Mock tests and analysis for SBI PO

A mock test is a rehearsal. Its value comes from what you change after it. Two aspirants can take the same thirty mocks; one improves steadily and the other stays flat. The difference is almost never the number of mocks. It is what happens in the hour after each one.

This guide covers when to start, how many to take, how to include the descriptive test, and a method of analysis you can follow after every mock, with a worked example.

Why mocks matter more in SBI PO than you might think

The SBI PO pattern has several features that only timed practice teaches:

• Sectional timers in both the prelims and the mains, so you must learn the right pace for each section separately.
• A 40-question English section in the prelims at 30 seconds a question, a pace most people have never worked at.
• Four sectional cut-offs plus a descriptive cut-off in the mains, so a single weak section can end your attempt.
• A 30-minute typed descriptive test straight after three hours of objective questions, which tests stamina as much as writing.

You can know every concept and still fail on any of these. Mocks are where you find out.

Types of mock and when to start

Memory-based previous papers are reconstructions by coaching institutes, not official papers, so treat them as a guide to style rather than exact questions.

How many?

There is no correct number, but a workable pattern is:

If analysis takes longer than the mock itself, that is fine. Take fewer mocks rather than skipping analysis. A mock without analysis gives you a score and nothing else.

Include the descriptive test

Many candidates take the objective mains mock and skip the descriptive part. On exam day the descriptive test comes immediately after three tiring hours of objective questions. Practise it in the same condition: tired, timed and typed.

If your mock series does not evaluate descriptive answers, score your own with a simple rubric:

Apply it to each of the three tasks and add the three scores. The descriptive test overview explains each format.

A structured analysis method

After every mock, work through these five steps, in writing:

1. Scores and attempts by section, entered in a log.
2. Every wrong answer: why? Classify it as a concept gap, a calculation slip, misreading the question, or a guess.
3. Every skipped question: attempt it untimed. Were there easy ones you never reached or passed over?
4. Time sinks: which questions took more than two minutes? Were they worth it?
5. Three actions for the next mock, specific enough to check.

Worked example

Suppose an aspirant's full prelims mock goes like this:

Net scores are correct answers minus a quarter mark for each wrong answer.

The analysis finds:

• Quant, 5 wrong: three were calculation slips in one DI set (units in lakhs misread as thousands), two were guesses on series.
• Quant skipped: two approximation questions were never reached because the DI set took nine minutes.
• English skipped: four RC questions were direct fact questions that took under 30 seconds each when attempted untimed.
• Reasoning: a clean section; one puzzle was left, correctly, after two minutes without progress.

Three actions for the next mock:

1. In quant, do approximation before DI, and write the units at the top of the rough sheet before starting any DI set.
2. No guesses in number series unless two options are eliminated.
3. In RC, read the questions first and answer fact questions before inference questions.

If these actions had been followed, the two approximation questions and four RC fact questions (assuming all six correct) would have added 6 marks, and removing the two series guesses would have saved 0.5. That is a potential 66 without learning a single new concept.

An error log

Read the log every Sunday. A reason that keeps appearing tells you what to study next, far more reliably than your overall score.

Percentiles, not raw scores

Mock series differ in difficulty, and so do individual mocks within a series. Your percentile (the share of test-takers who scored below you) is a better guide than marks.

Worked example. An aspirant ranks 1,200th out of 20,000 test-takers. The share below them is (20,000 − 1,200) ÷ 20,000 = 18,800 ÷ 20,000 = 94%, so they are at about the 94th percentile.

A steady or rising percentile over five or six mocks matters more than any single high or low score. Percentiles only mean something when many people take the same mock, so prefer a series with a large number of test-takers.

Choosing a mock series

• One good series is enough; a second can add variety close to the exam.
• Look for sectional timers, an interface close to the real exam, full solutions and descriptive practice, with feedback if possible.
• A large number of test-takers makes percentiles meaningful.
• Prefer a series that is honest about difficulty over one that makes you feel good.

Practice set

1. A candidate gets 32 right and 6 wrong in prelims English. What is the net score? 32 − 1.5 = 30.5.
2. In mains DI, 14 right and 4 wrong. Net score? 28 − 2 = 26 (2 marks each; 0.5 penalty each).
3. Rank 450 of 9,000 test-takers. Approximate percentile? (9,000 − 450) ÷ 9,000 = 95%.
4. Your last five mock scores are 58, 64, 55, 61, 63, but your percentile has risen each time. Are you improving? Yes. The raw scores vary with mock difficulty; the rising percentile shows you are doing better against the same field.
5. You have time for either two mocks without analysis or one mock with full analysis. Which? One mock with full analysis.

What to do next

• Set up an error log today, with the five columns above.
• Take your next mock at the time of day your exam slot is likely to be.
• Read the prelims strategy and mains strategy to decide your section order before the next mock.