IBPS PO Mains Data Analysis & Interpretation 2026: 40 Questions, 60 Marks, 45 Minutes
IBPS PO Mains Data Analysis & Interpretation 2026: 40 Questions, 60 Marks, 45 Minutes
The 2026 Mains paper on 4 October gives you 45 minutes for Data Analysis & Interpretation. Last cycle it also gave you 45 minutes, for 35 questions worth 50 marks. This year the same clock has to cover 40 questions worth 60 marks.
That single line is the whole article. Five more questions, ten more marks, not one extra second. Everything below is what that does to your attempt count, your accuracy floor, and your final rank.
The DI box at a glance
Objective total this year is 170 questions, 200 marks, 160 minutes, plus a 25-mark descriptive paper in 30 minutes. If you have not read the full pattern change yet, start with our IBPS PO Mains 2026 pattern breakdown and come back here.
What actually changed in DI, section by section
Most pages describing the 2026 Mains still print last year's table. Here is the honest comparison, with the 2025 column taken from a dated July 2025 page and the 2026 column cross-verified on two September 2026 pages.
Read the DI row again. GA got ten extra minutes to carry its extra load. DI got nothing.
You are being paid 5% more per question and given 12.5% less time to earn it. A DI set you solved comfortably in a 2025 mock will now finish outside the bell if you carry the same pace into October.
DI and Reasoning are now the same box
This is the structural fact of the 2026 paper and almost nobody has said it out loud. Reasoning and DI are identical: 40 questions, 60 marks, 45 minutes each. Same marks per question, same seconds per question, same seconds per mark. The only difference left between them is which one you can actually finish.
One DI question is worth 1.50 marks, which is 3.00 times one English objective question and 1.25 times one GA question. GA still pays the best rate per minute at 1.714, and we covered why in the Mains General Awareness revision map. But GA is a memory section with a ceiling. DI is 30% of the marks and it is the only block where a well-drilled candidate can pull clear of a crowd.
The negative marking number almost every page gets wrong
You will read "0.25 marks deducted for a wrong answer" on a dozen coaching pages. The rule is one-fourth of the marks assigned to that question, not a flat quarter mark.
A DI question carries 1.50 marks. One-fourth of 1.50 is 0.375. So a wrong DI answer costs you 0.375, which is 50% more than the 0.25 those pages imply.
The useful consequence is the guess break-even. You need p such that p x 1.50 equals (1 - p) x 0.375. That solves to p = 20.00%. Because the penalty is proportional to the marks, this same 20.00% holds in every section of the paper.
Note what a five-option paper does to the blind guess: it lands exactly on break-even, so random marking gains you nothing and costs you the seconds you spent doing it. In DI the elimination usually comes from a rough magnitude check, not from the options themselves.
Attempts you need, by accuracy
Net score in DI is A x (1.875p - 0.375), where A is attempts and p is accuracy. Turn that around and the attempts you need for a target are A = target / (1.875p - 0.375). This table is the one to pin above your desk.
"Impossible" is literal. The section has 40 questions. An entry above 40 means the arithmetic asks you to attempt more questions than exist.
Look at the 60% row. To reach 30 out of 60 marks at 60% accuracy you need exactly 40.00 attempts, which is every question in the section, at 67.5 seconds each, with zero margin for a single set that refuses to open. At 55% accuracy, 30 marks in DI is off the table completely.
That is the real message of the section. A DI accuracy problem is not something you can outrun with courage in the exam hall. Only practice closes it, and you have 27 days.
Why 25 careful questions beat 40 rushed ones
Here is the table that changed how we teach this section. It shows what skipping buys you: fewer attempts means more seconds on each attempt, which raises accuracy, which raises the net.
Now read the two cells that matter. 25 attempts at 85% accuracy gives you 30.47 marks. 40 attempts at 55% accuracy gives you 26.25. The candidate who answered 15 fewer questions finished 16.07% ahead.
And she did it with 108 seconds per question instead of 67.5, which is 60% more thinking time, which is exactly how her accuracy got to 85% in the first place. The two columns are not independent. Selection is what buys accuracy.
My honest opinion after watching this section for three cycles: the 25-to-30 attempt band at high accuracy is the right target for most candidates in October, and chasing 38 attempts is how people score 26 and blame their luck.
What DI is worth in your final rank
Final merit is Mains 80% plus Interview 20%, with the 225-mark Mains scaled to 100. That makes every Mains mark worth 0.3556 final points.
DI alone carries 21.33 final points. The entire interview carries 20.00. Your DI section outweighs your interview by 6.67%, and unlike the interview it is a skill you can measurably improve in four weeks.
Translated to the unit that stings: one Mains mark equals 1.78 interview marks. So one DI question is worth 2.67 interview marks, and one five-question DI set is worth 13.33 interview marks. Walking away from a set you could have opened costs you more than a bad answer on your strengths and weaknesses question.
If the difference between a qualifying mark and a merit mark is still fuzzy, our qualifying marks versus merit marks explainer works through it across six exams.
Count sets, not questions: the 5:37 rule
DI does not arrive as 40 loose questions. It arrives as sets. A caselet, a table, a bar graph, each carrying four or five questions off one data block. So your clock is not really 67.5 seconds per question.
At five questions per set, 40 questions is eight sets in 45 minutes. That is 337.5 seconds per set, or 5 minutes 37 seconds. Budget it that way.
The discipline that matters: give each new set 45 seconds of reading before you commit. If the data block has a missing value chain, a percentage-of-percentage layer, or three variables you have to hold at once, leave it and open the next one. That 45-second scan across eight sets costs you six minutes and it is the cheapest insurance on the paper.
Second discipline: once you are 90 seconds into a set with nothing solved, walk. Sunk time is not a reason to stay. This is the single habit that separates a 30 from a 20.
We are not printing a 2026 DI weightage table, and here is why
You have probably searched for "IBPS PO Mains DI topic wise weightage 2026" and found tables. We looked at them. We are not publishing one, and the reason is arithmetic, not caution.
Every historical DI weightage table is a count of questions in a 35-question section. The 2026 section has 40. Reprinting "Caselet: 18 questions" against a different denominator is not a forecast, it is a rounding error dressed as advice. Where the extra five questions land is a decision IBPS has not published.
What we can honestly give you is the history, labelled as history. This is a dated July 2026 compilation of actual Mains papers, and every column below describes the old 35-question section.
Two things are stable across six years and you can plan on them. DI and caselet never fell below 10 questions and touched 25. Arithmetic never disappeared.
Everything else moved around. So the sane preparation order is caselet and multi-graph DI first, arithmetic second, and the small formats third.
One more warning about sources. During research for this article we found a page updated 19 June 2026 still printing a 157-question Mains with an Essay and Letter descriptive. Letter writing was replaced by comprehension this cycle, which we documented in the Mains descriptive paper breakdown.
Another page dated 18 June 2026 printed section times that add to 180 minutes under a total that says 160. Check the arithmetic on any pattern table before you build a plan on it, including ours.
A 27-day DI plan that respects the clock
27 days, and DI is not the only section you owe time to. This plan assumes roughly 90 minutes of DI a day and that your basics (percentages, ratios, averages) are done. If they are not, spend the first four days there and compress the rest.
Log three columns after every timed section: solved clean, solved slow, guessed. Column two is your real syllabus. A question you solved in four minutes is a question you will not solve on 4 October, and it belongs in tomorrow's revision list, not in your confidence.
Five mistakes that cost DI marks in October
- Carrying 2025 pace into a 2026 paper. Your old comfortable set speed is now 12.5% too slow. Reset the timer, not the confidence.
- Treating 0.25 as the penalty. It is 0.375. Twelve wrong answers cost you 4.5 marks, which is three clean questions.
- Attempting a ninth set because eight felt too few. The table above says 25 good attempts beat 40 rushed ones by 16.07%.
- Skipping the 45-second scan. Committing to a set blind is how candidates lose seven minutes on a caselet they were never going to crack.
- Practising DI untimed for all 27 days. Untimed practice builds method. Only timed practice builds a score. Switch by 13 September.
Where these numbers came from
Verify the pattern against the official IBPS notification PDF and your Mains call letter before you finalise any plan. If a figure here ever disagrees with those, the official document wins.
Practising this section without an IBPS PO series
Honest note. ExamAtlas does not have an IBPS PO Mains series live yet. We have 14 test series and the closest banking one is the IBPS Clerk Prelims mock test series, which does carry Numerical Ability with the same arithmetic and DI base. What it does not carry is the 40-question, 60-mark, 45-minute Mains block, so treat it as calculation practice, not as a section simulation.
For the section itself, run this drill with a plain timer and any question bank. Eight sets, 45 minutes, phone in another room. Mark each set as opened-and-solved, opened-and-abandoned, or skipped on the scan. After ten repetitions you will know your real attempt band, which is the number this whole article exists to help you find.
If you are choosing between banking and a state exam this season, our IBPS PO versus IBPS RRB Scale-I comparison covers the pay and posting side of that decision.
FAQs
How many questions should I attempt in IBPS PO Mains DI 2026?
For most candidates, 25 to 30 questions with 80% or better accuracy. That band delivers 26.67 to 36.56 marks and leaves you 90 to 108 seconds per question. Attempt 35 or more only if your mock accuracy is already sitting above 85%.
Is the negative marking in DI 0.25 or 0.375?
0.375 marks. The rule deducts one-fourth of the marks assigned to the question, and a DI question carries 1.50 marks. The flat "0.25" you see on many pages is the rule applied to a one-mark question, which DI is not.
Should I guess in the DI section?
A blind guess on five options is exactly break-even at 20.00% and gains you nothing. Guess only after eliminating at least two options, where the expected value turns positive at +0.25 marks per guess, and only in the last two minutes when the time has no better use.
Is DI harder in 2026 than last year?
The questions are not necessarily harder. The clock is. The section went from 35 questions to 40 in the same 45 minutes, so you have 12.5% less time per question and 16.67% less time per mark than the 2025 paper allowed.
Which DI topic should I practise first with 27 days left?
Caselet and multi-graph DI. Across six years of Mains papers, DI and caselet never dropped below 10 questions in a 35-question section and once reached 25. Arithmetic-based questions come second. Data sufficiency and quantity comparison appear irregularly and are not worth your first two weeks.
Does DI have a sectional cut-off in Mains?
Do not plan on the answer either way from a coaching page. Sectional requirements are stated in the official notification and the call letter, and they have changed between cycles. What is certain is that DI carries 60 of the 200 objective marks and 21.33 of your 100 final points, so it is not a section you can sacrifice regardless of how the cut-off is structured.
Final word
The 2026 DI section is not a harder paper. It is a tighter one, and tightness punishes a specific habit: attempting everything.
You have 27 days. Spend the first twelve building method on caselets and the last thirteen building a timed attempt band you trust. On 4 October, open eight sets, commit to five or six, and let the other two go without a second thought. The arithmetic above says that candidate finishes ahead of the one who touched all forty.
One caveat before you close this tab: every number here rests on a pattern that two dated September 2026 pages agree on, not on a notification PDF we have read line by line. Download your call letter the day it drops and check the section table against this article. If IBPS has moved anything again, the call letter will say so first.