DILR guide
Two and a half sets
What that means in questions
- CAT 2025 DILR had 22 questions in 5 sets: two sets of 5 questions and three of 4. Half of them, 11, were TITA.
- Marking: +3 for a correct answer, −1 for a wrong MCQ, 0 for a wrong TITA. Half the section carries no penalty, so attempt every TITA you have worked on.
- Two full sets give you 8–10 questions; two more from a third set make about 11 attempts. At 85% accuracy that is 9 correct: 27 marks, or 25 if both misses were MCQs.
What the target is worth
| Percentile | CAT 2024 | CAT 2025 (initial estimates, by slot) |
|---|---|---|
| 99th | 37.8 | 30–34 |
| 95th | 27 | 23–26 |
| 90th | 22.5 | 17–19 |
| 85th | 19.5 | 13–15 |
| 80th | 17 | 11–13 |
Sources: Cracku (CAT 2024) and IMS (CAT 2025 initial estimates, so treat them as approximate).
Why 2.5 sets is realistic
IMS rated two of the five sets in every CAT 2025 slot as easy and one as medium; the other two were difficult. Your 2.5 sets are exactly those: the two easy ones and the first questions of the medium one. The whole skill is spotting them in the first five minutes.
The easy sets were not all puzzles. They included import–export data for five countries, trains moving between five stations, authors and books shown in two bar graphs, a ranking set, six friends calling each other and a chart of friends solving puzzles. Six of the fifteen sets that year were data-interpretation style (tables, bar graphs, a scatter plot, trade and currency data), so DI practice is not optional.
Exam day: choose first, solve second
Spend the first 5 minutes choosing, not solving. Picking the right two sets is worth more than any technique, and most low DILR scores come from sinking 20 minutes into one bad set.
Minute 0 to 5
1. Scan all five sets
Give each about a minute: read the opening lines of the setup and glance at its questions. Score it on the signals and rank the five.
Minute 5 to 19
2. Your best set
Draw the table, place what is certain, answer every question.
Minute 19 to 33
3. Your second set
Same again.
Minute 33 to 40
4. The half set
Open your third-ranked set and answer its two easiest questions: usually the ones that read straight off the data, or a TITA. In the last 30 seconds, enter your best reasoned value in any TITA you worked on.
Pick first
- A familiar structure (seating, a grid, a schedule, a table) with 4–6 entities.
- Totals that must add up: every total is a free equation.
- DI with complete, clean data and straightforward calculation questions.
- Several TITA questions, since they carry no penalty.
- Questions that read off the finished table (“which of these is true?”).
Leave for later
- Long lists of “if… then…” rules, or games with moves.
- Every question asking for a “maximum possible” or “which can be true” across many cases.
- Unusual setups: 3-D cubes, abstract processes, a long text for only 4 questions.
- Data you cannot organise into one table within 2 minutes.
The abandon rule
If you are 7 minutes into a set with no grid filled and no question answered, stop and move to your next-ranked set. Come back only if time is left at the end.
Order of questions inside a set
- Answer the questions that use only the main grid first.
- Questions that add a condition (“If X sits next to Y, then…”) are mini-cases; do them last.
- “Must be true”: the option has to hold in every surviving case. “Can be true”: one valid case is enough.
- MCQ guessing: only after you have ruled out two options. TITA: always enter your best answer.
Worked example: ranking five sets in five minutes
These are the kinds of opening lines you see when you open a DILR section. Rank the five sets before opening the answer.
- Set A
- Seven employees were rated by three managers on a 1–10 scale. A table shows some of the ratings, and each employee’s final rating is the median of the three. 4 questions, 2 of them TITA.
- Set B
- A bar chart shows the exports and imports of five countries in 2024, and a second chart shows each country’s trade balance in 2025. 5 questions, 3 of them TITA.
- Set C
- Eight players sit in a circle with tokens, and a long list of rules decides who passes tokens to whom in each round. 4 questions, every one asking “which of the following can be true”.
- Set D
- Six people sit around a table with eight chairs, so two are empty. Clues say who sits opposite or next to whom and how many empty chairs lie between pairs. 4 questions.
- Set E
- A large cube built from 27 small cubes is painted in three colours on different faces, and then some small cubes are removed. 5 questions.
Show ranking and whyHide ranking and why
- B first. Complete numbers, five questions and three TITAs; the work is the percentage and ratio arithmetic you already practise in QA.
- A second. One table, seven people and a simple rule (the median of three). Most questions read straight off the completed table.
- D as the half set. Seating is familiar, but empty chairs multiply the cases, so take only its one or two most direct questions.
- C skipped. Long rules plus “can be true” questions across many cases is the classic time sink.
- E skipped. It needs three-dimensional visualisation, which is slow to do reliably under time pressure.
Takeaway. The ranking takes about a minute per set: read the setup’s first lines, glance at the question types and count the TITAs. Practise it with the selection drill.