Study Plan · Product Guide

How the Study Plan decides what you study next

No AI guesses the ranking. It is one transparent rule you can explain to any learner. This page is the rule, in plain English.

The promise

A learner subscribes and we say: "We tell you exactly what to study next, ranked by what the exam actually asks, weighted by where you are weak." That promise is only honest if the order comes from real data. So the order is computed, the same way, every time. Here is how.

The whole idea, in one line

For every chapter, we ask two questions and multiply the answers:

Priority  =  How much the exam cares  ×  How weak you are
Importance — comes from the real question bank · Gap — comes from your practice history

Rank every chapter by that number, highest first. The top eligible chapter becomes "study this today." That is the engine. Everything below is just making those two numbers honest.

The two ingredients

Ingredient 1

Importance — "does the exam care?"

We counted every question from past exams (6,493 for JEE) and tagged each to a chapter. A chapter the exam asks constantly scores near 1.0. A chapter it barely touches scores near 0. Recent years count more than old ones. This number is the same for every learner.

Ingredient 2

Gap — "how weak are you here?"

From your practice and mock history. If you reliably get a chapter right, your gap is small. If you struggle or have never touched it, your gap is large (up to 1.0). This is the only part that is personal to you, and it works the same in any subject.

Why multiply, not add

Multiplying means a chapter only ranks high when both things are true: the exam cares AND you are weak. This is the part that makes the plan smart instead of obvious.

Exam cares a lot
Exam barely asks
You are weak
STUDY THISHigh payoff for your effort. The plan sends you here.
You are strong
Don't re-drillImportant, but you already own it. No point repeating.
The key insight for PMs
A learner's single most-tested chapter can still rank low if they already know it. And their weakest chapter can rank low if the exam never asks it. The plan refuses to waste exam-prep time on either. It hunts the green box: tested AND weak.

The one piece of "real" math: don't trust tiny samples

Say a learner answers 4 questions on Thermodynamics and gets all 4 right. Are they a master? No — 4 questions is luck, not proof. If we believed it, the plan would tell them to skip a chapter they might actually be shaky on.

So before trusting a score, we pull it toward the average until there is enough evidence. Four perfect answers might show as "68% confident" instead of "90%." After 30 or 40 questions, we trust the real number and stop pulling. One word for it: we discount scores we haven't earned yet. New learner with zero attempts? We show a dash, never an invented percentage.

Worked example — meet Priya, a JEE learner

Same exam date, five chapters. Watch how the order falls out — it is not the order you would guess.

ChapterSubjectExam caresHer gapPriorityWhy
Rotational MotionPhysics 0.701.000.70 Tested + never practiced → #1
ElectrostaticsPhysics 0.900.580.52 Heavily tested, still shaky
ThermodynamicsChemistry 0.850.320.27 4 lucky answers, discounted
CalculusMaths 0.950.230.22 Most-tested of all — but she owns it
Units & DimensionsPhysics 0.150.630.10 She's weak, but exam ignores it

Calculus is the single most-tested chapter (0.95) yet ranks 4th — Priya already knows it. Units is her 2nd-biggest weakness yet ranks last — the exam barely asks it. The plan spends her time where tested-and-weak overlap. That is the whole product in one table.

One more rule: you can't run before you walk

Rotational Motion ranked #1, but suppose it needs "Newton's Laws" first and Priya hasn't done those. The plan still lists Rotational Motion as her top goal, but "study today" skips down to the next chapter she's actually ready for (Electrostatics). It never tells her to start something she can't follow yet. The ranking says what matters; the prerequisites decide what she's allowed to start now.

"Readiness" — the one number she sees up top

A single percent: how exam-ready am I right now, weighted so that being strong in heavily-tested chapters counts more than being strong in trivia. Priya sits around 61%. It climbs as she closes high-importance gaps — which is exactly what the plan keeps pointing her at.

The three learner states

Same engine, three experiences depending on who shows up.

New / free user

Not subscribed. Sees a free outline teaser and an upgrade nudge. No real ranking until they go Pro. This is the conversion hook.

Pro, no history yet

"Cold start." We have no practice data, so accuracy shows as a dash. Order is driven by exam-importance alone, and "today" is the top chapter they're ready for. Phase: diagnose.

Active Pro user

Has practiced. Full personalized ranking, real accuracy, readiness percent climbing. Every practice session feeds back and re-ranks. Phase: master / simulate.

And the loop that connects them: a "no history" learner who practices one chapter becomes an "active" learner the very next load. The plan is never stale.

What to remember

One rule: study what's both heavily-tested and where you're weak. One personal number: your gap. One honesty rule: never invent a score we haven't earned — show a dash, or degrade to an empty state, rather than fake a plan. No AI in the loop deciding your rank, which is why every position is explainable to the learner who asks "why this chapter?"

Source of truth: issue #1127 (Study Plan v1 backend). JEE ships first; NEET, school grades, and corporate plug in later by swapping only the "Importance" ingredient — the rest of this page stays identical.