UPPSC PCS Prelims Marking Scheme: The Math Behind Negative Marking, Blind Guessing, and the Real Cutoff

The UPPSC Combined State/Upper Subordinate Services (PCS) Preliminary exam puts candidates through two very differently-scored papers, and most prep advice treats the headline cutoff -- "125 out of 200" -- as if it simply meant 62.5% accuracy. It doesn't, and the gap between what the raw cutoff looks like and what it actually requires is entirely explained by UPPSC's 1/3 negative-marking rule.

1. The real marking scheme, paper by paper. GS Paper I (General Studies) carries 150 questions for 200 marks -- 1.33 marks per correct answer -- with a deduction of one-third of that (0.44 marks) for every wrong answer; blanks cost nothing. CSAT (GS Paper II) carries 100 questions also worth 200 marks total -- 2 marks per correct answer, 0.67 deducted per wrong answer. Only GS Paper I counts toward the Prelims merit list; CSAT is purely qualifying, cleared at 33% (66 of 200 marks) regardless of how it compares to anyone else's score.

2. Why blind guessing is mathematically neutral, not "risky." Every UPPSC objective question offers four options, and the one-third penalty is calibrated so a pure random guess among four choices breaks exactly even, on both papers, regardless of how many marks the question is worth: a 1-in-4 chance of +m against a 3-in-4 chance of -m/3 nets to (1/4)m - (3/4)(m/3) = m/4 - m/4 = 0. This isn't a coincidence of GS Paper I's specific 1.33/0.44 numbers -- it holds for CSAT's 2/0.67 numbers too, because it's a property of "1/3 penalty on four options," not of the mark value. Practically: leaving a question blank out of fear of negative marking throws away a free coin flip. The moment you can confidently rule out even one of the four options, the guess among the remaining three turns strictly positive-EV: (1/3)m - (2/3)(m/3) = m/9 > 0.

3. What the real cutoff actually requires, in correct answers. UPPSC PCS Prelims 2023's finalized General-category cutoff was 125 out of 200 on GS Paper I -- a mark that reads like 62.5% accuracy but isn't. If a candidate attempts all 150 questions and gets x right (and 150-x wrong), their score is 1.33x - 0.44(150-x) = 1.78x - 66.7. Solving 1.78x - 66.7 = 125 gives x ≈ 108 -- meaning roughly 108 of 150 questions correct, about 72% accuracy, is needed to match that cutoff on a full attempt: nine and a half points higher than the naive mark-based reading. Every wrong answer doesn't just fail to help -- it actively erodes the buffer a correct answer built, which is why the real accuracy bar sits well above the raw percentage.

4. CSAT's 33% floor is a much lower bar by comparison. Running the same arithmetic on CSAT's 2/0.67 scheme against its 66-mark qualifying floor: 2x - 0.67(100-x) = 66 solves to x ≈ 50 -- just half of CSAT's 100 questions correct, on a full attempt, clears the qualifying hurdle. That's 22 percentage points easier than GS Paper I's effective cutoff bar, consistent with CSAT's design as a qualifying filter rather than a competitive section -- useful to know if time pressure forces a choice between double-checking GS Paper I answers and attempting every last CSAT question.

Camino Prep's UPPSC PCS question bank mirrors this exact split -- 150 General Studies questions, including a dedicated Uttar Pradesh-specific GK section no generic all-India PCS resource replicates, and 100 CSAT questions -- so practice scoring matches the real 1.33/0.44 and 2/0.67 marking schemes question-for-question, not a generic right-or-wrong tally. Check the exam calendar for the next Prelims cycle's dates before planning a study timeline around it.