How we estimate your marks

Every number Leibniz shows you starts from a single, honest question. Not “how much have you done?” and not “how hard have you tried?” It’s this:

We call the answer your demonstrated mark. It isn’t a prediction about the person you might become by exam day, and it isn’t credit for effort. It’s a measurement of where you stand right now, on the exam’s own terms: the mark you have already demonstrated you can produce.

That framing matters. Effort is easy to count and comforting to report, but exams don’t reward effort. They reward being able to do the questions. So everything you’re about to see is built to answer that one question as faithfully as possible, and to update the answer every time you practise.

Before you attempt a single question, the course already has a shape.

For every course on Leibniz, we’ve gathered the official record: real exam questions, written by the examiners, across many years. Picture each one as a point on a map. How hard the question is sets its place along the bottom. How much technique it demands (the craft of setting up the problem, not just the topic it names) sets its height. Questions that belong to the same topic settle into the same neighbourhoods.

The lake matters because it keeps us honest. An easy win is still visible on this map. It just lives in the shallow end, far from where the exam’s marks concentrate. By anchoring everything to official questions, we make sure “doing well on Leibniz” always means “doing well on the kind of thing the exam actually asks.”

When you practise, you don’t fill in a form. You disturb the water.

Each time you attempt a question on Leibniz, we place it on the same map: a new point, in a different colour. Your attempt lands near the official questions it resembles: the ones on the same topic, at a similar difficulty, calling for similar technique.

But a droplet doesn’t sit still. Your result ripples outward. If you handled a question well, the official items in its neighbourhood (the exam’s own versions of that skill) are read a little more hopefully. If you struggled, they’re read a little more cautiously. The closer an official item sits to your droplet, the more your result colours it.

This is how we can talk about parts of the course you’ve barely touched directly. You may never have sat that exact 2019 exam question, but you’ve done things that look like it, and those droplets have already reached it.

A mark earned in deep water says more than a mark earned in the shallows.

Two students can both answer a question correctly and tell us very different things. If the question was routine, the success is real but quiet: the exam was always going to ask harder things. If the question sat in the deep end (genuinely difficult, demanding real technique), that success rings loudly, because it speaks directly to how the exam’s top marks are earned.

This is difficulty alignment and technique alignment: your evidence is always weighed against where it landed on the map. It keeps the demonstrated mark honest in both directions: you can’t inflate it by farming easy questions, and you’re never under-credited for doing hard things early.

Who you were three months ago still counts. It just doesn’t get to interrupt.

Students change. Skills grow, rust, and grow again, so a reading of “where you stand right now” has to let recent work count for more than old work. Every droplet you’ve ever landed stays in the lake. A new attempt’s ripple is the one you notice; the older ripples settle slowly into the still water, quietened but never gone.

With enough droplets, your lake starts to look like the exam itself.

Here’s where the lake pays off. Because your attempts are seeded among official questions, your evidence gradually takes the shape of the real assessment: the topics the exam weights heavily collect the most droplets, the deep end fills in, and the map of what you’ve demonstrated begins to mirror the map of what the exam will ask.

We also don’t start from nothing. Every course has a known rhythm: roughly how marks tend to fall for students sitting it. That’s our starting assumption, the curve prior: a faint outline of the likely shape of things. Early on, it keeps your reading steady while there’s little evidence. As your droplets accumulate, your own evidence takes over and bends the curve toward you.

And because every student’s droplets land in the same official lake, we can make fair relational judgements: which of your topics is strongest relative to your others, where you’re ahead of where a typical student at your mark would be, and where the exam would catch you out today. Same water, same ruler, for everyone.

The course mark is the headline. The interesting story is underneath it.

Everything so far describes how we read the whole lake at once. That’s your course-level demonstrated mark. But the same map works at finer zoom. Read only the ripples inside one topic region and you have a topic-level mark. Zoom into a subtopic and you get a reading there too. It’s one map, read at different depths.

This is what turns a number into a diagnosis. “You’re at 82 in this course” is a fact. “You’re at 82 because Calculus is at 74, and Calculus is at 74 because Integration is at 66”. That’s a plan. Students see exactly where their marks are made and lost; teachers see the same layered picture for every student in the room.

The picture is never finished, and that’s the point.

Put all of it together (the lake, the droplets, the weighting, the fading, the starting shape) and you get a system that is alive in a specific sense: every attempt makes it a little sharper, and it always reflects who you are most recently.

A student with five attempts behind them gets an honest answer with wide margins. A student with five hundred gets a sharp one. Both are the same system telling the truth at the resolution the evidence allows, and the first student becomes the second just by practising.

A faithful, living map of where you stand is valuable on its own. But its real purpose is what comes next: because we can see which threads of the course are weakest (and which are cheapest to strengthen), we can point you at the practice that moves your mark the most. That’s a system of its own, with its own page.

A note on detail: this article is the conceptual picture, deliberately simplified. The production system carries more dials than shown here. The exact machinery is part of how Leibniz works, and stays ours.