FAQCAT PreparationWhat is the best strategy for CAT DILR when s
cat

What is the best strategy for CAT DILR when set selection is challenging?

Claude's answer·3 min read·568 words·✓ verified Mar 2026

Set selection is the skill that separates 99 percentile DILR scorers from everyone stuck at 85-90, and you can train it deliberately. The first 4-5 minutes of the DILR section should be pure scanning, zero solving.

Read every set's setup, count the conditions, judge the data density, and mentally rank all four sets before you touch a single question.

How to Scan and Rank Sets

Your scan should answer one question per set: "Can I see a path to the first answer within two minutes of starting?" Sets with clean tabular data, fewer than six conditions, and no circular dependencies are almost always faster to crack than those with nested logic or 10+ constraints. Scheduling sets and bar/line graph sets tend to have predictable solution structures. Heavy combinatorics and multi-variable matching sets tend to punish you with time sink.

Set TypeTypical DifficultyAvg. Solve Time
Bar/Line GraphLow-Medium8-10 min
SchedulingMedium10-12 min
Matching/ArrangementMedium-High12-15 min
Game Logic / NetworkHigh15+ min

Committing and Cutting Losses

Once you start a set, give it a hard 8-minute checkpoint. If you have cracked at least one question and see a clear path forward, stay.

If you have zero answers and no breakthrough, abandon immediately. Many students waste 15 minutes on a set they will never solve, then panic-rush the remaining time.

That is the single most common reason people score 85 when they have the aptitude for 95.

Post-CAT 2024, IIM Ahmedabad has leaned toward shorter, more elegantly constructed sets rather than brute-force data dumps. The difficulty is now in fast solvability judgment, not computation volume.

This actually rewards practiced test-takers more than raw intelligence.

Score Benchmarks to Target

Knowing what "good enough" looks like keeps you calibrated under pressure. Solving 2 sets fully with 90%+ accuracy typically lands you around 80-85 percentile.

Three sets solved cleanly pushes you to 95+ percentile. All four sets, with high accuracy, is the 99+ percentile zone.

For IIM ABC calls, you generally need to be above 99 in DILR, meaning three clean sets in 40 minutes with almost no errors is the realistic bar. This is hard, don't pretend otherwise.

Building Pattern Recognition

Practice sources matter. Use 2IIM's LRDI problem sets, IMS LRDI playlists, and official CAT previous year papers (2017-2024) as your core material.

Drill at least 100 sets across all five common types: scheduling, network optimization, bar/line graphs, matching, and game-logic. Each type has a standard first-move strategy.

Learn the first move for each type, and your scanning speed during the actual exam jumps significantly.

Mock Analysis is Where You Actually Improve

After every mock, revisit every set you abandoned. Was it actually hard, or did it just look intimidating at first glance?

If your skipped set was solvable in 10 minutes, that is a calibration error, and calibration is what you are training. Track your set-selection accuracy across mocks the same way you track your answer accuracy.

Coaches at IMS, Career Launcher, and 2IIM all emphasize this, and they are right.

Pro Tip: After each mock, write down the exact reason you skipped each abandoned set, then go back and solve it to check whether your read was correct. Four weeks of this habit will sharpen your solvability radar faster than any other drill.

Free · audit-verified data
Comparing two colleges?
Side-by-side fees, placements, audit status, and our verdict for any 2 or 3 of the 58 colleges in our database.
Tool · Free · Personalised

Got your answer? Now find which colleges fit YOUR profile.

The hard question isn't 'is this true' — it's 'which 5 colleges should I actually apply to'. Drop your score in 3 mins and Claude scores you against 20+ B-schools with real conversion %.

CHECK MY MBA CHANCES →
✓ 3 minutes✓ Real conversion %✓ 20+ schools
Related questions
AI