CAT DILR: How to Choose Which Sets to Attempt First

Ask any high scorer what mattered most in their DILR section, and "set selection" comes up almost every time, more often than raw logical ability, more often than calculation speed. DILR is unusual among CAT's three sections in that the questions within a set are tightly coupled. Get the setup right and four or five questions become straightforward. Misread one condition and the entire set can become unsolvable in the time you have. This makes choosing which sets to attempt, and in what order, a skill in its own right.

The first three minutes. Scan, don't solve

Before committing to any set, spend roughly sixty to ninety seconds per set scanning the full DILR section, not solving, just reading. You're looking for three things. How many variables are involved, whether the data is presented cleanly or requires significant inference to set up, and whether the question types attached to the set look standard, ranking, grouping, scheduling, or unusual.

Sets with more than four or five interacting variables, or that require you to infer missing data from indirect clues before you can even build your table, are higher risk. They're not necessarily unsolvable, but they take disproportionately long to set up, and a slow setup eats into time you need for the rest of the section.

A simple triage framework

Green, attempt first. Clean data, standard structure, you can mentally sketch the table within seconds of reading. These are your highest expected value sets, fast to set up, multiple questions become easy once the table is built.

Yellow, attempt second, time boxed. Workable but will take longer to set up, maybe six to eight variables, or a data structure that needs careful reading. Attempt these once you've banked the green sets, and give yourself a hard time cap, say twelve minutes, before deciding whether to push through or move on.

Red, attempt last, if at all. Sets where, even after your initial scan, you're not confident you understand what's being asked. These often have a low number of "gettable" questions even if you do partially solve them. If time allows after green and yellow sets, attempt the most approachable looking questions within a red set rather than the full set. Partial credit on a hard set is often a better use of remaining time than a slow, full attempt.

The sunk cost trap

The single most expensive DILR mistake isn't picking a hard set. It's refusing to abandon one once you're six minutes in and realize it's harder than expected. The emotional pull to finish what you started is strong, but DILR punishes this instinct severely. If you're past your time cap on a set and haven't cracked the core structure, the correct move is almost always to bank whatever partial answers you can infer and move to the next set, not to keep pushing on the assumption that the breakthrough is one minute away.

How to actually practice set selection

This is a skill that's hard to build from solving sets in isolation, which is exactly why full section DILR practice, attempting an entire section's worth of sets together, under time pressure, rather than one set at a time, matters more for this skill than topic isolated practice. When you only ever solve one set per session, you never practice the triage decision of which set to attempt next. You've already been told which one to do.

A useful drill. Take a full DILR section from a past paper, give yourself five minutes to scan and rank all the sets green, yellow, red before solving anything, then attempt them in that order with strict time caps. Afterward, compare your actual results against your initial ranking. Did the sets you called "green" really turn out to be the fastest? This calibration exercise, repeated across several papers, builds the pattern recognition that experienced solvers rely on almost instinctively.

DILR ability obviously matters. But two aspirants with similar logical reasoning skill can post very different DILR scores purely based on whether one of them wasted nine minutes on a set the other correctly identified as low value and skipped.