Why most mock analysis fails
A mock produces one number and a great deal of evidence. Most candidates keep the number and throw away the evidence. The score tells you where you stand today; the evidence tells you what to change, and only one of those affects your next mock.
There is a second, subtler failure. Analysis that produces a long list of weaknesses is barely more useful than no analysis, because you cannot act on eleven things at once and you will not be able to tell which change helped. Good analysis is aggressively reductive: it ends with two or three decisions.
The seven-step method
Do this in order, in one sitting, within a day of taking the mock while you still remember your reasoning on individual questions.
- Record the numbers before you look at solutions. For each section, write down questions attempted, questions correct, and time spent. Compute accuracy per section. Doing this before reading any solution stops hindsight from rewriting what happened.
- Separate the score into attempts and accuracy. A low score has two very different causes: attempting too few questions, or attempting too many at poor accuracy. The fix for each is opposite, so identify which one applies section by section.
- Classify every wrong and skipped question. Label each mistake as a concept gap, a calculation slip, a misread question, a selection error, or a time-pressure error. Counting the categories tells you what to actually work on.
- Find the questions you should have skipped. Look for questions where you spent more than three minutes and still got it wrong. These are selection errors and they cost time plus marks, making them the highest-value thing to fix.
- Find the questions you should have attempted. Re-solve the paper untimed. Anything you now solve comfortably but skipped in the exam was a selection error in the other direction, usually caused by panic or poor scanning.
- Write down two or three decisions. Convert the analysis into specific changes for the next two weeks, such as a topic to revise, a per-question time cap, or an attempt-count target for one section. More than three changes at once means you cannot tell which one worked.
- Log the mock and compare the trend. Record the section-wise result so you can see accuracy and attempt trends across mocks. One mock is noise; a trend across five is information.
Reading attempts and accuracy together
A section result only makes sense when you look at both numbers at once. Four patterns cover almost everything you will see.
Low attempts, high accuracy
You are leaving marks on the table. Push the attempt count up gradually and watch whether accuracy holds. This is the most comfortable problem to have.
High attempts, low accuracy
Negative marking is eating the score. The fix is selection discipline, not more practice volume—work on deciding faster what to leave alone.
Low attempts, low accuracy
A genuine preparation gap in this section. Return to fundamentals and topic-wise practice before expecting mock results to move.
High attempts, high accuracy
This section is working. Protect the time it gets and shift marginal effort to whichever section is weakest.
The arithmetic behind the second pattern is worth stating plainly. With +3 for a correct answer and -1 for an incorrect MCQ, each wrong answer swings you four marks away from a correct one. Twenty attempts at 60% accuracy nets 28 marks; fourteen better-chosen attempts at 80% nets 30. See the full marking scheme for how this interacts with TITA questions, which carry no penalty.
Classifying mistakes so the categories mean something
Go through every question you got wrong and every question you skipped, and give each one exactly one label. Forcing a single label is what makes the counts interpretable.
- Concept gap — you did not know the method. Fix by studying the topic, then revisiting it in your revision queue.
- Calculation slip — right method, wrong arithmetic. Fix with deliberate accuracy drills, not new theory.
- Misread — you solved a question that was not asked. Fix by re-reading the question stem before committing.
- Selection error — you should have skipped it, or should have attempted it. Usually the highest-value category.
- Time pressure — you knew it but ran out of room. Often a downstream symptom of selection errors earlier in the section.
After three or four mocks the distribution becomes obvious, and it is usually not what candidates expect. Concept gaps are rarely the largest category by the middle of preparation; selection errors usually are.
Section-specific things to check
VARC
Check accuracy on inference questions against direct questions separately. Note whether you lost marks on the passage you found least interesting—that is a stamina and engagement problem, not a comprehension one.
DILR
The decision that matters is which sets you chose. Record how long you spent scanning before committing, and whether the set you abandoned was genuinely harder than the one you finished.
QA
Separate topic gaps from speed gaps. If you solve a question correctly untimed in two minutes but skipped it in the exam, the problem is confidence and scanning, not the topic.
Across sections
Look for spillover. A rough VARC often shows up as a poor DILR set choice, because the sections are attempted back to back with no reset.
Turning analysis into the next two weeks
Close every analysis by writing two or three concrete changes, each specific enough that you will know in a fortnight whether it worked. Not “improve DILR” but “spend the first three minutes of DILR scanning all sets before starting any”. Not “revise Geometry” but “two Geometry sessions this week, then a mixed set”.
Then make those changes visible during the week rather than filed away in a notebook. Put the topics into your weekly timetable, let the revision queue bring them back before you forget them, and check the accuracy trend after the next two mocks — a single mock cannot tell you whether a change helped, but three in a row can.
Frequently asked questions
How long should CAT mock analysis take?
Plan for roughly the length of the mock itself, so about 90 to 120 minutes. If analysis regularly takes far longer, you are probably re-solving the entire paper rather than targeting what went wrong. If it takes 15 minutes, you are only reading solutions.
How many mocks should I take for CAT?
Quality of analysis matters more than volume. A common pattern is one mock a fortnight early in preparation, moving to one or two a week in the final two months. Never schedule a mock you do not have time to analyse—an unanalysed mock mainly teaches you your current score.
Why is my CAT mock score not improving?
The most common reasons are analysing scores instead of decisions, changing too many things at once so no change can be attributed, and taking mocks faster than you can absorb them. Check whether your accuracy is improving even when the score is flat, since rising accuracy at a stable attempt count usually precedes a score jump.
Should I attempt fewer questions to score more in CAT?
Often, yes. With +3 for a correct answer and −1 for a wrong MCQ, the gap between right and wrong is four marks. Attempting 14 questions at 80% accuracy can outscore attempting 20 at 60%. The right attempt count is whatever maximises net marks for you, and it is found through mock data rather than by rule.
What should I track after each CAT mock?
At minimum, record attempts, correct answers, and time for each of VARC, DILR, and QA. That gives you score, accuracy, and section balance. Adding a short note on the main error type makes the trend across mocks far easier to read later.