| name | second-order-thinking |
| description | Chain-reaction mapping — trace consequences beyond the obvious first effect to reveal hidden risks and opportunities. |
| group | thinking |
| keywords | ["second-order","consequences","chain-reaction","effects","howard-marks","systems","downstream"] |
| task_strategies | ["spike","investigation"] |
| stream_affinity | ["roadmap","research"] |
| allowed-tools | ["Read","Glob","Grep","AskUserQuestion"] |
| status | experimental |
| source | Howard Marks — The Most Important Thing (2011); Garrett Hardin — Filters Against Folly (1986) |
| acquired | 2026-06-15 |
Second-Order Thinking
And then what? And then what after that?
Howard Marks: "First-level thinkers look for simple formulas and easy answers. Second-level thinking is deep, complex, and convoluted." Most decisions fail not from their direct effects but from their cascades.
When to use
- Before implementing a change that affects many people or systems
- When evaluating a decision that seems obviously good (a sign first-level thinking is in play)
- When something feels like an improvement but you sense a catch
- When designing incentive structures, policies, or architecture
- Pairs with
/brana:systems-thinking for complex adaptive systems
Step 1 — State the action or decision
What change or decision are we analyzing?
AskUserQuestion: "What action or decision are we tracing consequences for?"
Step 2 — Map first-order effects
What are the direct, immediate consequences? These are what most people think about.
Format:
Action → [direct effect 1]
Action → [direct effect 2]
Action → [direct effect 3]
Step 3 — Map second-order effects
For each first-order effect: "And then what?"
[direct effect 1] → [consequence 1a]
[direct effect 1] → [consequence 1b]
[direct effect 2] → [consequence 2a]
Focus these lenses:
| Lens | Question |
|---|
| Behavioral | How will people adapt their behavior in response? |
| Competitive | How will competitors or opponents respond? |
| Resource | What gets consumed, depleted, or crowded out? |
| Systemic | What feedback loops does this trigger? |
| Temporal | What looks good short-term but bad long-term (or vice versa)? |
Step 4 — Map third-order effects (optional, high-stakes)
For the most significant second-order effects: "And then what after that?"
Stop at third order unless the chain is still producing surprises. Most useful insight lives in 2nd order.
Step 5 — Identify hidden risks and opportunities
From the full map:
Hidden risks — second/third-order effects that undermine the goal or create new problems:
Risk: [effect chain] → threatens [X]
Severity: high / medium / low
Hidden opportunities — second/third-order effects that could be amplified:
Opportunity: [effect chain] → enables [X]
How to capture: [specific action]
Step 6 — Output
**Second-order analysis of [action/decision]**
1st order: [summary of direct effects]
Critical 2nd-order surprises:
- [effect] → [consequence] [risk/opportunity]
- [effect] → [consequence] [risk/opportunity]
Decision impact:
→ Proceed / Modify / Avoid
Modification: [if needed]
Notes
- The goal is not to paralyze with consequences but to anticipate the most important surprises
- "Obviously good" decisions most need this analysis — consensus on first-level is a warning sign
- Hardin's filter: run the analysis through literacy (what does it say?), numeracy (at what scale?), and ecolacy (and then what?) lenses
- Time horizon matters: explicitly distinguish short-term vs long-term effects
- Pair with
/brana:pre-mortem after this analysis to stress-test the full consequence map