How Many DILR Sets Should You Actually Attempt Now?
This question circulates with confident answers attached. Attempt three. Attempt two and finish them. Attempt all four if you are aiming high. Every one of those is a guess dressed as a target, and the reason is the same in each case.
A set count cannot produce a percentile. Percentile is your position relative to everyone who sat the paper, and it moves with paper difficulty and with normalisation across slots, so no number of anything converts into it. Underneath that, the right count depends on how many sets on that particular paper are solvable by you in forty minutes, which is not knowable in advance and differs between candidates and between papers.
The count is an output of your judgment, not an input to it. Build the judgment across CAT DILR practice chapters.
- No set count produces a percentile, since percentile depends on the whole cohort and on normalisation.
- The right number depends on the paper and on you, and both are unknown until the section starts.
- Sets finished cleanly matter, not sets touched. Partial work on a set yields very little.
- Walking in with a target count converts selection into a quota and buys the worst sets on the paper.
- Decide the count after the opening scan, from what is actually in front of you.
Why a Count Cannot Produce a Percentile
Two separate breaks in the chain make the claim impossible, and both are worth seeing clearly.
The first is that sets attempted do not equal sets solved. A set touched and abandoned at minute nine contributes almost nothing, so a count of attempts describes activity rather than marks. Two candidates attempting three sets each can score entirely differently.
The second is that marks do not equal percentile. Your position depends on how everyone else performed on the same paper, and normalisation adjusts across slots that did not receive identical papers. Even a perfect prediction of your marks would not fix your percentile.
So "attempt N sets for 99 percentile" makes two unwarranted leaps at once. Published as advice it reads as precision, and it is arithmetic with the middle removed.
Walking into the section with a target number of sets in your head. It converts selection from a judgment into a quota, so you start committing to sets in order to reach the number rather than because they looked solvable. The quota reliably buys the worst sets on the paper, since those are what remains when you are chasing a count.
What Actually Determines the Right Number
Four variables decide it, and none of them is available before the section begins.
- How many sets the paper has. Not published in advance, and the section's composition has moved between years.
- How many are tractable for you. A set you recognise is a different proposition from the same set met by someone who has not.
- Question density per set. A set with five questions justifies more minutes than one with three.
- Your setup speed on those types. A candidate with standard representations finishes what another candidate cannot.
The Paper Half of the Equation
In the 2022 paper DILR carried 20 questions worth 60 marks with a 40 minute sectional limit, spread across a handful of sets. Recent papers have run roughly 66 to 68 questions overall, and the internal composition has moved with the paper's size. There is no fixed structure to plan a count against.
The Candidate Half
Two aspirants can look at the same four sets and correctly reach different decisions. One has done forty arrangement sets and can finish the arrangement set in seven minutes; the other has not and would need fifteen. The right count for them is genuinely different, and no article can know which of them you are.
Why Density Matters More Than Count
Two sets carrying five questions each are worth more than three carrying three each, and cost less total setup. Judging the section by sets rather than by questions attached is a small error that compounds over the forty minutes.
| What people ask | What decides it | When it becomes knowable |
|---|---|---|
| How many sets to attempt | How many you can finish | After the opening scan |
| Is three enough | Question density of those three | After the opening scan |
| Should I try a fourth | Minutes remaining and its tractability | Mid section |
| What percentile will that give | The whole cohort and normalisation | Never, in advance |
The Rule That Replaces the Number
A rule survives an unseen paper in a way a target cannot, and this one is short.
Read every set in the opening two to three minutes without solving. Judge each on whether you recognise its structure, whether it hands you firm constraints early, and how many questions hang off it. Then decide how many you will attempt properly, and divide the remaining time by that number.
Commit fully to the ones you choose, because questions within a set share a structure and half attention produces a half structure and almost no marks. Hold a walk-away time and leave anything that passes it, even when it feels close.
The count that results is whatever the paper allowed. On a paper with three tractable sets it will be three; on a harsher one it may be two, and two finished cleanly can be a good section.
Write your intended count and your walk-away time on the rough sheet after the scan and before you start solving. A number you wrote two minutes ago, chosen while looking at the actual paper, is far harder to argue with at minute nine than an intention formed last week.
Why Fewer Sets Often Scores More
This runs against instinct and it follows directly from how the section is built.
Questions arrive attached to sets, so the marks come from resolved structures rather than from time spent. A candidate who finishes three sets collects three sets of returns. One who touches five and completes none collects a handful of partial answers and has spent the same forty minutes.
Attempting more also degrades each attempt. Forty minutes across five sets gives eight minutes each, which for most sets is not enough to build the structure properly, so the extra coverage buys nothing and costs the depth that would have paid.
This is why the correct response to a paper with more sets is to become more selective rather than faster. More sets means more choice, which is an advantage if you use it to pick better and a burden if you read it as an obligation.
Mentors comparing good and poor DILR sections rarely find a difference in sets attempted. They find a difference in sets finished, and behind that a difference in how the first three minutes were spent. The candidate who chose after reading everything finished what they started; the one chasing a count did not.
Finished Sets, Not Attempted Sets
The count is an output. Read everything, choose deliberately, commit fully, leave early.
Work Through CAT DILR ChaptersHow to Find Your Own Working Number
If you want a personal figure, derive it from your own sections rather than borrowing one.
Take your last three timed DILR blocks. For each, record sets attempted, sets finished, and productive minutes, meaning the share of the forty that went to sets producing marks. The gap between attempted and finished is the number that matters.
If that gap is large, your count is too high for your current setup speed, and the fix is selectivity rather than pace. If the gap is small and you finished early, you were too conservative and can take one more.
Track it across a term and you will develop a reliable sense of how many sets you finish in forty minutes, which is the closest thing to a personal number that exists. It will still bend to the paper, which is why it informs the scan rather than replacing it.
The Drill That Improves the Judgment
Five sets, ninety seconds each without solving, ranked by expected solvability, then all five solved and the ranking checked. Ten repetitions and the ninety second judgment stops being a guess, which is what makes the in-section count decision reliable. Drawing the five sets from a mix of types, including selection based sets, keeps the drill informative.
What to Track Instead of the Total
Productive minutes rises when selection improves, often before the score does, because DILR totals carry a lot of paper variance. Our guide on spotting a solvable set in two minutes covers the signals the drill teaches you to read.
Why the Number Keeps Getting Quoted
It is worth understanding why this particular claim is so persistent, because the same shape shows up across CAT advice.
A set count is easy to remember, easy to act on and impossible to falsify. A candidate who follows it and scores well credits the target; one who follows it and scores badly assumes they executed poorly. Nothing about the outcome ever counts as evidence against the number, so it survives indefinitely regardless of whether it helped anyone.
It also comes with a plausible origin story. Somebody who scored well did attempt three sets, and reporting that is honest. The error enters when a description of what one person did on one paper becomes a prescription for what everyone should do on a paper nobody has seen.
The test for any strategy claim is whether it would still make sense on an unseen paper. Choose sets you can finish passes that test. Attempt three does not, because it was derived from papers that have already been written.
The Honest Answer
What is the ideal number of sets to attempt in CAT DILR? There isn't one, and anyone giving you a number is either guessing or repeating a guess.
Attempts do not convert to marks without resolution, and marks do not convert to percentile without the rest of the cohort and normalisation. Both links are missing from the claim, which is why it sounds precise and instructs badly.
What exists instead is a procedure that works on every paper. Read everything before committing, choose on structure, constraints and question density, commit fully to what you choose, and leave anything that passes your walk-away time. The count falls out of that, and it is the right count for that paper because it was derived from it.
- Do you decide your set count after the scan rather than before the section?
- How large is the gap between sets attempted and sets finished?
- Do you note question density per set during the scan?
- Is your walk-away time written down rather than intended?
If your DILR sections end with several sets half done, the count you are carrying is too high for your current setup speed. A CAT preparation strategy review will confirm it, and a personalised CAT preparation plan can hold the ranking drill in your week.
Derive the Count, Do Not Carry One In
Two sets finished cleanly beats four started. The paper decides which number that is.
Build My Weekly PlanFrequently Asked Questions About DILR Set Counts
How many DILR sets should I attempt for a 99 percentile?
No count produces a percentile. Attempts convert to marks only when a set is actually resolved, and marks convert to percentile only through the whole cohort's performance and normalisation across slots. Both links are missing from any such target.
Is attempting three sets enough?
It depends on the paper and on you. What matters is sets finished rather than attempted, and how many questions hang off each. Decide the count after reading every set in the opening minutes, not before the section starts.
Should I attempt more sets if the paper looks easier?
Possibly, but decide it from the scan rather than from a general impression. More sets on the paper means more choice available, and the candidates hurt by a larger set count are the ones who read the extra set as an obligation.
Why does attempting fewer sets often score more?
Because marks come from resolved structures rather than time spent. Spreading forty minutes across five sets leaves too little per set to build the structure, so the extra coverage yields little while costing the depth that would have paid.
Practice DILR sets, chapter by chapter
CAT DILR practice across LR puzzles and DI sets, each with a worked solution.
More from DILR
Continue reading
Using a Study Journal for CAT Without Overdoing It
How to Stop Comparing Your CAT Prep With Everyone Else
SPJIMR Admission Process: Why the Profile Leads It
How to Solve CAT Quant Faster Without Skipping Steps
Put it into practice