Your friend in the morning slot and you in the afternoon got nearly identical marks. You worked it out from the answer key — same number right, same number wrong, give or take one. Then the scorecards came out and her percentile was higher than yours. Not by a rounding error. By a real, visible gap. You stared at it and thought: how does the same score become two different percentiles? If you have ever felt that the CAT percentile calculation is a black box that does something quietly unfair to you, this is the explainer that opens the box. It is not unfair. It is just doing something most aspirants are never properly told about, and once you see the steps, the CAT percentile calculation turns out to be more sensible than secretive.
Start with what your raw score actually is
Before percentile, there is the raw score, and it is the simple part. Every correct answer gives you +3. Every wrong MCQ costs you −1. Anything you left blank, and the non-MCQ (TITA) questions you got wrong, cost you nothing. Add it up across all three sections — VARC, DILR, and QA — and you have a number out of a possible 198. That is your raw score. If you attempted 50 questions, got 40 right and 10 wrong, your raw score is 40 times 3 minus 10, which is 110.
Here is the first thing that surprises people: that raw score never appears on your scorecard. The official scorecard shows only scaled scores and percentiles. The raw number you carefully computed from the answer key is real, but it is only the starting ingredient. What happens to it next is the whole story of the CAT percentile calculation, and it is the step almost nobody explains in plain language.
Why CAT needs to adjust your score at all
CAT runs in three slots on a single day — morning, afternoon, evening. The three papers are different. The exam setters try hard to make them equally difficult, but you cannot perfectly match the difficulty of three entirely different question papers. One slot ends up a little tougher. Another a little easier. That is not a flaw anyone can fully remove; it is just what happens when lakhs of people cannot sit the same paper at the same time.
Now think about the unfairness that would cause if scores were left raw. If your slot's QA section was brutal and your friend's was gentle, her raw 40 and your raw 40 do not mean the same thing. Hers came easier. Yours was earned against a harder paper. Treating them as equal would punish you for the luck of your slot. So before anything else, the IIMs run a step to correct for exactly this. That correction step is the heart of the whole system, and understanding it is most of what the CAT percentile calculation asks of you.
The scaled score: where the magic happens
The fix is called normalization, and the output is your scaled score. In plain terms, the IIMs look at how everyone in your slot performed versus everyone in the other slots, and they adjust each section's raw scores onto one common scale. The principle is simple even though the statistics behind it are not: if your slot was genuinely harder, your raw score gets nudged upward. If your slot was easier, it may get nudged slightly down. The goal is a single ruler that measures everyone the same way regardless of which paper they faced.
This is why your scaled score can be higher or lower than the raw score you calculated. It is not an error in your maths. It is the system saying, "we have accounted for how hard your particular paper was." Normalization is done section by section first, and then your three scaled sectional scores are added to give your overall scaled score. Every ranking and percentile that follows is built on this scaled number, not your raw one. That single fact resolves most of the confusion people carry about the CAT percentile calculation, because almost every "that's not fair" reaction traces back to comparing raw marks the system never used.
One honest limit worth stating plainly: the exact statistical formula CAT uses is not released to the public. So nobody outside the IIMs can compute your final scaled score by hand, and every "percentile predictor" tool online is making an educated estimate from past trends, not a real calculation. Treat those predictors as rough guidance, never gospel. The official numbers come only from your scorecard, downloadable from the official CAT website at iimcat.ac.in.
From scaled score to percentile: the CAT percentile calculation itself
Once everyone has a scaled score, the percentile part is genuinely simple arithmetic. The IIMs rank every candidate by scaled score, best to worst. Then for any candidate, your percentile is just the share of people who scored below you, turned into a number out of 100. The formula is plain: percentile equals (N minus your rank) divided by N, times 100, where N is the total number of test-takers.
An example makes it concrete. Say 3 lakh people took CAT. If your scaled score puts you at rank 3,000 from the top, then 2,97,000 people are below you. Divide 2,97,000 by 3,00,000, multiply by 100, and you get a 99 percentile. That is all a percentile is: not your marks, but your position in the queue. A 99 percentile means you finished ahead of 99 percent of everyone who sat the exam that year. The CAT percentile calculation rounds this to two decimal places, which is why you see numbers like 98.74 rather than whole figures.
This is also why percentile is relative, not absolute. The same scaled score can map to a slightly different percentile in different years, because it depends entirely on how everyone else did. A score that earned a 99 last year might earn a 98.6 this year if the overall field was stronger. You are not competing against a fixed bar. You are competing against the specific crowd that showed up with you.
A worked example, start to finish
Numbers make this concrete in a way definitions cannot, so walk through one full path from answer key to scorecard. Imagine you attempted 54 questions, got 44 correct and 10 wrong. Your raw score is 44 times 3, which is 132, minus 10, leaving 122 out of 198. You write that number down feeling good about it. But that 122 is the last time you will ever see your raw marks, because everything after this point in the CAT percentile calculation runs on the scaled version instead.
Next, the normalization step looks at your slot. Suppose the evening slot you sat had a noticeably tougher DILR section, and on average people in your slot scored lower there than the morning crowd. The system reads that as "this paper was harder," and nudges the DILR portion of your score upward to compensate. Your raw 122 might become a scaled 128, or it might land at 119 if your slot was the easy one — you cannot know the direction in advance, because it depends entirely on how your slot compared to the others.
Then the ranking happens. Every candidate's scaled score is lined up from highest to lowest. Say your scaled score places you at rank 4,200 in a field of 3.3 lakh. That means roughly 3,25,800 people sit below you. The percentile step divides that by the total and multiplies by 100, landing you near 98.7. Notice what just happened: a clean raw 122 did not "equal" any fixed percentile. The CAT percentile calculation converted your marks into a position, and the position is what the scorecard reports. Two people with the same raw 122 from different slots could easily end up at 98.7 and 98.3, purely because of which paper they drew.
This is the whole machine in one pass: raw marks, then a slot-difficulty adjustment, then a rank, then a percentile from that rank. No step is mysterious on its own. The only genuinely hidden part is the precise statistical weighting inside the adjustment, which is why estimates are estimates.
What this means for how you read your scorecard
Understanding the CAT percentile calculation changes what you should actually look at when results arrive. The instinct is to fixate on the overall percentile alone. But the scorecard gives you sectional scaled scores and sectional percentiles too, and those often matter more for your specific shortlisting than the headline number.
Here is why. A candidate with a 99 overall but a weak 80-something in one section can miss calls from schools that enforce a high sectional cutoff, while someone with a slightly lower overall but balanced sections clears them comfortably. The overall percentile is the number people brag about; the sectional balance is the number that quietly decides outcomes. Reading your scorecard well means checking both, not just the big figure at the top.
So why did your friend's percentile beat yours?
Now the opening puzzle solves itself. You and your friend had the same raw score, but you sat different slots. If your slot was treated as slightly easier by the normalization step, your raw score scaled down a touch. If hers was treated as harder, hers scaled up. Different scaled scores mean different ranks, and different ranks mean different percentiles. Same raw marks, different outcome — and crucially, not because the system cheated you, but because it was trying to be fair to whoever drew the harder paper.
It can feel personal when you are on the wrong side of it. It is not. Over a large enough group the adjustment is even-handed, and in a different year the luck could fall your way. Understanding this does not change your number, but it changes how you carry it. The aspirants who handle results well are usually the ones who understand what the number is measuring before they panic about it, and the CAT percentile calculation rewards that calm more than people expect.
If you want to make real sense of your own scorecard — what your sectional percentiles imply, whether your profile clears specific cutoffs, what your number realistically means for calls — that is a conversation, not a chart. Talking to someone who has actually sat CAT, read their own scaled scores, and been through the call process can decode your specific situation faster than any predictor tool. Platforms like eSalahKaar let you talk to verified students and alumni from the IIMs and other top schools at per-minute pricing, so you pay only for the actual minutes of a real conversation. Worth bookmarking if you are staring at a scorecard and unsure what it actually means for you. You can see how it works on their how it works page, and common doubts are answered on the FAQ.
A few things the percentile is not
Clearing up what the number is also means clearing up what it is not, because a few myths cause real anxiety.
It is not your raw marks. People say "I got 99 percentile" as if it were a score; it is a rank position, and the underlying scaled marks behind a 99 can vary year to year. It is not a guarantee of a call either. IIMs shortlist on a composite score that adds your academics, work experience, and diversity factors on top of the CAT number, and they enforce sectional cutoffs — so a strong overall percentile with a weak section can still miss. And it is not something you can reverse-engineer precisely before results, because the formula is private and the field is unknown until everyone has sat the exam.
Knowing these boundaries keeps you from two common mistakes: over-trusting a predictor that promised you a number you did not get, and despairing over a percentile without checking whether your sectional balance and profile actually clear the schools you care about. The CAT percentile calculation tells you your position; it does not, on its own, tell you your outcome.
The one idea worth keeping
If you remember nothing else, remember this: your percentile is a measure of where you stand relative to everyone else, built on a scaled score that has already been adjusted for how hard your particular paper was. It is not your marks, and it is not a verdict on your worth. It is a position in a queue that the system tried genuinely hard to make fair. Understand that, and the scorecard stops being a black box and becomes just information — which is exactly what you need it to be when you sit down to plan your next move. That, in the end, is all the CAT percentile calculation is: a fair way of turning your marks into your place in a very large line.