SWAदेSH
2026-02-20

PYQ Trend Analysis: What UPSC Actually Asks

The syllabus tells you what's in scope. It doesn't tell you what's likely. Those are different questions, and only one of them can be answered with data — by looking at what UPSC has actually asked, year after year.

Subject weightage isn't uniform

Across three decades of Prelims papers, some subjects — Polity, Geography, Economy, Environment — show up far more consistently than others. This isn't a secret, but most aspirants still allocate revision time roughly evenly across the syllabus rather than weighting it toward what's historically asked most. A subject that appears in 15-20 questions a year deserves more repeated revision than one that appears in 2-3.

Topics within a subject aren't uniform either

The same unevenness holds one level down. Within Polity, constitutional bodies and amendments come up repeatedly; within Environment, biodiversity and climate policy dominate over more niche sub-topics. A topic-wise heatmap makes this visible in a way that skimming the syllabus list never does.

Difficulty trends shift over time

UPSC's average question difficulty isn't static — some years lean harder on factual recall, others lean harder on applied reasoning across multiple statements. Tracking this year-over-year matters for how you practice, not just what you study: if recent years trend toward multi-statement, elimination-heavy questions, that's the skill worth drilling, not just fact recall.

Let the data set your revision order

SWAदेSH's Prelims Analyser builds exactly this picture from every Prelims question since 1995 — subject distribution, topic-wise heatmaps, difficulty trends, and year-over-year change — so revision time can follow the historical record instead of guesswork. Explore the Prelims Analyser.

See this weighting play out directly by browsing a free sample of Prelims PYQs, and read UPSC Prelims vs Mains — What Actually Changes for how this same historical-pattern logic applies differently on the Mains side.

← Back to Blog