Topper Story: Arjun Mehta Scores 115 in JEE Main Maths

Arjun Mehta from Delhi scored 115 out of 120 in JEE Main Mathematics in 2025, placing him in the 99.7 percentile for the section. What makes Arjun's story particularly instructive is that he started preparation in Class 12 with a genuinely poor Mathematics foundation — consistent 55% in school tests — and rebuilt it systematically over fourteen months. His turnaround is not a story of natural aptitude but of a study system that prioritised the right things in the right order. This is that system in full.

The Diagnostic Start: Knowing Before Studying

Arjun's first month was pure diagnosis. He solved one previous-year JEE Main Mathematics paper without preparation, tracking every error and categorising each by chapter. The result was a precise picture of his weaknesses: he was strong in Trigonometry and Algebra but had almost no foundation in Calculus, and Coordinate Geometry was inconsistent. "Most students start studying without knowing what they don't know," he explains. "I knew my weaknesses before I started, so I never wasted a week on something I already understood."

His priority order came directly from this diagnostic: Calculus first (biggest gap + highest weightage), then Coordinate Geometry (inconsistent + high weightage), then Algebra and Trigonometry (already partially solid, just needed systemising). This evidence-based ordering is something every student can replicate by solving a past paper diagnostically before beginning formal preparation. Take a free JEE mock to run your own diagnostic.

Rebuilding Calculus From Zero

Arjun spent his first four months exclusively on Calculus. He started with the conceptual foundation — reading his NCERT Calculus chapters actively, not passively — then moved to the standard textbook problems, and only then to JEE-level PYQs. "The mistake students make in Calculus is jumping to JEE difficulty problems before the concept is solid," he notes. "Then they memorise solutions without understanding why, and the next variation stumps them."

For integration in particular, Arjun built a classification tree: when he saw an integral, he would identify its type (rational, trigonometric, product requiring IBP, or radical) in five seconds and select the corresponding technique. He trained this classification skill by solving 500 integrals of progressively increasing difficulty, each classified before being solved. By month four, he could classify any integral in under three seconds. Our definite integration guide and differential equations guide cover the Calculus topics he mastered in this phase.

The Coordinate Geometry Phase

Months five and six were Coordinate Geometry. Arjun's approach was formula-card based: for each conic section, he made a double-sided card with all key formulae on one side and five standard problem setups on the other. He would practice by reading the setup and writing the solution approach before looking at the worked solution. "The goal is to have the setup trigger the method automatically," he says. "Once that happens, each coordinate geometry question takes under three minutes, regardless of complexity."

Exam-Day Execution and the Three Mistakes

Arjun entered the exam having taken thirty full-length mocks. His exam-day strategy: first pass (45 minutes) to collect all instantly recognisable questions; second pass (60 minutes) for medium-difficulty questions requiring two to three steps; third pass (30 minutes) for the hardest remaining questions. He allocated his final 15 minutes to checking answers and correcting careless errors. "I made exactly three mistakes on exam day, all careless errors in my second pass," he recalls. "The mock test system had eliminated conceptual errors completely."

His advice: "Diagnose before you study. Rebuild gaps in order of importance. Train speed through deliberate classification practice. And use mocks as data, not ego checks." For students building the same system, our Math strategy guide, chapter weightage analysis, and calculus guide form the complete preparation toolkit.