What My Own Session Lengths Actually Look Like
Across 340 logged sessions in my own Karma Yogi account, the average length is 59 minutes. My single longest session is 190 minutes - just over three hours. Neither number is 25. If I had been running strict Pomodoro the whole time, both would be capped near it, and I would have needed the interruption machinery of Pomodoro to reassemble a 190-minute DILR block out of eight separate 25-minute intervals with forced breaks in between. I never did that, because for the sessions that actually ran long, the length was the point - not something to chop up.
That gap between 59 and 190 is not noise. It is two different kinds of session that get lumped together under "studying" and shouldn't be timed the same way.
What the Pomodoro Technique Actually Is
Francesco Cirillo built the Pomodoro Technique in the late 1980s as a personal tool for a specific problem: shallow, interruptible task-switching - the kind of day where you're juggling emails, small tasks, and the temptation to check your phone every few minutes. The cycle is 25 minutes of work, a 5-minute break, repeated four times, then a longer break. It is good at what it was built for: reducing decision fatigue and making an intimidating to-do list feel approachable in small chunks.
CAT prep is not that kind of work, and the mismatch shows up exactly where you'd expect - in the sections that require you to hold a mental structure in your head across many minutes at once.
Where a Fixed 25-Minute Timer Actively Hurts
Reading Comprehension needs the passage to stay live in your head
A CAT RC passage runs 500-700 dense words, usually argument-driven. Building an accurate mental model of the author's position takes a few minutes on its own, and then you're applying it across 4-6 questions. If a 25-minute timer goes off midway through that, the "break" it forces doesn't rest you - it erases the model you were holding, and you rebuild it from scratch afterward. That's not a pause; it's a restart with a five-minute tax attached.
DILR sets punish every restart
A real arrangement or scheduling puzzle asks you to hold six to ten constraints in working memory while you test configurations against them. Getting into that state takes several minutes of cold start on its own. My 190-minute session was a DILR-heavy block, and it stayed that long because stopping at any of the natural 25-minute marks inside it would have meant re-establishing the constraint set from nothing - twice, maybe three times, for no benefit. Long blocks aren't indulgence here; they're the only way the work actually gets done once.
A count of intervals isn't a study record
Finishing "six Pomodoros" tells you nothing about what happened inside them - which subject, which topic, whether accuracy moved. My own logs record subject and topic per session, which is why I can say the 190-minute outlier was DILR and not just "a long day." An interval count can't answer that question even in principle.
Where a Short Fixed Timer Genuinely Helps
This isn't an argument against timers - it's an argument against using one timer length for everything. Several kinds of CAT prep work are genuinely short and repetitive, and a fixed interval helps them the same way it helps any shallow task:
- Vocabulary drills - a round of new words plus a quick flip-card review is naturally a 15-20 minute task. A timer's main job here is just getting you to start.
- Formula and identity recall - running through algebraic identities or time-speed-distance formulas is shallow recall, not deep reasoning, and benefits from short, spaced bursts rather than one long sitting.
- Theory reading - reading a Quant chapter for concept understanding, as opposed to solving problems from it, tolerates a break every 25 minutes without losing much, because there's no live mental model to protect.
The dividing line isn't the subject - it's whether the task needs an uninterrupted mental model held across time, or whether it's small enough to survive being chopped into pieces.
A Concrete Example: Reading the Newspaper Doesn't Need 190 Minutes
I also track a subject called NEWSPAPER - the daily reading habit that feeds VARC comprehension without being a mock section itself. Those sessions are short by nature: read for 20-30 minutes, log it, done. Running a Pomodoro-style timer on that kind of session costs nothing, because there's no fragile mental model to protect across an interruption - unlike an RC set, the newspaper doesn't fall apart if you put it down and pick it back up later. It's a useful sanity check for the rule above: if stopping mid-task and resuming ten minutes later loses you nothing, a short fixed timer is fine or even helpful. If it loses you the whole structure you were building, it isn't.
What I Actually Do: Mixed Session Lengths, Logged Honestly
My own 59-minute average sits in between the two extremes on purpose. It's long enough to get through a real RC set or a short DILR puzzle without cutting it off, but most days don't demand a 190-minute block - that length happened once in 340 sessions because that specific day's material needed it. I don't set a timer length before I know what the session is for. I decide the task first - a puzzle, a passage, a vocabulary round - and let the timer match it, using stopwatch mode when I don't know exactly how long a DILR set will take and don't want an arbitrary cutoff to end it early.
What actually changed my prep wasn't picking Pomodoro or rejecting it - it was logging session length, subject and topic together, so the pattern above was visible to look back on rather than something I had to guess at. Two weeks of "felt like a lot of Quant" turns into an actual number once every session is logged with subject and duration - the same underlying habit that keeps a study streak alive rather than a vague once-a-week intention. A 190-minute outlier stops being a fuzzy memory and becomes a data point you can decide whether to repeat.
The Practical Rule
Use a short fixed timer for anything shallow and repetitive - vocabulary, formulas, theory reading. Use an open-ended or long timer for anything that requires holding a structure in your head - RC passages, DILR sets, full-length problem-solving blocks. Don't let the short bursts crowd out the long blocks; the long blocks are where the actual reasoning practice for CAT lives, and they're the ones a rigid 25-minute interval quietly undermines.
If you genuinely don't know how long a session will run - which is most DILR sets before you've read the puzzle - don't guess a Pomodoro count in advance. Start a stopwatch, work until the set is actually solved or you've hit a real wall, and stop then. The 190-minute session in my log wasn't planned as a 190-minute session; it became one because the material demanded it, and a pre-committed interval count would have forced an arbitrary stopping point in the middle of it for no reason other than the clock.
Start tracking free and let your own session lengths tell you which kind of studying you're actually doing on a given day.
End of essay
- Anish Guruvelli