Solve Textbook Exercises the Right Way: Understanding, Not Answers

There's a specific kind of frustration unique to textbook problems. You've read the chapter. You've watched the lecture. You understand the example the author worked through. Then you open the problem set and it's like none of that happened. You stare at problem 3 for forty minutes, try the obvious approach, fail, flip to the back of the book, find the answer is 137, and now you're even more confused because you don't know how they got 137.

This is the gap between watching someone solve a problem and solving one yourself. And it's the gap where the ROI of Ritsu is highest. Because the right answer doesn't teach you anything. The right process does.

This tutorial shows you how to use Ritsu to work through textbook exercises in a way that makes you better at the next problem — not just this one. It works for calculus, physics, linear algebra, economics, signals and systems, classical mechanics, organic chem, programming exercises, anything where "there's a problem and you need to get unstuck."

Prerequisites

• A free Ritsu account.
• A textbook exercise you're stuck on. Have the problem statement, the relevant chapter/section, and (if you have them) the notes or slides from your professor.
• About 15-30 minutes per problem. More if the problem is genuinely hard; Ritsu isn't trying to rush you.
• An honest commitment to do the problem with Ritsu, not copy from it. This is the one tool where cheating yourself costs you more than anyone else.

What you'll build

By the end you'll have:

1. A clean solution to your problem, written out step by step with your reasoning visible.
2. A mental model of the problem type — so the next problem in this category takes you half the time.
3. A targeted flashcard or two on the technique you just learned, scheduled for review before your next exam.

Notice what you won't build: a dependency on Ritsu where you can't solve problems without it. The protocol below is explicitly designed so you're doing the work, Ritsu is the guide. By problem four you'll be solving them on your own and using Ritsu only when genuinely stuck.

Steps

1. State the problem precisely

Open Ritsu. Start a session (name it "[Textbook] problem sets — chapter [N]" — you'll reuse this session for every problem in the chapter, which is better than one session per problem). Paste the problem statement, verbatim.

If the problem references figures or equations from the textbook, paste those too. Ritsu handles LaTeX, screenshots, and embedded images natively. Do NOT paraphrase the problem. Paraphrasing often introduces the exact ambiguity that was tripping you up.

Then — and this part is crucial — type what you've already tried:

I've tried: [what you tried]
I got stuck at: [where you got stuck]
I suspect the issue is: [your guess]

This three-line preamble changes everything. Without it Ritsu will default to a generic walkthrough that might skip the exact place you're stuck. With it, Ritsu knows to zoom in on your confusion point.

2. Ask for the approach, not the solution

Here's the command that makes all the difference:

/how should I approach this problem? Don't solve it yet — just tell me the method and where I might go wrong.

Ritsu returns the solution strategy: "This is a related-rates problem. The key insight is that both r and V depend on t; you'll differentiate V = (4/3)πr³ with respect to t and then substitute the given values. A common mistake is forgetting the chain rule."

Stop. Close Ritsu. Try the problem yourself with that strategy. You'd be surprised how often the approach is all you needed — the mechanics come easily once you know the method.

If you get the answer: great, run step 5 to cement it. If you're still stuck: go to step 3.

3. Step through it with /solve

When you're genuinely stuck — not "I haven't tried" stuck, but "I tried the approach and it didn't work" stuck — type:

/solve — but show me one step at a time and wait for my input before moving on

Ritsu presents the first step. You confirm you understand it (or ask a follow-up). Ritsu presents the second step. Repeat.

This is slower than just reading a solution. That's the point. When you read a worked solution all the way through, your brain stores it as a "thing I've seen." When you confirm each step actively, your brain stores it as a "thing I can do." Two weeks later the difference is massive.

At any step, ask:

/why that step? Why not [alternative approach]?
/what would happen if I used [different method]?

These meta-questions are where real problem-solving competence is built. You're not just learning how to do this problem — you're learning how to choose between approaches. That's the skill the exam tests.

4. Redo the problem from scratch, solo

This is the step people skip and it's the one that moves the needle. After you and Ritsu have worked through the problem together, close the chat. Open a fresh page. Solve the problem from the beginning, without help.

If you can't: go back to step 3 and find exactly which step tripped you. That's the knowledge gap. Patch it. Then try again.

You haven't solved the problem until you've solved it alone. The version where Ritsu nudged you is a scaffolded version — useful for learning, but not the same as being able to do it on an exam under pressure.

5. Lock in the technique with a flashcard

You've solved the problem. You can redo it. Now the only thing left is to make sure you still remember the technique in two weeks. Type:

/flashcard — create a card that captures the key technique from this problem, not the specific answer

Ritsu writes a card like:

Front: In a related-rates problem with a changing sphere, how do you relate dV/dt and dr/dt? Back: Differentiate V = (4/3)πr³ with respect to t: dV/dt = 4πr² · dr/dt. Then plug in the known values.

This is the actual skill being tested. Problem-specific answers (r = 4, V = 268) are worthless; generalisable techniques are what you'll actually be quizzed on.

The card goes into spaced-repetition review. You see it tomorrow, in four days, in two weeks. By the time the exam rolls around it's second nature.

Troubleshooting

"Ritsu gave me an approach but I still can't execute it." That means you have a prerequisite gap. Type /what do I need to know to execute this approach? — Ritsu backtracks and teaches you the underlying skill, then you come back to the problem. Takes an extra 15 minutes but patches the root cause rather than duct-taping over it.

"I'm worried I'll get dependent on Ritsu for problems." Use the "every fourth problem rule": for every three problems you solve with Ritsu's guidance, do the fourth completely solo. If you can, you're building real skill. If you can't, you've been over-relying and need to redo steps 2-4 more rigorously. This rule works.

"My textbook problem involves a specific piece of notation my professor uses that isn't standard." Paste your professor's notation guide or slide deck into the session first. Ritsu will use your course's conventions rather than the textbook's, which matters when you're graded on them.

"I'm doing coding exercises, not math." Everything above works identically. Use /code instead of /derive, and the "step through then redo from scratch" protocol applies even more strongly — typing code yourself is the only way to build muscle memory. Close the Ritsu window, open a blank editor, write the solution, run it, debug it, check it.

"The textbook answer and Ritsu's answer differ." First, check your algebra. Genuinely. Second, if they still differ, paste the textbook's answer into Ritsu and ask /reconcile — why do we differ?. Sometimes the textbook has a typo. Sometimes Ritsu went down a wrong path. Sometimes there are multiple correct answers in different forms. Whichever it is, you'll learn something.

Try it yourself

There's probably a problem you've been avoiding right now. The one you stared at yesterday and decided to "come back to later." Here's the deal: open Ritsu, paste the problem, work through steps 2-5 above. Total time is 20 minutes. At the end you'll have (a) solved the problem, (b) understood the technique, (c) a flashcard scheduled for review.

The alternative is spending another two hours tomorrow being stuck, or worse, Googling for an answer and learning nothing.

Click below to walk through a problem right now. Ritsu will ask what you've tried, coach you through the approach, and make sure you can redo it solo before you move on.