at all, the story I reconstructed in The SMS That Said 872 And Nothing Else.

Two Slopes Over The Same Marks

The chart draws both schemes as slopes across the same range of marks. The vertical distance between the two lines at any score is the size of the error a raw comparison makes at that score, and the distance grows as the marks climb.

At 600 marks the two readings differ by 4.55 points. At 1000 marks they differ by 7.58 points. This growth is why raw comparisons feel harmless for weak results and turn seriously misleading at the top of the range, exactly where families and merit lists care most. The cousins in this article scored in the zone where the gap runs past three points, and the elders ranking them had no sense that the ground under the two scores sloped at all.

When The Comparison Decides Money

Family pride is the gentle version of this error. The expensive version appears wherever a committee reads results from two boards on one sheet: a scholarship shortlist, a school admission file, a first screening of job applications by a clerk with forty files and one afternoon. I watched the expensive version nearly happen at a small private college where a relative of mine once taught.

The admission committee's first merit sheet, typed by an office assistant from the certificates, listed raw marks in a single column. The boy at the top of that sheet held 985 marks. The girl the faculty actually wanted, on the strength of her interview, sat fourth with 940, and the principal initialled the sheet before anyone asked the only question that mattered: out of how much?

The sheet was rebuilt with percentages before the offers went out, and the order changed at the top. The college lost nothing but an hour. Run the same error at the scale of a provincial scholarship, with thousands of files and no interview to catch it, and the error stops being an anecdote.

It becomes a list of students who lost funding to arithmetic nobody checked, and a list of students who hold awards a fair sheet would have given to someone else. Committees that handle mixed-board files professionally build the conversion into the sheet itself: the form asks for total marks beside obtained marks, and the percentage column is a formula, never a typing task. A form that asks for marks alone, with no total field, is a form designed to produce the cousins' argument at institutional scale.

A student on the receiving end has one defence, and it is cheap: wherever a form, a CV, or an application portal leaves room, write the percentage beside the marks and name the total in brackets. 940 of 1100 (85.45 percent) cannot be misread by a tired clerk. 940, alone, can be misread by anyone. I put that bracketed habit into every application that has left my house since the dinner-table argument ended, including my own, years before I built a calculator around the lesson.

Why Boards Run Different Totals At All

A scheme total is a bookkeeping outcome, not a difficulty rating. Boards build a total from their subject list: the number of papers, the marks per paper, and the weight of practical examinations inside science subjects. One board's intermediate scheme sums to 1100; another board, or the same board in a different scheme year, sums to 1200 because an extra paper or a heavier practical sits inside it.

Neither figure says the examination was easier or harder. The percentage exists precisely to wash that bookkeeping out of the comparison, by expressing every result as a share of its own whole.

Offices that compare results for a living do the division first and ask questions later. Universities convert every applicant's marks inside the applicant's own scheme before any merit arithmetic begins, and where qualifications cross bigger borders, the Higher Education Commission's equivalence procedures handle the reading. A family comparing two cousins needs the same discipline in miniature. The percentage is not a pedantic refinement of the comparison.

The percentage is the comparison, and the raw marks are two numbers from two different wholes that happen to share a page size. Where the comparison feeds an actual application, the components then enter an aggregate with fixed weights, a separate machine that I open up in the guide on The 10-40-50 Admission Formula.

A Fair Comparison, Step By Step

A fair comparison of any two results, from any two boards or years, runs in four steps and fits on an envelope. First, write each score as a fraction with its own total underneath; a score without its total is half a fact.

Second, divide each fraction and multiply by 100, rounding once at the end to two decimal places. Third, rank the percentages and ignore the raw gap entirely, however large it looks. Fourth, if the comparison decides something real, such as which certificate goes into an application file, note the board and scheme year beside each percentage, because a percentage read a decade later without its scheme invites the same argument all over again.

The two cousins are older now, one in a university lecture hall and one in a first job, and their certificates sit in the same family file. The sheet of paper from that dinner sits in the file too, at my insistence. Nobody apologised to either student at that dinner, and I never asked anyone to.

The change is quieter than an apology: nobody in my family has ranked those two by raw marks since, and a bare score now gets asked for its total before it gets believed. Two lines, four numbers, one division each. It is the shortest document in the file and the only one that ever settled anything.