Browse
Thinking

Argue Both Sides Before I Decide

For big decisions. Claude builds the strongest case for each option, exposes the assumptions underneath, and suggests a cheap test before you commit.

No votes yet this week
Output
Text
Added
Updated

The prompt

I have to decide: [THE DECISION, e.g. "Should we rewrite our app in a new framework?"]

Context: [FACTS, CONSTRAINTS, TIMELINE, WHAT IS AT STAKE]
Where I am leaning, and why: [YOUR CURRENT VIEW]

Do this in order:
1. Make the strongest honest case for each option, the way a smart advocate would. No weak versions.
2. For each side, list the assumptions it quietly depends on.
3. Tell me which of my own assumptions looks weakest, given the context.
4. If each option were the right call, what would I expect to see three to six months from now? Give me signals I could actually watch.
5. What is the cheapest experiment or piece of information that would reduce the uncertainty before I commit?
6. Only then give your recommendation, with a confidence level, and the one thing that would change your mind.
  • decisions
  • strategy
  • critical-thinking

What it generates

Example

The case for rewriting

Your current stack is why hiring takes four months. Good candidates see it and decline. Every feature takes about twice the estimate because of workarounds. A rewrite buys back years of speed, and the team is still small enough to pull it off.

The case for staying

Most rewrites overrun, and some never ship. Customers pay for features, not frameworks, and a nine month freeze hands your roadmap to competitors. The pain you describe sits in three modules, not in the whole app.

What each side assumes

  • Rewrite: the new framework is still the right choice in three years, and the team can keep two systems in their heads at once.
  • Stay: hiring is slow because of the stack, and not because of salary or location.

Your weakest assumption

"It will take four months." Your last two large projects ran 2.1 times over their estimates. Plan for eight or nine.

What you would see in six months

  • If rewriting is right: the first migrated module ships changes in half the time, and candidates stop asking about the stack in interviews.
  • If staying is right: refactoring the three painful modules removes most of the workarounds, and estimates start to hold.

Cheapest experiment

Rewrite only the billing module, behind a feature flag. Give it two weeks. Measure how long the next billing change takes, and whether old and new code can share authentication without hacks.

Recommendation

Migrate piece by piece instead of all at once. Confidence: about 75%. I would change my mind if the billing experiment shows the two systems cannot share authentication cleanly.

NoteThis is an example. Results change between runs and model versions. Prompt Podium is a fan project and is not affiliated with Anthropic.

Getting the most from this prompt

When to use it

Decisions that are expensive to reverse: a rewrite, a hire, a pricing change, a move to another city, leaving a job. It helps most when you already lean one way and want to know if you are fooling yourself.

Make it your own

  • Say which way you lean and why. Step three, which attacks your weakest assumption, only works if your assumptions are on the table.
  • With more than two options, list them all, but ask for the full treatment on the top two only.
  • For a team decision, paste each person's position and ask which disagreements are about facts and which are about values.

Watch out for

The recommendation at the end is the least valuable part. The cheap experiment in step five is the most valuable, because it turns an argument into something you can test in days. If both cases sound equally strong, the decision probably matters less than it feels, and the better move is to pick the option that is easier to undo.

How to use this prompt

  1. [01] Copy

    Copy the prompt

    Use the copy button above. The full text is included, nothing is cut off.

  2. [02] Fill in

    Replace the yellow parts

    Everything highlighted in yellow is a placeholder. Swap it for your own topic, audience or data.

  3. [03] Vote

    Run it in Claude and vote

    If the result helped, press the heart. Votes decide which prompts make the weekly top 10.

Learn the technique

All articles
2 min read

How the Weekly Ranking Works

One vote per prompt, a fresh count every Monday, and a simple tie-break. Everything about how prompts reach the top three, in plain words.

2 min read

How to Use XML Tags in Claude Prompts

Wrapping parts of your prompt in simple tags is one of the easiest ways to get more reliable answers from Claude. A practical guide with examples.