DILR set selection in CAT: how to pick the sets worth solving

In most CAT sections, a hard question costs you a couple of minutes. In DILR, a hard set can cost you fifteen minutes and return nothing, because you need to crack the set before any of its questions opens up. That is why strong DILR scorers are not always the fastest solvers. They are the best choosers.

This guide gives you a scan routine, the signals that separate a friendly set from a trap, a worked ranking, and a rule for when to walk away.

Why fewer sets done fully beat more sets done partly

Compare two hypothetical ways an aspirant could spend the 36 or so minutes left after a four-minute scan, on four sets of five or six questions.

The numbers are illustrative, but the shape is real. Most of the time in a set goes into cracking it. Once cracked, its remaining questions come cheaply. Nine or ten minutes on a set is often just enough to understand it and not enough to cash in.

The scan: four to five minutes, all sets

Open each set and spend about a minute on it. Do not solve anything. Look at:

1. The type. Is it a type you have practised often?
2. The size. How many entities and how many attributes? Five people on five days is small. Eight teams with points, goals and match results is large.
3. The conditions. Are they direct ("R is on Thursday") or indirect ("the number of people between R and S is equal to…")?
4. The questions. Do they ask for direct values, or for "maximum possible", "minimum possible", "which of the following must be true"?
5. The numbers. For DI, are the values friendly (multiples of 5 and 10) or awkward (four-digit values with decimals)?

Signals of easier and harder sets

Looks hard, often isn't

• Long introduction, plain conditions. A set with ten lines of story but direct conditions is often friendlier than a four-line set of abstract constraints.
• Large table, most of it given. If the gaps can be filled from totals, the size is not the difficulty.

Looks easy, often isn't

• A short set with "at least" and "at most" everywhere. These often lead to maxima-minima reasoning with several cases.
• A familiar type with an unusual twist, such as a round-robin tournament where the points system is non-standard.

A scoring rubric

Rate each set from 0 to 2 on four questions.

Take the highest scorer first. A score of 3 or less is a set to open last, if at all.

Worked example: ranking four sets

Suppose the four sets on your screen look like this.

Order: A, D, C, then B.

• A is a classic fill-the-gaps table. See table-based DI.
• D is solvable but slow. Take it second, and approximate wherever the options are far apart.
• C is middling. Knockouts have a clear structure, but "who could have met whom" questions need cases.
• B has no fixed starting point, and it asks for counts of possible arrangements. It is the most likely time sink.

Open B only if you finish the others with at least ten minutes left, and then look for any question you can answer without solving the whole set.

When to walk away

• The five-minute check. Five minutes into a set, you should have placed at least one fact and ruled out at least one case. If you have neither, move on.
• The contradiction loop. If you have restarted the grid twice because cases keep breaking, the set is either harder than it looked or you misread a condition. Re-read the conditions once. If nothing changes, leave.
• The sunk-cost trap. Time already spent is gone either way. Ask only: "Is my next ten minutes worth more here or on an untouched set?"

Partial solving

A set you cannot crack fully can still give marks.

• Questions with their own "if". Some questions add a condition that fixes the arrangement, which makes that one question solvable even when the set is not.
• Questions answerable from one condition. A question about the total of a column might need only the given totals.
• TITA answers. With no negative marking, enter your best value on any TITA question you have worked on.

Practice drill

Rank these three sets using the rubric, then check the answers.

1. Seven people stand in a line. Four of their positions are given directly, and the remaining conditions are about who stands next to whom. Questions ask for positions.
2. A table of monthly production of four plants over six months, with half the cells missing. Only some totals are given, and questions ask for the "maximum possible" production of one plant.
3. A round-robin league of five teams (each team plays every other once), with a win worth 2 points and a draw 1. Final points are given for all teams, and questions ask for the result of specific matches.

Answers:

1. First. Four fixed positions, direct conditions and definite questions: the rubric scores it at or near 8.
2. Last. Missing totals and "maximum possible" questions signal a maxima-minima set with several cases.
3. Second. A clear structure (10 matches in all, since 5 × 4 ÷ 2 = 10) and definite questions, though some deduction is needed.

What to do next

• In your next five DILR sectionals, spend the first four to five minutes scanning and write your rubric scores on the rough sheet.
• After each test, add a column to your set log: "Was my ranking right?"
• Practise the five-minute check on every set this week, even in untimed practice.
• Read DILR time management for how to fit the scan into the 40 minutes.