What actually decides a UK degree class
Ask what separates a 2:1 from a 2:2 and the intuitive answer is marks. Near a boundary — which is where the question always gets asked — that is only half true. The other half is two administrative decisions carrying no marks at all: how many years count and what they weigh, and what the regulations do in the last half mark before a line. This guide runs one transcript through both, using rules three UK universities publish in their own words.
The transcript
Year two, six 20-credit modules: 52, 55, 58, 60, 61, 62. That is 120 credits and a credit-weighted mean of exactly 58. Final year, a 40-credit dissertation at 65 plus four 20-credit modules at 58, 62, 52 and 61: 120 credits again, weighted total 7,260, mean exactly 60.5. Nothing dramatic, no resits, no missing marks. A perfectly ordinary borderline student.
The boundaries are the settled part
Three institutions, three separate documents, one line. The University of the Highlands and Islands awards a First at "an average (mean) mark of 70% or more," then 60–69, 50–59 and 40–49. Manchester Metropolitan lists "70-100% (First Class)", "60-69%", "50-59%" and "40-49%". Leeds publishes 70.0 and above, 60.0–69.9, 50.0–59.9 and 40.0–49.9. If anything here is a convention you can lean on, it is this one.
The weighting is not
Manchester Metropolitan's calculation page describes "Level 5 contributing 25% towards the final mark and Level 6 contributing 75%", and an alternative in which "only marks from a your final year count towards your degree outcome". Leeds uses a "1:2 weighting of grades between programme years 2 and 3". UHI classifies on the final level's credits. Same transcript, three real schemes:
| Weighting | Degree mark | Exact comparison |
|---|---|---|
| 25 / 75 (Levels 5 and 6) | 59.875 | Lower second (2:2) |
| 1:2 between years two and three | 59.667 | Lower second (2:2) |
| Final year only | 60.5 | Upper second (2:1) |
The whole spread is 0.833 of a mark, and the class flips inside it. One set of exams; an enrolment decision made years earlier setting what they were worth. It is the most useful thing to look up in your own regulations, and the reason the calculator takes weights as editable ratios rather than baking one scheme in.
The half mark at the bottom
Take the 25 / 75 result, 59.875, and read the three published borderline rules against it. Manchester Metropolitan: "Rounding is never used to determine classification of an award" — the mark is 59.875, the boundary is 60, the answer is a 2:2. UHI: "Each student's average (mean) mark is rounded to the nearest whole number," with the worked example that "a student would need to get an average (mean) mark of 59.5 or above to be rounded up to 60" — 59.875 rounds to 60, so a 2:1. Leeds: "Classification averages within 0.5 of the next class boundary will be awarded the higher classification" — the gap is 0.125, so a 2:1.
Two of those three rules are worded very differently and do exactly the same arithmetic. Rounding to the nearest whole number promotes a mark when it is at least half a point below a whole boundary; a within-0.5 rule promotes on precisely the same condition. They diverge only if a boundary stops being a whole number — on a 59.7 line, a 59.4 average rounds to 59 and misses, but sits 0.3 away and is promoted. The calculator keeps them as separate options because your regulations will use one wording or the other, and because the boundary table is editable.
Credits are not decoration
Average the five final-year marks as if they were equal and you get 59.6. Weight them by credit, as UHI spells out — "40-credit modules will be counted as two instances of the same mark" — and you get 60.5. That dissertation is worth 0.9 of a mark on its own. Run the credit-blind version through the 25 / 75 scheme and the degree comes out at 59.2, which is a 2:2 under all three borderline rules, promotion or no promotion. The credit weighting is what put this student within range of the line in the first place.
If the final year is still ahead
Carrying a 58 into a final year worth 75 of the 100 weight, here is what that year has to average, target by target:
| Target | Boundary | Final-year average needed |
|---|---|---|
| First | 70 | 74 |
| Upper second (2:1) | 60 | 60.667 |
| Lower second (2:2) | 50 | 47.333 |
| Third | 40 | 34 |
The student above finished on 60.5 and needed 60.667 — short by 0.167 of a final-year mark, which is the same 0.125 of a degree mark seen through a 75 percent weight. That is the honest scale of a borderline: not a grade, not a module, a rounding error's worth of coursework. It is also why the number worth knowing in second year is not your average but the average you still need, which the grade needed calculator solves one course at a time.
What none of this models
Weighted averages are only part of a real regulation. Manchester Metropolitan runs a second method alongside the first — a lower average paired with a minimum volume of credit already in the higher band — and applies "whichever method is most beneficial". Others discount weak modules, condone failed ones, or leave an exam board discretion inside a borderline zone. None of that is arithmetic and none of it is portable between universities, so the tool does not pretend to it. Get the number here; get the verdict from your regulations.