First Principles Thinking

A skill for rethinking from fundamentals when stuck or things get too complex.

When to Use

• Implementation got too complex
• Lost sight of what you were trying to build
• Fixated on a technical solution
• Struggling with trade-off decisions
• Don't understand "why it's this way"

Thinking Framework

Step 1: Identify the Essential Problem

## What are we trying to solve right now?

**Surface problem**:
(Technical implementation talk)

**Essential problem**:
(Value to user/business)

**Verification questions**:
- If this problem is solved, who benefits and how?
- If this feature didn't exist, what problems would occur?

Step 2: List and Question Assumptions

## Implicit Assumptions

List assumptions taken for granted in current implementation:

1. [ ] <assumption 1> → Is this really true?
2. [ ] <assumption 2> → Why do we think this?
3. [ ] <assumption 3> → Is there another way?

**Assumptions to question**:
- "Must be..."
- "Impossible to..."
- "Everyone does it this way"
- "It's always been this way"

Step 3: Re-verify Constraints

## Real Constraints vs Assumptions

| Constraint | Type | Basis |
|------------|------|-------|
| Budget $10K | 🔒 Fixed | Already approved |
| Use React | ⚠️ Verify | Just because "team is familiar"? |
| REST API | ⚠️ Verify | Why not GraphQL? |
| Complete in 1 week | 🔒 Fixed | Release date confirmed |

Step 4: Zero-Base Options

## If starting over from scratch?

**Options thought from constraints only, from blank slate**:

1. **Option A**: <description>
   - Pros:
   - Cons:

2. **Option B**: <description>
   - Pros:
   - Cons:

3. **Option C (current direction)**: <description>
   - Pros:
   - Cons:

**What's the simplest solution?**
(Cut features, different approach, option to not do it)

Step 5: Decision and Next Actions

## Conclusion

**Choice**: Option X

**Reasons**:
1. <reason1>
2. <reason2>

**What to drop**:
- <what to drop>

**Next actions**:
1. [ ] <action1>
2. [ ] <action2>

Question List

Questions to ask yourself when stuck:

About the Problem

• "First of all, is this a problem worth solving?"
• "Is the problem definition correct?"
• "Who is this feature for?"

About the Solution

• "What's the simplest solution?"
• "What if we don't build this feature?"
• "How to get 80% of value with 20% of effort?"
• "What if it doesn't have to be perfect?"

About Assumptions

• "Why do we think this?"
• "Is it a fact or speculation?"
• "Could the opposite be true?"
• "Will this assumption still be valid in 5 years?"

Output Format

## First Principles Analysis

### Current State
**What we're trying to do**:
<description>

**Where we're stuck**:
<description>

---

### Essence

**The real problem to solve**:
<core problem>

**Definition of success**:
<what success looks like>

---

### Assumption Verification

| Assumption | Verification Result |
|------------|---------------------|
| <assumption1> | ✅ Valid / ❌ Invalid / ⚠️ Needs verification |

---

### Options After Rethinking

1. **Simple option**: <description>
2. **Modified current option**: <description>
3. **Don't do it option**: <description>

---

### Recommendation

**Choice**: <option>

**Reason**: <rationale>

**What to cut**: <what to cut>

Notes

• Avoid analysis paralysis: Don't overthink, connect to action
• Don't seek perfection: Sometimes a 70% solution is enough
• Share with team: Don't hold alone, borrow perspectives
• Time-box: If no conclusion in 30 minutes, consult