Steel-Manning: How to Make Your Opponent's Argument Better Than They Can

Last time, we landed on a hard truth: if you argue to win, you usually lose. The goal of a good argument is to find what's true, not to defeat a person.

So here's the single most powerful technique for actually doing that. It feels backwards. It works like nothing else.

Before you argue against someone's position, you make their case for them — stronger and clearer than they did.

It's called steel-manning, and it's the exact opposite of what most people do.

Strawman vs. steelman

The strawman is the move you already know. You take the weakest, dumbest version of the other person's argument — or a distortion of it — and demolish that. It feels like winning. It's actually fighting a scarecrow you built yourself. You've defeated no one.

The steelman is the reverse. You take their argument and rebuild it into the strongest possible version — the one their smartest ally would make — and engage that.

"So the best version of your view is this: ___, and the strongest reason for it is ___. Have I got that right?"

Then, and only then, do you respond.

Dennett's rules

The philosopher Daniel Dennett laid out how to criticize someone well. The order matters:

• •Re-express their position so clearly and fairly that they say, "Thanks — I wish I'd put it that way."
• •List the points you agree on (especially anything non-obvious).
• •Say what you've learned from them.
• •Only now are you allowed to rebut.

Three of the four steps happen before you disagree. That's not politeness. It's strategy — and it's the same reason you can disagree hard without damaging the relationship.

Why it's so powerful

Steel-manning does three things at once, each worth the effort alone:

• ✓It disarms them. Someone who feels genuinely understood stops defending and starts listening. You've removed the threat, so they no longer need their armor. (People don't soften because they lost — they soften because they were heard.)
• ✓It makes your disagreement bulletproof. Beat the strongest version of their argument and your conclusion is solid. Beat only the weak version and you never actually won.
• ✓It protects you from being wrong. Sometimes, building the best case for the other side, you discover… it's better than yours. That's not a defeat. It's the entire point — you came to find the truth, and you found it.

The ultimate test

There's a brutal way to check whether you've earned the right to dismiss a view: the Ideological Turing Test. Can you state the opposing position so convincingly that a neutral observer couldn't tell you're not a true believer?

If you can't — if the best you can manage is a clumsy, obviously hostile version — then you don't understand the argument well enough to reject it yet. You're not disagreeing with their view. You're disagreeing with your caricature of it.

Most of the positions we're most sure are wrong, we've never actually steel-manned. That should make all of us a little less certain.

Why this is the engine of good decisions

Here's where it stops being a debate trick and becomes the core of how teams decide well.

A group only makes a good decision if every real option gets its strongest hearing. If the alternatives get strawmanned — "well, obviously we can't do X" — the team isn't choosing between the real options. It's choosing between one real option and a row of scarecrows.

Steel-manning is how you make sure the best argument actually wins: give each side its most powerful form, then let them compete. It's slower. It's also the difference between a decision that survives contact with reality and one that collapses the moment someone outside the room makes the case you never let anyone make. Building it into the process is what structure is for.

The strongest thing you can do for your own argument is to first make the other one as strong as it can possibly be.

The deeper treatment — where the technique comes from, the research on whether restating the other side first really makes you more persuasive, and how to run it in a meeting — is here: Steelmanning: argue the other side better than they can.

Pick a view you're confident is wrong. Can you state its strongest version — the one a smart believer would give — well enough that they'd nod? Try it in the replies.

I'll tell you honestly whether your steelman holds, or whether you've built a scarecrow.