mingyao743/make-rational-decision/hat-haircut-tattoo

Hat–Haircut–Tattoo Theory

A three-layer model for classifying how hard something is to change in a person:

观点像帽子，性格像发型，价值观像刺青。

Opinions are hats. Character is a haircut. Values are tattoos.

The core insight: these three layers have radically different change costs. Misjudging which layer a conflict belongs to is the most common reason debates fail and persuasion backfires.

When to Use

• A discussion keeps going in circles despite good arguments on both sides
• You are preparing to persuade someone and want to pick the right strategy
• You must decide whether a disagreement is worth arguing at all
• You are trying to change your own behavior and keep failing
• You are evaluating a person (candidate, partner, partner-team) and want to know what to trust: words, patterns, or trade-offs

The Three Layers

Diagnostic Questions

Before engaging, classify which layer you are actually dealing with:

1. Hat test: "Would this person plausibly change their mind this week given new information?" — If yes, it's an opinion. Argue with data.
2. Haircut test: "Does this person acknowledge the trait, intend to change it, but repeatedly fail?" — It's character/habit. Arguing harder won't help; only structure and time will.
3. Tattoo test: "Is this tied to their identity, upbringing, or worldview — would changing it feel like becoming someone else?" — It's a value. Stop optimizing for persuasion.

Strategy Selection

Once classified, choose deliberately:

• Hat conflict → Engage. Ladder up: facts → interpretation → proposal. Cheap to try, cheap to lose.
• Haircut conflict → Coach, don't convince. Replace argument with systems: triggers, reminders, accountability, habit loops. Measure progress in weeks, not conversations.
• Tattoo conflict → Navigate, don't convert. Options in order of preference:

1. Find overlap ("we differ on X, agree on Y")
2. Set boundaries ("let's not discuss X")
3. Accept the trade-off explicitly (in hiring, partnership, relationships)
4. Walk away if the tattoo directly conflicts with your own tattoos

Applying It to Yourself

Self-change follows the same ladder:

• Changing an opinion: read one good counterargument
• Changing a habit: 60+ days of deliberate practice, environmental friction removal
• Changing a value: usually requires lived experience, not reasoning — travel, loss, responsibility, proximity to different people

If you keep "knowing" something but not acting on it, stop polishing the argument (hat work) and start building the system (haircut work).

Using It to Read People

Don't judge by what people say — judge by layer:

• Hats (what they claim today): weakest signal
• Haircut (recurring behavior patterns over months): moderate signal
• Tattoos (what they consistently sacrifice for, under pressure): strongest signal

In interviews, dating, and partnerships: listen to hats, observe haircuts, watch what they protect when costs arrive — that's the tattoo.

Output Format

## Hat–Haircut–Tattoo Analysis: [Topic / Conflict]
 
### Classification
| Item | Layer | Evidence | Confidence |
|------|-------|----------|------------|
| [Position/behavior 1] | 🎩 Hat / 💇 Haircut / 🖋 Tattoo | [Why] | High/Medium/Low |
 
### Recommended Strategy
- [For each item: engage / coach / navigate — with concrete approach]
 
### What NOT to Do
- [Predictably wasted efforts, e.g. "debating the tattoo-level topic"]
 
### Relationship / Self Plan
- [Coexistence arrangements, boundaries, or self-change systems]

Extensions

This skill is enhanced by the following shared extensions:

• `_shared/skill-workflows.md` — Workflow 8 (Disagreement Resolution) uses this skill as the triage step before Six Hats
• `_shared/examples-library.md` — Add a worked example when available

Rules

• Never treat a values-level disagreement as an opinion-level one — this is the #1 failure mode
• Classification is per-topic, not per-person: the same person holds hats, haircuts, and tattoos simultaneously
• Layers can migrate (a repeated opinion becomes a value; a shock event removes a tattoo) — reassess periodically
• When in doubt, probe once with a genuine question rather than assuming the layer
• Winning the argument is not the goal; choosing the right intervention is