How to Attempt DILR in CAT: Scan, Choose, Commit, Switch

Here is how to attempt DILR in CAT. Scan the available sets, triage them by how clearly each offers a path to answers, commit to the strongest opportunity, and check actual progress rather than the clock. Continue while uncertainty is shrinking. Reallocate when effort has stopped producing deductions and a better opportunity remains. Spend the remaining time on the best marks still available, whether that is a full set, part of a set or single questions.

The full framework:

Two things are deliberately absent. There is no universal scan duration, and there is no universal number of sets to attempt. Both depend on the paper in front of you and on what you have shown in mocks.

This article covers exam-hall execution. It assumes you can already solve some DILR sets in practice and want to lose fewer marks to decisions under time pressure. It does not teach how to solve any particular set type.

The DILR Section Is a Resource-Allocation Problem

Understanding how to attempt DILR in CAT starts with one shift in framing. DILR is a solving test, but it is also an allocation problem. The section has fixed time, and different sets return different numbers of correct answers per minute. The useful question is not “Can I solve this set?” but “Where is my next minute most likely to produce correct answers?”

Solving ability and decision quality both matter. A candidate who cannot build representations quickly will not be rescued by clever triage. A candidate who solves well but keeps investing in the wrong set leaves marks unclaimed.

Where marks are actually lost

These are primarily execution errors. They can coexist with solving-skill gaps, but better decision habits can reduce them even when your underlying DILR ability stays the same.

Why a fixed plan fails

The structure of the section has not been constant. These are structural facts from past papers:

• CAT 2021: 20 DILR questions in four sets of 4, 4, 6 and 6.
• CAT 2022 and CAT 2023: 20 questions each, in four sets of five.
• CAT 2024 and CAT 2025: 22 questions each, in five sets: two of five questions and three of four.

The section ran for 40 minutes in each of these years. In the latest completed CAT papers, it moved from four sets to five. Nothing guarantees that a future paper will follow any of these patterns, so a plan such as “always attempt N sets” rests on an assumption the exam does not promise to honour.

Scan First, but Know What a Scan Can and Cannot Tell You

Scanning is an uncertainty-reduction exercise, not clairvoyance. Its job is to give you enough information to compare the available sets and decide where to begin. It cannot tell you with certainty which set is easiest.

What you can assess while scanning	What you know only after entering a set
Amount and format of information	Whether your chosen representation actually works
Whether a natural table, grid or diagram suggests itself	Whether a first useful inference emerges
Number of entities being tracked	Whether constraints collapse possibilities or casework expands
Presence of fixed anchors	Whether questions open independently or stay locked behind the full solution
Apparent calculation load	How much calculation the questions actually demand
Apparent branching	Whether solving is step-by-step or all-or-nothing
Whether questions look dependent or potentially local	Whether a hidden interpretation trap exists
Your familiarity with this structure	Whether familiarity translates into speed on this instance

The left column supports your first ranking. The right column is why that ranking must stay provisional. A clean-looking set can stall at the first inference, and a heavy-looking set can settle once one anchor is found.

How long should you scan?

Keep the initial scan brief, and treat any exact duration as a heuristic. Coaching and test-taking approaches differ on how long to scan and when to run time checkpoints, and no single figure suits every candidate or paper. What matters is the output: a rough ranking of sets, not a partial solution of any of them.

Whether to look at every set before starting is also a judgement. Looking at all of them improves your comparison but costs time. Looking at fewer saves time but raises the chance of missing a better opportunity. Your mock history is the best guide to which side suits you.

Formats are visible at scan time. CAT 2023 Slot 2, for example, included a set built on two 3×3 tables using mean, median and mode. You would register that format in seconds, but how much effort such a set needs can only be judged from the actual set on the day.

Triage Your Sets: Attack, Hold, Defer

After scanning, sort what you see into three provisional categories. No scoring formula is needed. Attack sets have a clear representation, a visible entry point and manageable execution risk. Hold sets are potentially solvable but carry real uncertainty. Defer sets look unclear, heavily branched, expensive to calculate or lacking an entry point. Defer does not mean discard. It means “not first,” and a deferred set can become the best remaining option later.

Dimension	Attack	Hold	Defer
What you observe	You can see how you would draw the table or diagram; an anchor is visible; load looks bounded	Setup is possible but something is unresolved: an ambiguous statement or an unsizeable workload	You cannot picture a representation; conditions branch widely; nothing gives an entry point
What it suggests	Low setup risk, likely early return	Possible return with uncertainty attached	High setup risk relative to other options
What to do	Start here; pick the best Attack if several exist	Keep as fallback; start here only if no Attack exists, choosing the least uncertain	Leave for later; revisit if better options run out
What could change it	Setup fails to yield a usable representation: downgrade	A key statement clarifies or an anchor appears: upgrade. Ambiguity persists: downgrade	A second reading reveals a clean entry point: upgrade

Classify by structure, not topic label. “This is an arrangement set” tells you little. “I can see a fixed anchor and few entities” tells you more.

Personal difficulty vs paper difficulty

A set being regarded as manageable in general does not automatically make it the best set for you. “Easy in the paper” is not “easy for you.” Common profiles include:

• Strong at arrangements but weaker in calculation-heavy data;
• Comfortable with charts but uncomfortable with casework;
• Highly accurate but slow at setup;
• Fast, but prone to missing a constraint.

Each of these should rank the same set differently. Your mock history is evidence about which families you solve cleanly and which cost you time or marks. Use it to inform triage without turning preferences into absolute rules. Familiarity helps you recognise structure quickly, but it does not prove this instance will cooperate.

Historical fact: Tournament and games formats have appeared across several CAT papers (CAT 2021 Slot 2 “Games & Tournament”, CAT 2021 Slot 3 “Javelin Tournament”, CAT 2022 Slot 3 “Four-Person Games”, CAT 2024 Slot 1 “Tournament”). Arrangement formats appeared in different CAT 2025 slots (Slot 1 “Circular Arrangement — Four Persons / Seven Chairs” and Slot 3 “Circular Arrangement — Seven Students”).

Analytical inference: Because families recur across papers and slots, the family label alone is insufficient for selecting a set. This is our reasoning, not a finding that any of those named sets was easier or harder than another.

Illustrative scenario (not a claim about any named set): Suppose games and tournament sets are consistently your weakest family in mocks. Even if one looks tidy at scan time, your own record is a reason to rank a comparable set from another family above it, unless the tournament set shows an unusually clear entry point.

Choosing the first set

Prefer an Attack set. If several qualify, break ties using personal fit, apparent calculation risk and question count. If none qualifies, begin with the least uncertain Hold. Be cautious with two shortcuts: “shortest set first” confuses length with ease, and “DI first” or “LR first” assumes a difficulty hierarchy that is not established.

The Progress Checkpoint: Continue or Switch?

This is the core decision of the section. Once inside a set, the useful question is not “How long have I been here?” but “Is my uncertainty shrinking?”

Elapsed time prompts reassessment; it does not decide anything by itself. Some coaching approaches suggest specific time checkpoints, and the numbers differ from source to source, so treat any such figure as a heuristic to test in mocks, not a law.

A hard set is not a bad set. What matters is whether the difficulty is yielding.

Indicator	Difficult but progressing	Difficult and stagnant
Representation	Stable; holds up under new statements	Unusable, or redrawn repeatedly
Information gained	Definite values or placements accumulating	Rereading produces nothing new
Possibilities	Alternatives reducing	Cases multiplying, not collapsing
Reasoning	You are making deductions	Assumptions are replacing deductions
Questions	One or more becoming answerable	Calculation expands without unlocking any question
Remaining uncertainty	Bounded; you can say roughly what is left	Interpretation itself is still uncertain
Likely decision	Continue	Consider reallocating if a better shortlisted set exists

Signals will not always agree, so weigh them together. One stubborn inference does not make a set stagnant if everything else is stable. Several stagnation signals appearing together are a strong reason to reassess.

Illustrative scenarios

Network-format sets have appeared in CAT (CAT 2022 Slot 2 “Warehouse and Four Destinations — Network”; CAT 2024 Slot 3 “Network of Roads and ATMs”). That is a historical fact about the family recurring. The scenarios below are hypothetical and make no claim about how those sets behaved.

• Illustrative continue case: Suppose a network set gives you a stable representation quickly and values begin to accumulate with each condition. Even if the set is demanding, the evidence says you are progressing. Continue.
• Illustrative switch case: Suppose that after rereading you still cannot build a usable representation, and you notice you are guessing at relationships the statements do not support. The evidence says you are stagnant. If another shortlisted set looks more promising, reallocate.

The tie-breaker

When signals are mixed, compare the expected return from your next few minutes here with the expected return from your best remaining alternative. If staying looks at least as good, stay. If the alternative is clearly better, move. It is a judgement, not a calculation, but asking the question explicitly keeps you honest.

Sunk Cost, Leaving a Set and Recovering from a Wrong First Choice

The sunk-cost trap

The familiar thought is: “I have already spent so long on this set that I cannot leave now.”

Time already spent cannot be recovered, whether you stay or go. So separate two things:

Past investment is not a reason to continue, and it is not a reason to leave either. If the representation is stable and completion appears near, continuing can be entirely rational after a long stay. If effort keeps producing no new deductions while better opportunities remain, reallocation is rational however much time you have already sunk.

Park, don't abandon

When you leave a set, do it deliberately. Jot down what you established, whether any question looked answerable from it, and why you left. Return only for a reason: a fresh idea about the representation, or nothing better remaining. Do not return out of a sense of unfinished business.

After a wrong first pick

A poor first choice is a decision error to correct, not a section failure. Recognise it through the progress signals, park the set with notes, re-triage the remaining sets (your view of them may have changed), and take the best remaining Attack or Hold.

Structurally, a single slot can put several formats in front of you at once. CAT 2024 Slot 1 contained five sets across different families: Stock Prices; Web Surfers / Stars; Three Persons Visiting Different Countries; Tournament; and Two Candidates Contesting Election. That makes re-triage a genuine comparison.

Illustrative scenario (not a claim about any named set): Suppose your first pick stalls after setup. You park it, look again at the remaining four and find one whose representation is now obvious. That is the process working. Recovery is not guaranteed, but a calm re-triage usually beats digging in.

Whole Set or Individual Questions?

Set level. Early in the section, solving a whole set is often efficient because one representation can unlock several questions. Setup cost is paid once and the return is spread across the set.

Question level. If full resolution becomes expensive, check whether particular questions can be answered through local conditions, bounds, direct data, specific cases or partial deductions. This is an opportunity to check, not a guarantee. Some sets contain such questions and some do not.

Dimension	Whole-set thinking	Question-level thinking
When it fits	Earlier in the section; representation stable; deductions accumulating	A set resists full solution; time is shrinking; a parked set may still have easy marks
What you do	Complete the representation and work through the questions	Look for questions needing only a local condition, bound or directly readable value
Verify first	That your representation is holding	That the question does not silently depend on the unresolved part
Main risk	Sinking time into a set that will not close	Answering from an assumption a later condition would overturn
Main benefit	One setup cost across several answers	Marks from a set you would otherwise leave entirely

Set sizes matter too. CAT 2021 had sets of 4, 4, 6 and 6 questions, which is a structural fact: payoff differed by question count within the same section. Analytical inference from question count, not a historical difficulty finding: a larger set may offer greater payoff but also greater concentration risk if the representation fails. That is a reason to weigh set size against your confidence in the setup, and nothing more.

How Remaining Time Changes Your Decisions

As the section progresses, starting a fresh set costs a full setup that weighs more heavily as time shrinks. The goal shifts from “finish what I started” to “capture the best marks still available.” That might be the partial questions of a parked set, a Hold set with a bounded route, a single question needing only local reasoning, or a TITA whose answer is already within reach. Do not rush to meet a predetermined attempt target; hurrying through a fresh set to keep pace with a plan tends to produce errors.

How many DILR sets should you attempt?

• The actual difficulty of the paper and the quality of the sets in your slot;
• Your mock history and strengths by set family;
• Your setup speed, inference speed and accuracy;
• Your progress inside the section and the time remaining;
• The partial-question opportunities available.

The variation across CAT 2021 to CAT 2025 shows why a single number cannot be right, and none of those structures is promised for the future. No number of sets guarantees any percentile. The number you attempt is an outcome of your decisions and solving, not a goal to hit. What counts is correct answers, net of negative marking.

If you are slower, focus on faster commit-or-park decisions and on checking for question-level marks, so a stalled set does not absorb the section. If you are strong, refine triage, review easier sets you rejected, and guard against overreach. No reliable evidence establishes “attempt everything” as a universal strategy.

MCQ, TITA and the Calculator

In recent CAT papers, an incorrect MCQ attracts negative marking, while TITA (type-in-the-answer) questions do not carry the same wrong-answer penalty. Check the official notification for your year, since marking details are set by the exam conductors. That difference is often summarised as “TITA is free,” which is wrong.

• If you have already reached an answer, entering it costs essentially nothing.
• If a TITA needs meaningful further work and your confidence is low, the absence of a penalty does not make the time worthwhile. Compare it against your other options.
• “Always attempt every TITA” ignores the cost side.

For MCQs, the same logic applies with added risk: an unsupported guess can cost marks under the marking scheme in force.

Calculator. Use the on-screen calculator where arithmetic is genuinely cumbersome, such as percentages, ratios or verification. The principle is: solve the structure first; calculate only what the structure requires. CAT 2025 Slot 2 included a set labelled “Pollution Index — Weighted Average”, which shows only that weighted-average data sets have appeared. The calculator guidance does not depend on it.

Build Your Personal DILR Decision Profile from Mocks

You cannot rehearse instincts you have never measured. Knowing how to attempt DILR in CAT in theory is one thing; seeing your own decision errors in data is what makes it usable. Keep the log small. Five fields you actually review beat a giant spreadsheet you never open.

After each DILR section in a mock, record:

Optional extras are setup time, solve time after setup and time to first correct answer.

After a handful of mocks, look for patterns rather than single cases; the number needed varies, so treat any figure as a heuristic. Ask which families repeatedly make you regret a choice, whether you reject sets too quickly or persist too long, and where your progress typically stalls. Use the profile to adjust triage. “I usually do well on arrangements” is a reason to lean towards them, not to ignore a clearer set of another type. This is a decision log, not a full mock-analysis method.

Common DILR Attempt-Strategy Myths

Where the evidence does not establish something, the table says so, and it avoids replacing one unsupported claim with its opposite. Several of these myths shape how to attempt DILR in CAT for many candidates without ever being tested.

#	Myth	Better decision
1	“Always solve two full sets.”	No fixed count fits every paper or candidate. Let the paper, your mock record and your progress set your attempt level.
2	“Spend the first five minutes only scanning.”	Scan time is a heuristic that varies by practitioner. Scan briefly for a provisional ranking and adjust from mocks.
3	“DI is easier than LR.”	No reliable hierarchy exists in either direction. Difficulty depends on the set and on you.
4	“Choose the shortest set first.”	Shorter is not easier. Judge by representation clarity, entry point and execution risk.
5	“Choose the most familiar topic first.”	Familiarity aids recognition but does not prove this instance is solvable. Treat it as one factor.
6	“Never leave once you have invested several minutes.”	Past time is sunk. Decide on current progress: stay if completion is near, leave if stagnant.
7	“TITA questions should always be attempted.”	Entering an answer you already have costs little. Spending real time on a low-confidence TITA has an opportunity cost.
8	“Strong students should target all sets.”	No reliable evidence supports this as universal. Attempt what you can solve accurately given the paper.
9	“The first displayed set is rarely the easiest.”	No reliable evidence establishes this or its opposite. Evaluate each set on its merits.
10	“You must fully solve a set before answering.”	Some questions can be answered from local conditions, bounds or direct data. Check, but do not assume every set allows it.
11	“Set selection matters more than solving ability.”	Both matter. Good selection cannot replace solving ability, and strong solving can be wasted by poor allocation.
12	“If the first set goes wrong, the section is ruined.”	A wrong first pick is a decision error. Park it, re-triage and take the best remaining opportunity. Recovery is not guaranteed, but panic makes it less likely.

How to Attempt DILR in CAT: The Complete Exam-Hall Decision Flow

Here is how to attempt DILR in CAT as a single flow. It is a decision aid to rehearse in mocks, not an algorithm that guarantees an outcome.

The important loop is the return to triage. Every time you park a set, re-rank what remains, because your picture of the other sets has changed.

If this happens → do this

Situation	What to do
Your first set stalls right after setup	Run the progress check. If stagnation is clear and another set looks better, park with notes, re-triage and move.
Two sets look equally good	Break the tie by personal fit, then question count, then apparent calculation risk. Keep the other as first fallback.
You keep rereading the same statements	Treat it as a stagnation signal. Try one different representation; if that also fails, reassess against alternatives.
Cases are multiplying, not collapsing	Look for a missed anchor or constraint that prunes them. If none and another set is available, consider reallocating.
You are close to finishing a set you have spent long on	Past time is sunk. If the representation is stable and remaining work is bounded, continuing may be right.
A question looks answerable from one local condition	Check that it does not depend on the unresolved part, then answer it.
A TITA needs a long derivation and you are unsure	Weigh the time against other options. No negative marking does not make the time free.
Time is short and a parked set had partial progress	Check its questions for ones answerable from what you established, and compare against starting something new.
You realise you rejected a set too quickly	Re-triage. If it is now the best remaining option, go back.
You feel tempted to speed up to reach a set count	Return to the decision variable: which action produces the most correct answers from here?

Final DILR Exam-Day Checklist

Use this checklist to rehearse how to attempt DILR in CAT before test day.

Scan and triage

• Scan each set briefly and note only observable features.
• Sort into Attack / Hold / Defer by structure and personal fit, not topic label alone.
• Pick your first set from Attack, or the least uncertain Hold.

Inside a set

• When the clock prompts you, ask whether uncertainty is reducing and anything is becoming answerable.
• Distinguish “difficult but progressing” from “difficult and stagnant.”
• Ask whether past time or current evidence is driving your decision.

Leaving and returning

• Note what you established before parking a set.
• Re-triage before starting anything new.
• Return to a parked set only for a reason.

Questions and marking

• Check for questions answerable from local conditions, bounds or direct data.
• Enter answers you have already reached, including TITA.
• Do not spend meaningful time on a low-confidence TITA just because it has no negative marking.
• Calculate only what the structure requires.

Last stretch

• Spend remaining time on the best available marks, not on finishing whatever you started.
• Do not rush to hit a predetermined set count.

Minimum Viable DILR Mock Log

• Easier set missed: which set could you have solved productively, and why did you skip it?
• Wrong first-set choice: did your first pick stall, and what did the scan miss?
• Time lost in abandoned sets: how much time went into sets you left without marks?
• Accuracy by set type: which families gave clean answers, and which cost wrong ones?
• Time to first useful progress: how long did each set take to yield a usable representation?

Optional: setup time, solve time after setup, time to first correct answer.

Frequently Asked Questions

01How to attempt DILR in CAT: what is the core method?

The core method for how to attempt DILR in CAT is to scan the sets, sort them into Attack, Hold and Defer using structure and your own strengths, and commit to the strongest option. Then judge by progress, not elapsed time. Continue while uncertainty is reducing, reallocate when effort produces no meaningful progress and a better opportunity exists, and use remaining time on the best marks still available.

02How much time should I spend scanning DILR sets?

There is no universal scan duration. Coaching approaches differ, and any specific number is a heuristic. Keep the scan brief and aim for a provisional ranking, not the start of a solution. Use mocks to see whether your scans are too short (you miss good sets) or too long (you lose solving time).

03Should I scan all sets before solving?

It depends on you and the slot. Scanning all sets improves comparison but costs time; scanning fewer saves time but raises the chance of missing a better set. There is no universal rule, so test both in mocks.

04How do I choose the easiest DILR set?

You cannot identify it with certainty, since scanning reduces uncertainty without removing it. Look for a clear representation, a visible entry point and manageable-looking load, combine that with your personal fit from mocks, and let the first few minutes of progress confirm or correct the choice.

05How do I choose my first DILR set?

Prefer a set you would classify as Attack. If several qualify, break ties using personal strengths, apparent calculation risk and question count. If none qualifies, start with the least uncertain Hold.

06How many DILR sets should I attempt?

There is no universal number. It depends on the paper's difficulty, set quality, your mock history, strengths by family, speed, accuracy, progress and time remaining. The section's structure has also varied across recent CAT papers, so a fixed target is not a safe assumption.

07Does the number of sets I attempt determine my percentile?

No number of attempted sets guarantees any percentile. What counts is correct answers, net of negative marking on MCQs. Attempting more sets can raise or lower your score depending on accuracy and time spent.

08When should I leave a DILR set?

Consider leaving when progress has stalled and a better opportunity exists. Signs include rereading that yields nothing new, an unusable representation, cases multiplying instead of collapsing, assumptions replacing deductions, and calculation growing without unlocking questions. Elapsed time prompts the check, but evidence of progress decides.

09How do I avoid sunk-cost traps in DILR?

Separate past investment from current evidence. Time already spent cannot be recovered, so compare what the next few minutes are likely to produce with your best alternative. Continuing can be right if the representation is stable and completion is near; leaving can be right if you have stopped making deductions.

10What if my first DILR set goes wrong?

Treat it as a decision error, not the end of the section. Park the set with notes, re-triage the remaining sets and take the best remaining Attack or Hold. Recovery is not guaranteed, but calm re-ranking usually serves you better than staying with a stalled set.

11Should I attempt DI or LR first?

Neither has a reliable claim to be easier, so there is no fixed order. Choose by representation clarity, entry point, execution risk and your record in mocks.

12Is a familiar DILR set necessarily easier?

No. Familiarity helps you recognise structure quickly but does not prove this particular set will cooperate. Treat it as one factor alongside how the set looks and how similar sets have gone in your mocks.

13Is the shortest set the best one to start with?

Not necessarily. Fewer questions mean a smaller payoff, and a short set can still be hard to set up. Judge by representation clarity and execution risk, using set size as one input.

14Can I solve individual questions without completing the whole set?

Sometimes. Questions may be answerable from local conditions, bounds, direct data, specific cases or partial deductions. Not every set has such questions, so check whether a question truly depends on the unresolved part before answering.

15Should I return to an abandoned set?

Return when you have a reason: alternatives are worse than expected, time has opened up, or you have a fresh idea about the representation. Notes taken when you parked the set make a return faster.

16Are TITA questions safer?

They carry no wrong-answer penalty in recent CAT formats, which removes one kind of risk, but they still consume time. Zero negative marking does not mean zero opportunity cost, so a TITA needing long work for a low-confidence answer is not automatically a good use of time.

17Should I attempt every TITA?

No. If you already have an answer, entering it has little downside. But do not spend meaningful time deriving a low-confidence TITA only because it has no negative marking.

18How should slower and stronger DILR candidates approach the section?

Slower candidates should build discipline around commit-or-park decisions and question-level marks, so one stalled set does not absorb the section. Stronger candidates should refine triage, review easier sets they rejected in mocks, and guard against overreach. Neither group has a fixed target.

19How can I identify my best DILR set types, and what should I track in mocks?

Keep a compact log of easier sets missed, wrong first-set choices, time lost in abandoned sets, accuracy by set type and time to first useful progress. After a handful of mocks, patterns in your set-family accuracy and rejection errors become visible, and they should shape your triage.

20How should I use the calculator in DILR?

Solve the structure first and calculate only what the structure requires. The calculator helps with cumbersome arithmetic such as percentages and ratios, and with verification. Avoid computing every figure in a table without knowing what the questions need.

21Does the CAT DILR pattern change every year?

The structure has changed across recent papers. CAT 2021 had four sets of 4+4+6+6 questions; CAT 2022 and CAT 2023 had four sets of five; CAT 2024 and CAT 2025 had five sets, two of five questions and three of four. Plan around a process that adapts to the paper rather than a fixed structure.