GATE Mock Test Analysis: How to Review a Mock and Plan the Next Week

You finish a GATE mock, see a score, and wonder what to do with it. The answer is that the score is the least useful number on the page. A proper review turns one afternoon's paper into a plan for the next seven days. This guide teaches that review step by step, for every GATE branch, with one fully worked example.

Key takeaways

• Score the mock with GATE's own marking scheme before you read a single solution.
• Give every wrong, skipped and slow question one row in an error log, with one cause from a fixed list of five.
• Four numbers summarise a mock: accuracy, attempt rate, marks lost to negatives and marks per minute.
• Split lost marks into "didn't know" and "knew but lost". The second group is the fastest to recover.
• Each cause has its own fix, and the next seven days should go to the cause that cost the most marks.
• Compare mocks by their rates and causes, not by the raw score, because mock difficulty changes.

The words this review uses

Every GATE paper has 65 questions for 100 marks in 3 hours, and each question is worth 1 or 2 marks. Those two facts fix the split: 30 one-mark and 35 two-mark questions, since $30 + 35 = 65$ and $30 + 2 \times 35 = 100$. General Aptitude is 15 of the 100 marks in every paper. Confirm the current pattern at gate2027.iitm.ac.in.

Five terms carry the rest of this guide.

• Net marks: marks for correct answers minus the negative marks for wrong MCQs.
• Accuracy: the share of attempted questions you got right.
• Attempt rate: the share of the 65 questions you attempted at all.
• Marks lost: for one wrong question, the marks it carried plus any penalty it charged.
• Marks per minute: net marks earned in a stretch of the paper, divided by the minutes it took.

Step 1: score it exactly as GATE would

Score the paper with the answer key first, and leave the solutions closed. Reading solutions first makes every wrong answer look like a slip, and the log turns dishonest.

The scheme is the same for every GATE paper, and it is set out in the GATE exam pattern and marking scheme guide, which applies to all branches. Check the current figures on the official question paper pattern page before you rely on them.

Trap: An MSQ with one extra option ticked is simply wrong. Log it as wrong, not as "nearly right".

Step 2: build the error log

The error log is a table with one row for every question that did not earn full marks, plus any correct one that took far too long. Use the same columns every time, so that logs from different mocks can be added together.

Each row gets exactly one cause. If two seem to fit, pick the one that happened first.

Step 3: compute the four numbers

Accuracy is correct answers over attempted questions, and attempt rate is attempted questions over all 65:

Here $C$ is the number of correct answers and $A$ the number attempted. Net marks are the positive marks $P$ minus the penalty, where $W_1$ and $W_2$ count wrong 1-mark and 2-mark MCQs:

A worked example

The figures below are invented to show the arithmetic. They are a hypothetical mock, not anyone's real result.

Positive marks come from each group's correct answers times its marks. The penalty applies only to the nine wrong MCQs:

Now the rates. Accuracy is $\displaystyle \frac{34}{48} \approx 70.8\%$ and the attempt rate is $\displaystyle \frac{48}{65} \approx 73.8\%$.

The 14 wrong answers cost more than the penalty, because each also gave up the marks it carried:

Finally, marks per minute. Suppose your timestamps show the first 120 minutes earned 38 net marks and the last 60 minutes earned the remaining $\displaystyle \frac{22}{3} \approx 7.33$. That is $\displaystyle \frac{38}{120} \approx 0.32$ marks a minute against $\displaystyle \frac{7.33}{60} \approx 0.12$. That last hour also carried $\displaystyle \frac{11}{3}$ of the $\displaystyle \frac{14}{3}$ penalty marks. The late, low-yield stretch is where the guessing happened.

Remember: Score first, read solutions second. A review built on solutions you have already read hides your real causes.

The virtual calculator guide is worth reading if many of your slips happen on the on-screen calculator. Every one of the six GATE Guide books carries ten full-length mocks with worked solutions, so you can run this review on papers built to the official pattern. See what each book includes.

Step 4: separate "didn't know" from "knew but lost"

Group the log by cause. In the example the 14 wrong answers split like this, with marks lost including penalties:

The first two rows are "didn't know": $\displaystyle \frac{26}{3} + \frac{16}{3} = 14$ marks. The last three are "knew but lost": $\displaystyle \frac{25}{3} + \frac{7}{3} + 2 = \frac{38}{3} \approx 12.67$ marks. Without studying a single new topic, this candidate's ceiling was:

That gap of nearly 13 marks is the most useful number in the whole review. It is recoverable with habits, not new chapters.

In one line: "Didn't know" needs study; "knew but lost" needs practice and habits, and it is usually cheaper to fix.

Step 5: a seven-day plan for each cause

Give most of the next week to the cause with the largest marks lost. The table shows what each cause asks of you.

Guessing deserves one line of arithmetic. A blind guess on a 1-mark MCQ has an expected value of $\displaystyle \frac{1}{4}(1) - \frac{3}{4}\left(\frac{1}{3}\right) = 0$. Rule out one option and it becomes $\displaystyle \frac{1}{3}(1) - \frac{2}{3}\left(\frac{1}{3}\right) = \frac{1}{9}$. So guess an MCQ only after an elimination, and always enter an answer for MSQ and NAT.

If General Aptitude rows keep appearing, those 15 marks are often the quickest to secure; the General Aptitude preparation guide covers that section for every branch.

Comparing mocks without chasing the raw score

Mocks differ in difficulty, so two raw scores are not directly comparable. Track the rates and the knew-but-lost total instead. Here are three hypothetical mocks, the first being the worked example:

Mock 3's score fell, yet accuracy rose, negative marks fell and knew-but-lost marks more than halved since mock 1. That is a candidate improving on a tougher paper. The real exam's marks are later turned into a GATE score, as the score and normalisation guide explains, so steady habits matter more than one mock's number.

Trap: Repeating a mock you have already reviewed inflates the score. Use repeats only for timed practice, never for the trend table.

Every two or three mocks, add the logs together and sort by section and cause. A cause that tops the list twice in a row is where your next full week belongs.

Quick revision

1. Score with the official scheme first: $\displaystyle -\frac{1}{3}$ and $\displaystyle -\frac{2}{3}$ for wrong MCQs, nothing for MSQ and NAT.
2. Every wrong, skipped or slow question gets one row and exactly one cause.
3. The five causes are concept gap, guess, calculation slip, misread and time.
4. Accuracy is $\displaystyle \frac{C}{A}$, attempt rate is $\displaystyle \frac{A}{65}$, and marks lost include both the marks forgone and the penalty.
5. Net marks plus knew-but-lost marks gives your ceiling without new study.
6. Spend the next seven days on the cause that cost the most marks.
7. A blind MCQ guess is worth zero on average; guess only after ruling out an option.
8. Compare mocks by accuracy, negatives and knew-but-lost marks, not by raw score alone.