It's Wednesday night. Your exam is Saturday morning. You have twelve slide decks, a 300-page textbook, two problem sets, and the creeping certainty that you're going to fail.

I want to be honest with you: three days is not ideal. If you had three weeks and did this protocol end-to-end you'd walk into the exam relaxed. But three days is what you have, and three days is enough — if you stop doing what cramming-under-panic tells you to do.

The classic mistake is to re-read everything, panic more, switch to highlighting, panic more, then try to redo all the practice problems starting from problem 1. That's a recipe for spending 30 hours on 40% of the material. You end up memorising the easy stuff (already knew it) and skipping the hard stuff (the reason the exam is hard).

This tutorial walks you through the opposite approach: the Ritsu 72-Hour Drill Protocol. You triage ruthlessly. You drill what you don't know. You review the schedule. You pass.

Prerequisites

A free Ritsu account (Plus tier if you're uploading >5 large slide decks — otherwise the free tier is fine).
Every piece of material the exam could pull from: slide decks, lecture notes, assigned readings, problem sets. Don't curate. Dump everything in.
Honest self-awareness about what you understand vs. what you've just glanced at. This protocol works by being brutally honest about gaps.
6-8 hours per day for three days. That's the minimum to drill a full course's worth of material. More is better but not by much — quality of reps matters more than quantity.

What you'll build

You're building a drill machine around yourself over three days:

Day 1 (Triage): a weakness map across all material. You'll know exactly where your 5-10 biggest gaps are.
Day 2 (Drill): targeted active recall on those gaps, plus past-problem work to build speed.
Day 3 (Review): spaced repetition of everything from Days 1-2, plus full-mock-exam pressure testing.

The protocol is designed so Day 1 is the most important. Get the weakness map right and the next two days are execution. Skip it and you're back to flat cramming.

Steps

1. Day 1, Hour 0 — Dump everything into one session

Open Ritsu. Create a new session and name it [Course Name] Finals Drill. Upload every slide deck, every lecture note, every PDF, every problem set. Even the handout you skimmed once and put in a drawer.

Ritsu handles multi-document sessions — it'll index everything together so queries span the whole corpus. This matters because real exams test across topics, and you want your drill to mirror that.

While it's indexing, do one thing: find the course syllabus or any "what's on the exam" email from your instructor and paste it into the chat. A single line saying "final covers chapters 1-8, weighted heavily on integration techniques" changes the entire drill plan.

2. Day 1, Hour 1 — Build the weakness map

Here's the most important command in this whole protocol:

/quiz 20 questions across ALL material — spread evenly across topics, not one topic per question

Take the quiz. Be honest. If you half-remember the answer, that counts as wrong. The goal is not to make yourself feel better; it's to find gaps.

When you're done, type:

/progress — which topics did I score worst on?

Ritsu returns a weakness breakdown: "You're at 30% on Fourier transforms, 45% on convergence tests, 70% on basic integration, 85% on series representations." This is your target list. You will spend the next 48 hours drilling the bottom three, lightly reviewing the middle, and completely ignoring the top.

Write those 3-4 topics down somewhere you'll see them. On a post-it. On your wallpaper. Wherever. They are the only thing that matters until Saturday.

3. Day 1, Hour 2-8 — Deep drill the #1 weakness

Pick your worst topic. In our example, "Fourier transforms at 30%." Type:

/start Fourier transforms

Ritsu gives you the structured learning track for that topic within the course materials. Work through it with these commands in order:

/eli5 — force the core intuition
/why — understand when this tool is used
/derive — walk through the key formulas (intensely useful for math/physics/engineering)
/quiz — 10 questions, targeted at this topic
/solve [a specific problem from your problem sets] — Ritsu walks you through problem-solving, not just answer

Rinse and repeat until your quiz score on that topic hits 80% or higher. Depending on the gap, this is 2-6 hours of focused work. Do not move to the next weakness until this one is drilled.

By end of Day 1 you should have closed your single worst gap. This is more progress than most people make in three days of cramming.

4. Day 2 — Close the next two gaps + schedule reviews

Wake up Day 2 and do the hardest thing: don't re-drill yesterday's topic. You closed it yesterday; revisiting it now is comfort-seeking. Trust the protocol.

Instead, type:

/next

Ritsu moves to your #2 weakness and builds the same workflow — eli5, why, derive, quiz, solve — but now you have Day 1's confidence momentum. Expect this closure to take 4-5 hours, not 6.

Do the same for #3 weakness in the afternoon. Three weaknesses in one day.

Critically: at the end of Day 2, type:

/flashcard — generate cards for everything from Day 1 and Day 2

Review those flashcards before bed. Ten minutes. This is the ONLY thing you do before sleep. Spaced repetition during exam week is a cheat code — it halves how much you forget overnight.

5. Day 3 morning — Mock exam under pressure

This is where most students mess up. They keep drilling new material on Day 3. That's the worst thing you can do. By Day 3 you've closed your three biggest gaps; more drilling has diminishing returns. The limiting factor now is retrieval speed under pressure — can you produce the answer fast, in exam conditions, with a ticking clock.

Here's the protocol for Day 3 morning:

/test — full mock exam matching the real exam format

Ritsu builds a timed mock exam from all your material. If you have a past exam from the instructor, upload it — Ritsu will mirror its style. Sit the mock exam like the real thing. Phone in another room. Water bottle. Timer.

When you finish, DON'T grade it yet. Stand up, walk for 10 minutes, come back. Then type:

/review — grade my mock exam and explain every question I got wrong

Ritsu explains each miss, maps it to the underlying concept, and tells you whether it's a knowledge gap (study more) or an execution gap (sloppy, go slower tomorrow). Knowledge gaps are where you spend the afternoon.

6. Day 3 afternoon — Targeted patches + full review

Spend 2 hours on any new gaps the mock exam surfaced. Same commands as before: /eli5, /why, /quiz. But keep it short — you're patching, not overhauling.

Then spend the last 2 hours on review:

/flashcard --review-all

Blitz every flashcard from Days 1-3. Each one should take 3-5 seconds. If you pause on one for more than 10 seconds, tag it for one more review before bed.

Stop studying by 8pm. Don't look at your notes again tonight. Sleep 8 hours. Seriously. Sleep consolidates memories and cramming at 2am undoes the entire protocol.

Troubleshooting

"I have more than 3 days but less than a week." You're lucky. Run Day 1 across two days (more triage, deeper initial drilling), then Days 2-3 as written, then spend extra days on additional weaknesses. Don't just double the time on the same gaps — breadth plus depth.

"I'm overwhelmed uploading 15+ files." Upload them in two Ritsu sessions: one for slide decks, one for readings. Run the weakness map on both separately, then combine the top-3 from each. This is 10 minutes of extra setup and the gain in clarity is worth it.

"My weakness map has 8 topics I scored below 50% on." Your course has a lot of ground to cover and three days is genuinely not enough. Take the top 3 and ignore the rest completely. Better to ace 60% of the exam than half-cram 100% of it.

"I don't have past exams." Type /test --style=textbook and Ritsu builds a mock from the structure of assigned problem sets. Not as good as a real past exam, but close enough.

"I keep panicking during the mock exam." That IS the point. Day 3 morning's mock is supposed to simulate the panic so Saturday feels familiar. If you flatten your nervous system by sitting the mock under real conditions, the actual exam is a repeat performance. First time always feels worse.

Try it yourself

The 72-Hour Drill Protocol isn't a hack. It's what every student who ends up getting an A under time pressure is already doing intuitively — just slower, less structured, and without active recall. Ritsu makes the structure explicit and handles the grading + scheduling so your brain only does the actual learning.

If your exam is three or four days away right now and you're reading this instead of studying, stop reading and go run Step 1. Open Ritsu, upload your materials, get the weakness map. The 20-minute investment of Day 1 Hour 1 is going to be worth more than the next six hours of panic-reading combined.

Start your 72-hour protocol below. Ritsu will walk you through upload, triage, and drill planning. Pull out your slide decks and go.