Can You Hit 99 Percentile in DILR With Only 2 Sets?
The question circulates in a specific form: someone scored very well having attempted only two sets, therefore two sets is enough. It is an appealing story because it suggests the section can be won with restraint rather than with capability, and restraint is easier to adopt than skill.
Two things are wrong with it. No set count produces a percentile, so the arithmetic in the claim does not connect. And the candidate in the story did not succeed because they attempted two sets; they attempted two because that was what the paper and their own speed permitted, which is a completely different causal story.
The count is an output of 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.
- Sets attempted is the wrong number. Sets finished, and the questions attached to them, is what scores.
- Two well chosen sets can be a good section, and two is not a target to walk in carrying.
- Question density matters: two sets carrying five questions each is a different proposition from two carrying three.
- The story gets the causation backwards, since the count was a consequence rather than a strategy.
Why a Count Cannot Produce a Percentile
Two separate breaks in the chain, and both are worth seeing clearly.
The first is that attempts do not equal marks. A set touched and abandoned contributes almost nothing, while a set finished cleanly yields the questions attached to it. So a count of attempts describes activity rather than outcome, and two candidates attempting two sets each can score entirely differently.
The second is that marks do not equal percentile. Your percentile is your position relative to everyone who sat the exam, it moves with how the whole cohort performed, and normalisation adjusts across slots that did not receive identical papers. There is no fixed mapping between a raw score and a percentile.
So a claim of the form "N sets produces X percentile" makes two unwarranted leaps at once. It sounds precise and it is arithmetic with the middle removed.
Adopting a low attempt target because a high scorer reported one. A target count converts selection from a judgment into a quota, and a quota of two makes you stop after two even when a third tractable set is sitting there with eleven minutes left.
What the Story Actually Describes
The anecdote is usually true and the inference drawn from it is not.
- The paper offered limited tractable sets. On a harder paper, two may be all that resolves in forty minutes for anyone.
- They read everything first. The count followed from a deliberate choice rather than preceding it.
- They finished what they started. Two completed sets yield their whole question clusters, which is where the marks came from.
- The cohort found it hard too. Percentile is relative, so a modest raw score sits high when scores fall across the room.
The Causation Runs the Other Way
They did not choose two and then score well. They chose well and the count came out as two. Reversing that, and walking in intending to attempt two, produces a candidate who stops choosing after two decisions rather than one who chooses correctly throughout.
Question Density Is Doing Hidden Work
Two sets carrying five questions each is ten questions. Two carrying three each is six. Those are very different sections and the anecdote never specifies which. Judging the section by sets rather than by questions attached is a small error that compounds across forty minutes.
And the Cohort Matters
On a paper where most candidates struggled, a section built on two resolved sets sits well in the distribution. On a gentler paper the same section sits lower. Neither outcome is a property of the number two.
| What people hear | What actually happened |
|---|---|
| Two sets is the strategy | Two was the result of choosing well |
| Attempting less scores more | Finishing what you start scores more |
| The count produced the percentile | The cohort and normalisation produced it |
| Two sets is enough | Enough depends on density and on the room |
What Is Genuinely True in It
The story is not nonsense, and separating the sound part from the unsound one is where the value is.
It is true that fewer sets finished cleanly frequently outscores more sets touched. Questions arrive attached to sets, so marks come from resolved structures rather than from time spent, and spreading forty minutes across five sets leaves too little per set to build any of them properly.
It is true that a correct answer adds three marks while a wrong multiple choice answer deducts one and a blank costs nothing, which means half-built sets producing guesses are worse than untouched ones.
And it is true that restraint is undervalued in a preparation culture that treats attempts as the measure of effort. The correction is real; the number attached to it is the invention.
Decide your intended count after the opening scan, not before the section, and write it on the rough sheet along with your walk-away time. A number chosen while looking at the actual paper is a judgment. A number carried in from an anecdote is a quota.
How to Decide the Number Properly
The procedure replaces the target and it works on a paper nobody has seen.
Read every set in the opening two to three minutes without solving. Note how many there are, how many questions hang off each, and whether you recognise the structure of each one.
Then decide how many you can realistically finish in the remaining time, given your own measured setup speed. That figure is personal: a candidate who lays out a distribution set in thirty seconds can take more than one who needs two minutes, and both should know which they are from practice.
Then divide the remaining time by that number, commit fully to the ones you chose, and hold a walk-away point. The count that results is whatever the paper allowed, and on some papers it genuinely will be two.
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 completed what they started; the one carrying a target did not.
Finished, Not Attempted
Two finished sets can be a good section. Two as a target is a quota that stops you choosing.
Work Through CAT DILR ChaptersFinding Your Own Working Figure
If you want a personal number, derive it from your own sections rather than borrowing one.
Take your last three timed DILR blocks and record, for each, sets attempted, sets finished, and productive minutes, meaning the share of the forty that produced marks. The gap between attempted and finished is the number that tells you something.
A large gap means your working count is too high for your current setup speed, and the fix is selectivity rather than pace. A small gap with time left over means you were too conservative and can take one more.
Across a term that figure stabilises into a realistic sense of how many sets you finish in forty minutes. It still bends to the paper, which is why it informs the scan rather than replacing it.
The Drill Behind 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 rather than arbitrary.
What to Track Instead of the Total
Productive minutes rises when selection improves, often before the score does, because DILR totals carry heavy paper variance. Our guide on spotting a solvable set in two minutes covers the signals that make the opening judgment work.
The Risk in a Low Count Nobody Mentions
Most discussion of this treats a low attempt count as the safe option, and there is a real downside worth stating.
If you commit to two sets and one of them fails to resolve, you have spent a large share of the section on a single successful set, with very little time left to recover. A candidate attempting three has more tolerance for one going wrong, because the failure is a smaller fraction of the whole.
So the restraint argument has a limit. Fewer sets finished properly beats more sets touched, and it also concentrates your outcome into fewer decisions, which is the same mechanism that makes DILR high variance in the first place. Deliberately narrowing to two increases that concentration rather than reducing it.
The balanced position is to choose as many sets as you can genuinely finish and no more, and to keep a walk-away time so that a failure is contained early enough to leave room for a replacement. That is not a count strategy at all, which is rather the point.
The Answer
Can two sets be enough in CAT DILR? On some papers, yes, and the sentence is incomplete without the paper. What is not true is that two sets produces any particular percentile, because attempts do not convert to marks without resolution and marks do not convert to percentile without the rest of the cohort and normalisation.
The candidate in the anecdote attempted two because reading everything told them two were worth taking, and then they finished both. That is the transferable part: read before committing, choose on structure and question density, finish what you start, and leave what will not resolve.
Do that and the count takes care of itself. It might be two, it might be three, and on a gentle paper with high density it might be more. Carrying a number in the door is the one version guaranteed to be wrong on the specific paper in front of you.
- 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?
- Do you know your own setup speed for the set types you handle?
If your sections end with several sets half done, the count you are carrying is too high for your current speed. A CAT preparation strategy review will confirm it, and a personalised CAT preparation plan can hold the ranking drill in your week.
The Paper Decides the Count
Your job is to read it properly first. The number is what falls out of that.
Build My Weekly PlanFrequently Asked Questions About DILR Set Counts
Can attempting only two DILR sets produce a 99 percentile?
No count produces a percentile. Attempts convert to marks only when a set actually resolves, and marks convert to percentile only through the whole cohort's performance and normalisation across slots. Both links are missing from that claim.
Is two sets enough?
On some papers yes, depending on how many questions attach to each and how the cohort performed. Decide the count after reading every set in the opening minutes rather than carrying a target in from someone else's result.
Why do some high scorers attempt so few sets?
Because they read everything first and committed only to sets they could finish. The low count was the consequence of choosing well, not the strategy that produced the score, and reversing that causation produces a quota rather than a judgment.
What number should I track instead?
Sets finished rather than attempted, the gap between the two, and productive minutes, meaning the share of your forty minutes that produced marks. All three are informative in a way that a target count never is.
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
Why Word Problems Feel Harder in CAT Quant Than School
Why High Scorers Skip an Entire CAT RC Passage Fully
Why DILR Sets Take Longer to Understand Than Solve
Do Mocks in a Noisy Environment Help or Hurt CAT Prep?
Put it into practice