A Brilliant alternative for when the streak isn't becoming ability

The 200-day Brilliant streak that still can’t sit a calculus problem

Here is the person this page is written for. She is a product manager, or a nurse, or a software developer who never did the maths module; it does not much matter. Two hundred and some days ago she opened Brilliant on her phone, and she has opened it every morning since, ten minutes over coffee, tap, drag, aha, close. The streak counter is the longest-running commitment in her life outside her job. She has finished the calculus course, most of the probability one, and a good deal of the logic track for fun.

Then a colleague slides a whiteboard problem across, a rate of change with a chain rule buried in it, and asks her to have a go. She knows what a derivative is, vividly. She can picture the tangent sliding along the curve. And she cannot start. Not because Brilliant taught her wrongly, but because at no point in two hundred days did anything require her to produce a derivative cold, from a blank page, a month since she last saw the idea. The practice sets that would have asked exactly that were there, hints off, but always optional, and the streak was just as happy with a lesson.

If you recognise her, the rest of this page is about what happened, why it is not her fault, and what the app would need to look like for a streak to turn into ability.

Why Brilliant’s interactive lessons feel like learning

The tap-through format is not a gimmick. When you drag a slider and watch a Riemann sum tighten around a curve, the idea arrives through your hands, and that beats a lecture for building a mental picture. Each lesson is short, the feedback is instant, and the questions are pitched so that you almost always get them with a nudge. The whole experience is engineered for the small, reliable hit of I get it, and it delivers that hit with remarkable consistency.

That consistency is exactly the problem, and it takes a moment to see why. Learning that sticks is effortful: you struggle, retrieve, fail a bit, and try again without the scaffolding. A lesson designed so that you almost never fail is a lesson designed to feel good, and the feeling of fluency it leaves behind is mostly recognition. You recognise the derivative when the app shows it to you. That is a real and useful thing to have. It is just not the same thing as being able to produce one.

Three things are thin, and they are thin by design, because a product built for curiosity does not need them. Placement is a light touch: sign-up asks about your level to pick a starting path, but nothing probes concept by concept whether your algebra can carry a calculus course. There is no mastery gate between one concept and the next: on Premium you can jump to any lesson in any course, so you can carry a gap forward indefinitely and never be told. And the return to old material is opt-in: Brilliant’s practice sets do use spaced repetition on your weak areas, and they drop the hints, but finishing a course never requires them, and the streak is satisfied by a lesson just as well. The streak keeps you busy every day. It never once measures what you can now do.

What Brilliant is genuinely best at, and who should stay

It would be dishonest to pretend the product is anything other than superb at its own job. For breadth, nothing in the space comes close: the same subscription wanders from group theory to neural networks to physics, and every course is made to the same high standard. The interactive explanations are, in our view, the best intuition-building material available to a general audience, and the visual design is a pleasure in a category that mostly is not. If we are picking a first exposure to a hard idea for a curious adult, we would still pick it.

So stay, without guilt, if any of these describe you. You are learning recreationally and the streak is its own reward. You want to sample several fields rather than go deep in one. You already have working maths from school or university and want to keep it warm and enjoy it, rather than rebuild it. Or you are a teacher looking for a beautiful way to show a concept rather than a way to drill it. In all four cases the thing we are about to describe would feel like a narrower, more demanding product, because it is.

What changes when a map, not a streak, decides your next Brilliant-style lesson

Consiliae keeps the part of the format that works: short explanations, then problems you solve yourself, immediately. What it adds is the skeleton the streak was standing in for. It starts with a placement scan, which is the cold test our product manager was never made to take: up to 25 adaptive questions, about 12 minutes, no timer, and it is free and needs no account. The result lands on a prerequisite knowledge graph you can see: every concept from foundations through school maths into early university, and the arrows that say which one depends on which. You pick a goal, the route to it is computed through the graph, and the concept in front of you is there because the map says it is next, not because it looked interesting in the catalogue.

Then two things the streak never enforced. Each step is gated: you move on when a mastery check says the concept holds, and when it does not, the map routes you back to the prerequisite that failed rather than letting the gap ride along. And spaced repetition is built in, so the chain rule you passed in March is scheduled back just before you would have forgotten it, and again after that, until it stays. Retention stops being something you hope for and becomes something the system measures.

The honest cost of that trade: our catalogue is maths only, for learners 16 and over, our problems are chosen for mastery rather than delight, there is no built-in tutor to talk you through a step, and some days the map will hand you a foundation you would rather skip. We have XP, a weekly league and a streak too, but the streak comes with Recovery Days and never decides what you study next. If you want a fireworks show of ideas across all of science, that is the other product, and it is very good. If you want, in six months, to be handed a rate of change problem and simply do it, the map is the shorter road, and you can walk the first stretch of it free.