CAT Attempt Strategy: Why Toppers Skip More Questions

Look through any post-result CAT thread and the pattern repeats itself. Someone who attempted 52 questions scores a 99.5 percentile. Someone who attempted 61 scores a 91. CAT toppers routinely attempt fewer questions than the average test taker, not more, and that is not luck. CAT's marking scheme rewards accuracy so heavily, and punishes a wrong MCQ guess so directly, that skipping a shaky question is almost always the higher value move. This guide breaks down the real CAT attempt strategy behind selective attempting, what verified 99-percentile scorers say about it, and a simple filter you can use to make the same call in real time on exam day.

Want to see where your own attempt-and-accuracy balance actually lands? Run a mock through the CAT score predictor and check your percentile trajectory before exam day.

The CAT Attempt Strategy Everyone Gets Backwards

Most first-time CAT aspirants walk in assuming attempts are the scoreboard. Attempt more, know more, score more. It feels intuitive, and it is almost exactly wrong for an exam with negative marking.

CAT currently runs 120 minutes across three sections (VARC, DILR, and QA), each locked to a 40-minute window, with roughly 66 to 68 questions on the full paper depending on the year. A correct MCQ earns plus 3. A wrong MCQ costs minus 1. An unattempted question earns exactly zero.

Once you sit with that scheme for a minute, the topper behavior stops looking strange. Attempting a question you are genuinely unsure about is a bet with negative expected value unless your odds of getting it right are better than roughly 1 in 4, which is exactly why accuracy-first thinking sits at the center of serious CAT preparation.

What the Scoreboard Actually Rewards

Test-prep analysis of past CAT results (not an official IIM dataset, since the exam conveners do not publish an attempts-by-percentile table) shows a consistent pattern: the number of attempts you need to reach a strong percentile drops sharply as your accuracy rises. The relationship is not linear. It compounds, because every additional wrong answer both removes a correct-answer opportunity and adds a penalty on top of it.

Notice what happens between the 90 percent and 70 percent rows. Dropping accuracy by 20 points does not just cost you marks on the questions you got wrong. It forces you to attempt 15 to 20 more questions just to land in the same neighborhood of correct answers, because a growing share of every extra attempt now works against you instead of for you.

This is the entire reason CAT preparation content built around raw attempt counts misleads people. The attempts number is a symptom of your accuracy, not a target in its own right. A quick section-wise snapshot from topper interviews:

Section-wise, toppers tend to cap their attempts rather than maximize them. Test-prep guidance built on topper interviews commonly cites ranges around 16 to 18 attempts in VARC at 80 percent-plus accuracy (our CAT 2026 VARC time allocation blueprint breaks this down section by section), and 18 to 20 in QA, best paired with a clear CAT Quant Decision Tree for choosing what to attempt first.

DILR runs narrower still, often just 12 to 14 attempts, meaning two sets solved cleanly rather than a shallow pass across four or five, a habit covered in our guide to choosing DILR sets before solving them. DILR punishes half-finished sets hardest, since a set you attempt but do not fully crack usually returns low accuracy on that entire block.

The R3 Filter: A 3-Second Decision System

Knowing the math does not automatically change your behavior under exam pressure. What actually changes behavior is a decision rule you can run in seconds, before you get pulled into a question. Call it the R3 Filter: Read, Rate, Route.

The R3 Filter works because it forces the skip decision to happen before you have invested anything, which is exactly when it is easiest to make rationally. Most aspirants who struggle with over-attempting are not bad at math. They are making the skip decision three minutes too late, after they have already sunk time into a question that Read and Rate would have flagged as low-confidence in the first fifteen seconds.

Why Your Brain Fights the Smart Move

If selective attempting is this straightforward on paper, why do so many well-prepared aspirants still over-attempt on exam day? The likely answer sits in a well-documented decision bias called the sunk cost effect: once you have spent time or effort on something, you feel pressure to keep going rather than waste it, even when quitting is the better move from where you stand right now.

Peer-reviewed research on the sunk cost effect, published in the journal Judgment and Decision Making, found this pull is driven mainly by a reluctance to feel wasteful, not by a rational recalculation of what happens next. On CAT day, this shows up in a specific, repeatable sequence:

• You read a question and it looks solvable, so you commit roughly 90 seconds to it.
• It turns out messier than expected, and you are now 90 seconds in with nothing to show for it.
• The rational move is to abandon it immediately, but the sunk cost effect pushes you to keep going "since you are already this far in."
• That second round of time is almost always the more expensive one, since it is stolen directly from a question you could have solved cleanly.

What Real CAT Toppers Say About It

This is not just a theoretical framing. It shows up directly in how verified 99-plus percentile scorers describe their own exam-day approach. Himanish Sinhababu, who scored 99.55 percentile, put it plainly in an interview: "Don't trade-off accuracy for the number of attempts. Even if your number of attempts is low, maintain at least 85 to 100 percent accuracy, and you will see a significant improvement in your scores."

Sujay Srivastava, who scored 99.94 percentile in CAT 2024, described a similar discipline around DILR specifically, timing himself to roughly 12 to 13 minutes per set so he could solve three sets thoroughly rather than spreading thin attempts across four.

His own reflection on the exam: "I got more than expected in Verbal and slightly less than expected in Quant, so overall I was quite pleased with my results." That kind of measured, section-by-section satisfaction, rather than a scramble to finish every question, is a recurring theme across topper interviews.

Common Mistakes While Learning To Skip

Selective attempting is a skill, and like any skill, it is easy to apply badly in the first few mocks. Building consistent reps against real CAT exam practice questions is what makes the Read and Rate steps fast enough to trust under time pressure. These are the mistakes that show up most often.

Frequently Asked Questions