Initial commit

skills/hypotheticals-counterfactuals/resources/evaluators/rubric_hypotheticals_counterfactuals.json

@@ -0,0 +1,211 @@
+{
+  "criteria": [
+    {
+      "name": "Scenario Plausibility",
+      "description": "Scenarios are possible given current knowledge, not fantasy. Counterfactuals were realistic alternatives at decision time.",
+      "scale": {
+        "1": "Implausible scenarios (magic, impossible foreknowledge). Counterfactuals couldn't have been chosen at the time.",
+        "3": "Mostly plausible but some unrealistic assumptions. Counterfactuals stretch believability.",
+        "5": "All scenarios plausible given what was/is known. Counterfactuals were genuine alternatives available at decision time."
+      }
+    },
+    {
+      "name": "Minimal Rewrite Principle (Counterfactuals)",
+      "description": "Counterfactuals change as little as possible to isolate causal factor. Not multiple changes bundled together.",
+      "scale": {
+        "1": "Many factors changed simultaneously. Can't tell which caused different outcome. 'What if X AND Y AND Z...'",
+        "3": "Some attempt at isolation but still multiple changes. Primary factor identified but confounded.",
+        "5": "Perfect isolation: single factor changed, all else held constant. Causal factor clearly identified."
+      }
+    },
+    {
+      "name": "Causal Mechanism Specification",
+      "description": "Explains HOW change leads to different outcome. Not just stating result but tracing causal chain.",
+      "scale": {
+        "1": "No mechanism specified. Just outcome stated ('sales would be higher') without explanation.",
+        "3": "Partial mechanism. Some causal steps identified but incomplete chain.",
+        "5": "Complete causal chain: initial change → immediate effect → secondary effects → final outcome. Each step explained."
+      }
+    },
+    {
+      "name": "Probability Calibration",
+      "description": "Scenarios assigned probabilities based on evidence, base rates, analogies. Not all weighted equally.",
+      "scale": {
+        "1": "No probabilities assigned, or all scenarios treated as equally likely. No base rate consideration.",
+        "3": "Rough probabilities assigned but weak justification. Some consideration of likelihood.",
+        "5": "Well-calibrated probabilities using base rates, analogies, expert judgment. Sum to 100%. Clear reasoning for each."
+      }
+    },
+    {
+      "name": "Pre-Mortem Rigor",
+      "description": "For pre-mortems: follows 6-step process, generates novel failure modes specific to context, not just generic risks.",
+      "scale": {
+        "1": "Generic risk list copied from elsewhere. Hindsight bias ('obvious' failures). No structured process.",
+        "3": "Some specific risks but mixed with generic ones. Process partially followed.",
+        "5": "Rigorous 6-step process: silent brainstorm, round-robin, voting, mitigations. Context-specific failure modes identified."
+      }
+    },
+    {
+      "name": "Action Extraction",
+      "description": "Clear extraction of common actions, hedges, options, and decision points from scenarios. Not just stories.",
+      "scale": {
+        "1": "Scenarios developed but no actions extracted. 'Interesting stories' with no operational implications.",
+        "3": "Some actions identified but incomplete. Missing hedges or options.",
+        "5": "Comprehensive action plan: common actions (all scenarios), hedges (downside), options (upside), decision triggers clearly specified."
+      }
+    },
+    {
+      "name": "Leading Indicator Quality",
+      "description": "Indicators are observable, early (6+ months advance), and actionable. Clear thresholds defined.",
+      "scale": {
+        "1": "No indicators, or lagging indicators (show scenario after it's happened). No thresholds.",
+        "3": "Some leading indicators but vague thresholds or not truly early signals.",
+        "5": "High-quality leading indicators: observable metrics, 6+ months advance notice, clear thresholds, trigger specific actions."
+      }
+    },
+    {
+      "name": "Scenario Diversity",
+      "description": "Scenarios are qualitatively different, not just magnitude variations. Cover meaningful range of futures.",
+      "scale": {
+        "1": "Scenarios differ only in magnitude (10% growth vs 15% vs 20%). Basically same story.",
+        "3": "Some qualitative differences but scenarios too similar or narrow range.",
+        "5": "Meaningfully different scenarios: qualitative distinctions, broad range captured, distinct strategic implications for each."
+      }
+    },
+    {
+      "name": "Bias Avoidance (Hindsight/Confirmation)",
+      "description": "Avoids hindsight bias in counterfactuals, confirmation bias in scenario selection, anchoring on current trends.",
+      "scale": {
+        "1": "Strong hindsight bias ('we should have known'). Only scenarios confirming current view. Anchored on status quo.",
+        "3": "Some bias awareness but incomplete mitigation. Mostly avoids obvious biases.",
+        "5": "Rigorous bias mitigation: re-inhabits decision context, considers disconfirming scenarios, challenges assumptions, uses base rates."
+      }
+    },
+    {
+      "name": "Monitoring and Adaptation Plan",
+      "description": "Defined monitoring cadence (quarterly), indicator tracking, scenario probability updates, adaptation triggers.",
+      "scale": {
+        "1": "No monitoring plan. Set-it-and-forget-it scenarios. No updates planned.",
+        "3": "Informal plan to review occasionally. No specific cadence or triggers.",
+        "5": "Detailed monitoring: quarterly reviews, indicator dashboard, probability updates, clear adaptation triggers and owner."
+      }
+    }
+  ],
+  "guidance_by_type": {
+    "Strategic Planning (1-3 year horizon)": {
+      "target_score": 4.2,
+      "key_criteria": ["Scenario Diversity", "Probability Calibration", "Action Extraction"],
+      "common_pitfalls": ["Too narrow scenario range", "No hedges against downside", "Monitoring plan missing"],
+      "specific_guidance": "Use three-scenario framework (optimistic/baseline/pessimistic) or 2×2 matrix. Assign probabilities (optimistic 15-30%, baseline 40-60%, pessimistic 15-30%). Extract common actions that work across all scenarios, plus hedges for downside. Quarterly monitoring."
+    },
+    "Pre-Mortem (Project Risk Identification)": {
+      "target_score": 4.0,
+      "key_criteria": ["Pre-Mortem Rigor", "Action Extraction", "Bias Avoidance"],
+      "common_pitfalls": ["Generic risks (not context-specific)", "Hindsight bias ('obvious' failures)", "No mitigations assigned"],
+      "specific_guidance": "Follow 6-step process rigorously. Silent brainstorm 5-10 min to prevent groupthink. Generate context-specific failure modes. Vote on top 5-7 risks. Assign mitigation owner and deadline for each."
+    },
+    "Counterfactual Learning (Post-Decision Analysis)": {
+      "target_score": 3.8,
+      "key_criteria": ["Minimal Rewrite Principle", "Causal Mechanism Specification", "Bias Avoidance"],
+      "common_pitfalls": ["Changing multiple factors (can't isolate cause)", "No causal mechanism (just outcome)", "Hindsight bias ('knew it all along')"],
+      "specific_guidance": "Change single factor, hold all else constant. Trace complete causal chain (change → immediate effect → secondary effects → outcome). Re-inhabit decision context to avoid hindsight. Use base rates and analogies to estimate counterfactual probability."
+    },
+    "Stress Testing (Decision Robustness)": {
+      "target_score": 4.0,
+      "key_criteria": ["Scenario Diversity", "Causal Mechanism Specification", "Action Extraction"],
+      "common_pitfalls": ["Only optimistic/pessimistic (no black swan)", "No mechanism for how extremes occur", "Decision not actually tested"],
+      "specific_guidance": "Test decision against optimistic, pessimistic, AND black swan scenarios. Specify HOW extreme outcomes occur. Ask 'Does decision still hold?' for each scenario. Extract hedges to protect against downside extremes."
+    },
+    "Assumption Reversal (Innovation/Pivots)": {
+      "target_score": 3.5,
+      "key_criteria": ["Scenario Plausibility", "Action Extraction", "Bias Avoidance"],
+      "common_pitfalls": ["Reversed assumptions implausible", "Interesting but no experiments", "Confirmation bias (only reverse convenient assumptions)"],
+      "specific_guidance": "Reverse core assumptions ('customers want more features' → 'want fewer'). Test plausibility (could reversal be true?). Design small experiments to test reversal. Challenge assumptions that support current strategy, not just peripheral ones."
+    }
+  },
+  "guidance_by_complexity": {
+    "Simple (Routine Decisions, Short-Term)": {
+      "target_score": 3.5,
+      "focus_areas": ["Scenario Plausibility", "Causal Mechanism Specification", "Action Extraction"],
+      "acceptable_shortcuts": ["Informal probabilities", "Two scenarios instead of three", "Simple pre-mortem (no voting)"],
+      "specific_guidance": "Quick pre-mortem (30 min) or simple counterfactual analysis. Two scenarios (optimistic/pessimistic). Extract 2-3 key actions. Informal monitoring acceptable."
+    },
+    "Standard (Strategic Decisions, 1-2 year horizon)": {
+      "target_score": 4.0,
+      "focus_areas": ["Probability Calibration", "Scenario Diversity", "Leading Indicator Quality", "Monitoring and Adaptation"],
+      "acceptable_shortcuts": ["Three scenarios (not full 2×2 matrix)", "Quarterly vs monthly monitoring"],
+      "specific_guidance": "Three-scenario framework with probabilities. Extract common actions, hedges, options. Define 5-7 leading indicators. Quarterly scenario reviews and updates. Assign owners for monitoring."
+    },
+    "Complex (High Stakes, Multi-Year, High Uncertainty)": {
+      "target_score": 4.5,
+      "focus_areas": ["All criteria", "Rigorous validation", "Comprehensive monitoring"],
+      "acceptable_shortcuts": ["None - full rigor required"],
+      "specific_guidance": "Full 2×2 scenario matrix or cone of uncertainty. Rigorous probability calibration using base rates and expert judgment. Comprehensive pre-mortem with cross-functional team. Leading indicators with clear thresholds and decision triggers. Monthly monitoring, quarterly deep reviews. All mitigations assigned with owners and deadlines."
+    }
+  },
+  "common_failure_modes": [
+    {
+      "name": "Implausible Counterfactuals (Fantasy)",
+      "symptom": "Counterfactuals require magic, impossible foreknowledge, or weren't real options at decision time. 'What if we had known pandemic was coming?'",
+      "detection": "Ask: 'Could a reasonable decision-maker have chosen this alternative given information available then?' If no, implausible.",
+      "fix": "Restrict to alternatives actually available at decision time. Use 'what was on the table?' test. Avoid hindsight-dependent counterfactuals."
+    },
+    {
+      "name": "Multiple Changes (Can't Isolate Cause)",
+      "symptom": "Counterfactual changes many factors: 'What if we had raised $3M AND launched EU AND hired different CEO...' Can't tell which mattered.",
+      "detection": "Count changes. If >1 factor changed, causal isolation violated.",
+      "fix": "Minimal rewrite: change ONE factor, hold all else constant. Want to test funding? Change funding only. Want to test geography? Change geography only."
+    },
+    {
+      "name": "No Causal Mechanism",
+      "symptom": "Outcome stated without explanation. 'Sales would be 2× higher' but no WHY or HOW.",
+      "detection": "Ask 'How does change lead to outcome?' If answer vague or missing, no mechanism.",
+      "fix": "Trace causal chain: initial change → immediate effect → secondary effects → final outcome. Each step must be explained with logic or evidence."
+    },
+    {
+      "name": "Scenarios Too Similar",
+      "symptom": "Three scenarios differ only in magnitude (10% growth vs 15% vs 20%). Same story, different numbers.",
+      "detection": "Read scenarios. Do they describe qualitatively different worlds? If no, too similar.",
+      "fix": "Make scenarios qualitatively distinct. Different drivers, different strategic implications. Use 2×2 matrix to force diversity via two independent uncertainties."
+    },
+    {
+      "name": "No Probabilities Assigned",
+      "symptom": "All scenarios treated as equally likely, or no probabilities given. Implies 33% each for three scenarios regardless of plausibility.",
+      "detection": "Check if probabilities assigned and justified. If missing or all equal, red flag.",
+      "fix": "Assign probabilities using base rates, analogies, expert judgment. Baseline typically 40-60%, optimistic/pessimistic 15-30% each. Justify each estimate."
+    },
+    {
+      "name": "Hindsight Bias in Counterfactuals",
+      "symptom": "'Obviously we should have done X' - outcome seems inevitable in retrospect. Overconfidence counterfactual would have succeeded.",
+      "detection": "Ask: 'Was outcome predictable given information at decision time?' If reasoning depends on information learned after, hindsight bias.",
+      "fix": "Re-inhabit decision context: what was known/unknown then? What uncertainty existed? Acknowledge alternative could have failed too. Use base rates to calibrate confidence."
+    },
+    {
+      "name": "Generic Pre-Mortem Risks",
+      "symptom": "Pre-mortem lists generic risks ('ran out of money', 'competition', 'tech didn't work') not specific to this project.",
+      "detection": "Could these risks apply to any project? If yes, too generic.",
+      "fix": "Push for context-specific failure modes. What's unique about THIS project? What specific technical challenges? Which specific competitors? What particular market risks?"
+    },
+    {
+      "name": "Scenarios Without Actions",
+      "symptom": "Interesting stories developed but no operational implications. 'So what should we do?' question unanswered.",
+      "detection": "Read scenario analysis. Is there action plan with common actions, hedges, options? If no, incomplete.",
+      "fix": "Always end with action extraction: (1) Common actions (all scenarios), (2) Hedges (downside protection), (3) Options (upside preparation), (4) Leading indicators (monitoring)."
+    },
+    {
+      "name": "Lagging Indicators (Not Leading)",
+      "symptom": "Indicators show scenario after it's happened. 'Revenue collapse' indicates pessimistic scenario, but too late to act.",
+      "detection": "Ask: 'Does this indicator give 6+ months advance notice?' If no, it's lagging.",
+      "fix": "Find early signals: regulatory votes (before law passed), competitor funding rounds (before product launched), adoption rate trends (before market share shift). Leading indicators are predictive, not reactive."
+    },
+    {
+      "name": "No Monitoring Plan",
+      "symptom": "Scenarios developed, actions defined, then filed away. No one tracking which scenario unfolding or updating probabilities.",
+      "detection": "Ask: 'Who monitors? How often? What triggers update?' If no answers, no plan.",
+      "fix": "Define: (1) Owner responsible for monitoring, (2) Cadence (monthly/quarterly reviews), (3) Indicator dashboard, (4) Decision triggers ('If X crosses threshold Y, then action Z'), (5) Scenario probability update process."
+    }
+  ],
+  "minimum_standard": 3.5,
+  "target_score": 4.0,
+  "excellence_threshold": 4.5
+}

skills/hypotheticals-counterfactuals/resources/methodology.md

@@ -0,0 +1,498 @@
+# Hypotheticals and Counterfactuals: Advanced Methodology
+
+This document provides advanced techniques for counterfactual reasoning, scenario planning, pre-mortem analysis, and extracting actionable insights from alternative futures.
+
+## Table of Contents
+1. [Counterfactual Reasoning](#1-counterfactual-reasoning)
+2. [Scenario Planning Techniques](#2-scenario-planning-techniques)
+3. [Extracting Insights from Scenarios](#3-extracting-insights-from-scenarios)
+4. [Monitoring and Adaptation](#4-monitoring-and-adaptation)
+5. [Advanced Topics](#5-advanced-topics)
+
+---
+
+## 1. Counterfactual Reasoning
+
+### What Is Counterfactual Reasoning?
+
+Counterfactual reasoning asks: **"What would have happened if X had been different?"** It's a form of causal inference through imagined alternatives.
+
+**Core principle**: To understand causality, imagine the world with one factor changed and trace the consequences.
+
+**Example**:
+- **Actual**: Startup raised $5M Series A → burned through runway in 14 months → failed to reach profitability
+- **Counterfactual**: "What if we had raised $3M instead?"
+  - Hypothesis: Smaller team (8 vs 15 people), lower burn, forced focus on revenue, reached profitability in 12 months
+  - Reasoning: Constraint forces discipline. Without $5M runway, couldn't afford large team. Would prioritize revenue over growth.
+  - Insight: Raising more money enabled premature scaling. Constraint would have been beneficial.
+
+### The Minimal Rewrite Principle
+
+When constructing counterfactuals, change **as little as possible** to isolate the causal factor.
+
+**Bad counterfactual** (too many changes):
+- "What if we had raised $3M AND competitor had failed AND market had doubled?"
+- Problem: Can't tell which factor caused different outcome
+
+**Good counterfactual** (minimal change):
+- "What if we had raised $3M (all else equal)?"
+- Isolates the funding amount as causal variable
+
+**Technique**: Hold everything constant except the factor you're testing. This reveals whether that specific factor was causal.
+
+### Constructing Plausible Counterfactuals
+
+Not all "what ifs" are useful. Counterfactuals must be **plausible** given what was known/possible at the time.
+
+**Plausible**: "What if we had launched in EU first instead of US?"
+- This was a real option available at decision time
+
+**Implausible**: "What if we had magically known the pandemic was coming?"
+- Requires impossible foreknowledge
+
+**Test for plausibility**: Could a reasonable decision-maker have chosen this alternative given information available at the time?
+
+### Specifying Causal Mechanisms
+
+Don't just state outcome; explain **HOW** the change leads to different result.
+
+**Weak counterfactual**: "Sales would be 2× higher"
+**Strong counterfactual**: "Sales would be 2× higher because lower price ($50 vs $100) → 3× higher conversion rate (15% vs 5%) → more customers despite 50% lower margin per customer → net revenue impact +2×"
+
+**Framework for causal chains**:
+1. **Initial change**: What's different? (e.g., "Price is $50 instead of $100")
+2. **Immediate effect**: What happens next? (e.g., "Conversion rate increases from 5% to 15%")
+3. **Secondary effects**: What follows? (e.g., "Customer volume triples")
+4. **Final outcome**: Net result? (e.g., "Revenue doubles despite lower margin")
+
+### Using Counterfactuals for Learning
+
+**Post-decision counterfactual analysis**:
+
+After a decision plays out, ask:
+1. **What did we decide?** (Actual decision)
+2. **What was the outcome?** (Actual result)
+3. **What else could we have done?** (Alternative decision)
+4. **What would have happened?** (Counterfactual outcome)
+5. **What was the key causal factor?** (Insight for future)
+
+**Example**: Hired candidate A (strong technical, weak communication) → struggled, left after 6 months. Counterfactual: B (moderate technical, strong communication) would have stayed longer, collaborated better. **Insight**: For this role, communication > pure technical skill.
+
+### Avoiding Hindsight Bias
+
+**Hindsight bias**: "I knew it all along" — outcome seems inevitable after the fact.
+
+**Problem**: Makes counterfactual analysis distorted. "Of course it failed, we should have known."
+
+**Mitigation**:
+- **Re-inhabit decision context**: What information was available then (not now)?
+- **List alternatives considered**: What options were on the table at the time?
+- **Acknowledge uncertainty**: How predictable was outcome given information available?
+
+**Technique**: Write counterfactual analysis in past tense but from perspective of decision-maker at the time, without benefit of hindsight.
+
+---
+
+## 2. Scenario Planning Techniques
+
+### Three-Scenario Framework
+
+**Structure**: Optimistic, Baseline, Pessimistic futures
+
+**When to use**: General strategic planning, forecasting, resource allocation
+
+**Process**:
+
+1. **Define time horizon**: 6 months? 1 year? 3 years? 5 years?
+   - Shorter horizons: More specific, quantitative
+   - Longer horizons: More qualitative, exploratory
+
+2. **Identify key uncertainties**: What 2-3 factors most shape the future?
+   - Market adoption rate
+   - Competitive intensity
+   - Regulatory environment
+   - Technology maturity
+   - Economic conditions
+
+3. **Develop three scenarios**:
+   - **Optimistic** (15-30% probability): Best-case assumptions on key uncertainties
+   - **Baseline** (40-60% probability): Expected-case, extrapolating current trends
+   - **Pessimistic** (15-30% probability): Worst-case assumptions on key uncertainties
+
+4. **Describe each vividly**: Write 2-4 paragraph narrative making each world feel real
+
+5. **Extract implications**: What should we do in each scenario?
+
+**Example (SaaS startup, 2-year horizon)**:
+
+**Key uncertainties**: (1) Market adoption rate, (2) Competition intensity
+
+**Optimistic scenario** (20% probability): "Market Leader"
+- Adoption: 40% market share, viral growth, $10M ARR
+- Competition: Weak, no major new entrants
+- Drivers: Product-market fit strong, word-of-mouth, early mover advantage
+- Implications: Invest heavily in scale infrastructure, expand to adjacent markets
+
+**Baseline scenario** (60% probability): "Steady Climb"
+- Adoption: 15% market share, steady growth, $3M ARR
+- Competition: Moderate, 2-3 well-funded competitors
+- Drivers: Expected adoption curve, competitive but differentiated
+- Implications: Focus on core product, maintain burn discipline, build moat
+
+**Pessimistic scenario** (20% probability): "Survival Mode"
+- Adoption: 5% market share, slow growth, $500k ARR
+- Competition: Intense, major player launches competing product
+- Drivers: Slow adoption, strong competition, pivot needed
+- Implications: Cut burn, extend runway, explore pivot or acquisition
+
+### 2×2 Scenario Matrix
+
+**Structure**: Two key uncertainties create four quadrants (scenarios)
+
+**When to use**: When two specific uncertainties dominate strategic decision
+
+**Process**:
+
+1. **Identify two critical uncertainties**: Factors that:
+   - Are genuinely uncertain (not predictable)
+   - Have major impact on outcomes
+   - Are independent (not correlated)
+
+2. **Define axes extremes**:
+   - Uncertainty 1: [Low extreme] ←→ [High extreme]
+   - Uncertainty 2: [Low extreme] ←→ [High extreme]
+
+3. **Name four quadrants**: Give each world a memorable name
+
+4. **Develop narratives**: Describe what each world looks like
+
+5. **Identify strategic implications**: What works in each quadrant?
+
+**Example (Market entry decision)**:
+
+**Uncertainty 1**: Regulatory environment (Strict ←→ Loose)
+**Uncertainty 2**: Market adoption rate (Slow ←→ Fast)
+
+|  | **Slow Adoption** | **Fast Adoption** |
+|---|---|---|
+| **Strict Regulation** | **"Constrained Growth"**: Premium focus, compliance differentiator | **"Regulated Scale"**: Invest in compliance infrastructure early |
+| **Loose Regulation** | **"Patient Build"**: Bootstrap, iterate slowly | **"Wild West Growth"**: Fast growth, grab market share |
+
+**Strategic insight**: Common actions (build product), hedges (low burn), options (compliance prep), monitoring (regulation + adoption)
+
+### Cone of Uncertainty
+
+**Structure**: Range of outcomes widens over time
+
+**When to use**: Long-term planning (5-10+ years), high uncertainty domains (technology, policy)
+
+**Visualization**:
+```
+Present → 1 year: Narrow cone (±20%)
+        → 3 years: Medium cone (±50%)
+        → 5 years: Wide cone (±100%)
+        → 10 years: Very wide cone (±200%+)
+```
+
+**Technique**:
+1. **Start with trend**: Current trajectory (e.g., "10% annual growth")
+2. **Add uncertainty bands**: Upper and lower bounds that widen over time
+3. **Identify branch points**: Key decisions or events that shift trajectory
+4. **Track leading indicators**: Signals that show which path we're on
+
+**Example (Revenue forecasting)**:
+- Year 1: $1M ± 20% ($800k - $1.2M) — narrow range, short-term visibility
+- Year 3: $3M ± 50% ($1.5M - $4.5M) — wider range, product-market fit uncertain
+- Year 5: $10M ± 100% ($5M - $20M) — very wide range, market evolution unknown
+
+### Pre-Mortem Process (Prospective Hindsight)
+
+**What is it?**: Imagine future failure, work backward to identify causes
+
+**Why it works**: "Prospective hindsight" — imagining outcome has occurred unlocks insights impossible from forward planning
+
+**Research foundation**: Gary Klein, "Performing a Project Premortem" (HBR 2007)
+
+**6-Step Process**:
+
+**Step 1: Set the scene**
+- Future date: "It is [6 months / 1 year / 2 years] from now..."
+- Assumed outcome: "...and the [project has failed completely / decision was disastrous]."
+- Make it vivid: "The product has been shut down. The team disbanded. We lost $X."
+
+**Step 2: Individual brainstorm (5-10 minutes, silent)**
+- Each person writes 3-5 reasons WHY it failed
+- Silent writing prevents groupthink
+- Encourage wild ideas, non-obvious causes
+
+**Step 3: Round-robin sharing**
+- Each person shares one reason (rotating until all shared)
+- No debate yet, just capture ideas
+- Scribe records all items
+
+**Step 4: Consolidate and cluster**
+- Group similar causes together
+- Look for themes (technical, market, team, execution, external)
+
+**Step 5: Vote on top risks**
+- Dot voting: Each person gets 3-5 votes
+- Distribute votes across risks
+- Identify top 5-7 risks by vote count
+
+**Step 6: Develop mitigations**
+- For each top risk, assign:
+  - **Mitigation action**: Specific step to prevent or reduce risk
+  - **Owner**: Who is responsible
+  - **Deadline**: When mitigation must be in place
+  - **Success metric**: How to know mitigation worked
+
+**Pre-mortem psychology**:
+- **Permission to dissent**: Failure assumption gives license to voice concerns without seeming negative
+- **Cognitive relief**: Easier to imagine specific failure than abstract "what could go wrong?"
+- **Team alignment**: Surfaces hidden concerns before they become real problems
+
+---
+
+## 3. Extracting Insights from Scenarios
+
+### Moving from Stories to Actions
+
+Scenarios are useless without actionable implications. After developing scenarios, ask:
+
+**Core questions**:
+1. **What should we do regardless of which scenario unfolds?** (Common actions)
+2. **What hedges should we take against downside scenarios?** (Risk mitigation)
+3. **What options should we create for upside scenarios?** (Opportunity capture)
+4. **What should we monitor to track which scenario is unfolding?** (Leading indicators)
+
+### Identifying Common Actions
+
+**Common actions** ("no-regrets moves"): Work across all scenarios
+
+**Technique**: List actions that make sense in optimistic, baseline, AND pessimistic scenarios
+
+**Example (Product launch scenarios)**:
+
+| Scenario | Build Core Product | Hire Marketing | Raise Series B |
+|----------|-------------------|----------------|----------------|
+| Optimistic | ✓ Essential | ✓ Essential | ✓ Essential |
+| Baseline | ✓ Essential | ✓ Essential | △ Maybe |
+| Pessimistic | ✓ Essential | △ Maybe | ✗ Too risky |
+
+**Common actions**: Build core product, hire marketing (work in all scenarios)
+**Not common**: Raise Series B (only makes sense in optimistic/baseline)
+
+### Designing Hedges
+
+**Hedges**: Actions that reduce downside risk if pessimistic scenario unfolds
+
+**Principle**: Pay small cost now to protect against large cost later
+
+**Examples**:
+- **Pessimistic scenario**: "Competitor launches free version, our revenue drops 50%"
+- **Hedge**: Keep burn low, maintain 18-month runway (vs. 12-month)
+  - Cost: Hire 2 fewer people now
+  - Benefit: Survive revenue shock if it happens
+
+- **Pessimistic scenario**: "Regulatory crackdown makes our business model illegal"
+- **Hedge**: Develop alternative revenue model in parallel
+  - Cost: 10% of eng time on alternative
+  - Benefit: Can pivot quickly if regulation hits
+
+**Hedge evaluation**: Compare cost of hedge vs. expected loss × probability
+- Hedge cost: $X
+- Loss if scenario occurs: $Y
+- Probability of scenario: P
+- Expected value of hedge: (P × $Y) - $X
+
+### Creating Options
+
+**Options**: Prepare to capture upside if optimistic scenario unfolds, without committing resources now
+
+**Real options theory**: Create flexibility to make future decisions when more information available
+
+**Examples**:
+- **Optimistic scenario**: "Adoption faster than expected, enterprise demand emerges"
+- **Option**: Design product architecture with enterprise features in mind (multi-tenancy, SSO hooks), but don't build until demand confirmed
+  - Low cost now: Design decisions
+  - High value later: Fast enterprise launch if demand materializes
+
+- **Optimistic scenario**: "International markets grow 3× faster than expected"
+- **Option**: Hire one person with international experience, build relationships with international partners
+  - Low cost now: One hire
+  - High value later: Quick international expansion if opportunity emerges
+
+### Defining Leading Indicators
+
+**Leading indicators**: Early signals that show which scenario is unfolding
+
+**Characteristics of good leading indicators**:
+- **Observable**: Can be measured objectively
+- **Early**: Visible before scenario fully plays out (6+ months advance notice)
+- **Actionable**: If indicator triggers, we know what to do
+
+**Example (Market adoption scenarios)**:
+
+| Scenario | Leading Indicator | Threshold | Action if Triggered |
+|----------|------------------|-----------|---------------------|
+| Optimistic | Monthly adoption rate | >20% MoM for 3 months | Accelerate hiring, raise capital |
+| Baseline | Monthly adoption rate | 10-20% MoM | Maintain plan |
+| Pessimistic | Monthly adoption rate | <10% MoM for 3 months | Cut burn, explore pivot |
+
+**Monitoring cadence**: Review indicators monthly or quarterly, update scenario probabilities based on new data
+
+### Decision Points and Trigger Actions
+
+**Decision points**: Pre-defined thresholds that trigger specific actions
+
+**Format**: "If [indicator] crosses [threshold], then [action]"
+
+**Examples**:
+- "If monthly churn rate >8% for 2 consecutive months, then launch retention task force"
+- "If competitor raises >$50M, then accelerate roadmap and increase marketing spend"
+- "If regulation bill passes committee vote, then begin compliance implementation immediately"
+
+**Benefits**:
+- **Pre-commitment**: Decide now what to do later, avoids decision paralysis in moment
+- **Speed**: Trigger action immediately when condition met
+- **Alignment**: Team knows what to expect, can prepare
+
+---
+
+## 4. Monitoring and Adaptation
+
+### Tracking Which Scenario Is Unfolding
+
+**Reality ≠ any single scenario**: Real world is blend of scenarios, or something unexpected
+
+**Monitoring approach**:
+1. **Quarterly scenario review**: Update probabilities based on new evidence
+2. **Indicator dashboard**: Track 5-10 leading indicators, visualize trends
+3. **Surprise tracking**: Log unexpected events not captured by scenarios
+
+**Example dashboard**:
+
+| Indicator | Optimistic | Baseline | Pessimistic | Current | Trend |
+|-----------|-----------|----------|-------------|---------|-------|
+| Adoption rate | >20% MoM | 10-20% | <10% | 15% | ↑ |
+| Churn rate | <3%/mo | 3-5% | >5% | 4% | → |
+| Competitor funding | <$20M | $20-50M | >$50M | $30M | ↑ |
+| NPS | >50 | 30-50 | <30 | 45 | ↑ |
+
+**Interpretation**: Trending optimistic (adoption, NPS), watch competitor funding
+
+### Updating Scenarios
+
+**When to update**:
+- **Scheduled**: Quarterly reviews
+- **Triggered**: Major unexpected event (pandemic, regulation, acquisition, etc.)
+
+**Update process**:
+1. **Review what happened**: What changed since last review?
+2. **Update probabilities**: Which scenario looking more/less likely?
+3. **Revise scenarios**: Do scenarios still capture range of plausible futures? Add new ones if needed
+4. **Adjust actions**: Change hedges, options, or common actions based on new information
+
+**Example**: Before pandemic: Optimistic 20%, Baseline 60%, Pessimistic 20%. After: Add "Remote-first world" (30%), reduce Baseline to 40%. Action: shift from office expansion to remote tooling.
+
+### Dealing with Surprises
+
+**Black swans**: High-impact, low-probability events not captured by scenarios (Taleb)
+
+**Response protocol**:
+1. **Acknowledge**: "This is outside our scenarios"
+2. **Assess**: How does this change the landscape?
+3. **Create emergency scenario**: Rapid scenario development (hours/days, not weeks)
+4. **Decide**: What immediate actions needed?
+5. **Update scenarios**: Incorporate new uncertainty into ongoing planning
+
+**Example**: COVID-19 lockdowns (not in scenarios) → Assess: dining impossible → Emergency scenario: "Delivery-only world" (6-12 mo) → Actions: pivot to takeout, renegotiate leases → Update: add "Hybrid dining" scenario
+
+### Scenario Planning as Organizational Learning
+
+**Scenarios as shared language**: Team uses scenario names to communicate quickly
+- "We're in Constrained Growth mode" → Everyone knows what that means
+
+**Scenario-based planning**: Budgets, roadmaps reference scenarios
+- "If we hit Optimistic scenario by Q3, we trigger hiring plan B"
+
+**Cultural benefit**: Reduces certainty bias, maintains flexibility, normalizes uncertainty
+
+---
+
+## 5. Advanced Topics
+
+### Counterfactual Probability Estimation
+
+**Challenge**: How likely was counterfactual outcome?
+
+**Approach**: Use base rates and analogies
+1. **Find analogous cases**: What happened in similar situations?
+2. **Calculate base rate**: Of N analogous cases, in how many did X occur?
+3. **Adjust for specifics**: Is our case different? How?
+4. **Estimate probability range**: Not point estimate, but range (40-60%)
+
+**Example**: "What if we had launched in EU first?" — 20 similar startups: 8/20 chose EU-first (3/8 succeeded = 37.5%), 12/20 chose US-first (7/12 = 58%). Our product has EU features (+10%) → EU-first 35-50%, US-first 50-65%. **Insight**: US-first was better bet.
+
+### Scenario Narrative Techniques
+
+**Make scenarios memorable and vivid**:
+
+**Technique 1: Use present tense**
+- Bad: "Adoption will grow quickly"
+- Good: "It's January 2026. Our user base has grown 10× in 12 months..."
+
+**Technique 2: Add concrete details**
+- Bad: "Competition is intense"
+- Good: "Three well-funded competitors (FundedCo with $50M Series B, StartupX acquired by BigTech, OpenSource Project with 10k stars) are fighting for same customers..."
+
+**Technique 3: Use personas/characters**
+- "Sarah, our typical customer (marketing manager at 50-person B2B SaaS company), now has five alternatives to our product..."
+
+**Technique 4: Include metrics**
+- "Monthly churn rate: 8%, NPS: 25, CAC: $500 (up from $200)"
+
+### Assumption Reversal for Innovation
+
+**Technique**: Take core assumption, flip it, explore implications
+
+**Process**:
+1. **List key assumptions**: What do we take for granted?
+2. **Reverse each**: "What if opposite is true?"
+3. **Explore plausibility**: Could reversal be true?
+4. **Identify implications**: What would we do differently?
+5. **Test**: Can we experiment with reversal?
+
+**Examples**:
+
+| Current Assumption | Reversed Assumption | Implications | Action |
+|-------------------|---------------------|--------------|--------|
+| "Customers want more features" | "Customers want FEWER features" | Simplify product, remove rarely-used features, focus on core workflow | Survey: Would users pay same for product with 50% fewer features but better UX? |
+| "Freemium is best model" | "Paid-only from day 1" | No free tier, premium positioning, higher LTV but lower top-of-funnel | Test: Launch premium SKU, measure willingness to pay |
+| "We need to raise VC funding" | "Bootstrap and self-fund" | Slower growth, but control + profitability focus | Calculate: Can we reach profitability on current runway? |
+
+### Timeboxing Scenario Work
+
+**Problem**: Scenario planning can become endless theorizing
+
+**Solution**: Timebox exercises
+
+**Suggested time budgets**: Pre-mortem (60-90 min), Three scenarios (2-4 hrs), 2×2 matrix (3-5 hrs), Quarterly review (1-2 hrs)
+
+**Principle**: Scenarios are decision tools, not academic exercises. Generate enough insight to decide, then act.
+
+---
+
+## Summary
+
+**Counterfactual reasoning** reveals causality through minimal-change thought experiments. Focus on plausibility, specify mechanisms, avoid hindsight bias.
+
+**Scenario planning** (three scenarios, 2×2 matrix, cone of uncertainty) explores alternative futures. Assign probabilities, make vivid, extract actions.
+
+**Extract insights** by identifying common actions (no-regrets moves), hedges (downside protection), options (upside preparation), and leading indicators (early signals).
+
+**Monitor and adapt** quarterly. Track indicators, update scenario probabilities, adjust strategy as reality unfolds. Treat surprises as learning opportunities.
+
+**Advanced techniques** include counterfactual probability estimation, narrative crafting, assumption reversal, and rigorous timeboxing to avoid analysis paralysis.
+
+**The goal**: Prepare for uncertainty, maintain strategic flexibility, and make better decisions by systematically exploring "what if?"

skills/hypotheticals-counterfactuals/resources/template.md

@@ -0,0 +1,304 @@
+# Hypotheticals and Counterfactuals Templates
+
+Quick-start templates for counterfactual analysis, scenario planning, and pre-mortem exercises.
+
+## Focal Question Template
+
+**What are you exploring?**
+
+**Type**: [Counterfactual (past) / Hypothetical (future)]
+
+**Core question**:
+- Counterfactual: "What would have happened if [X] had been different?"
+- Hypothetical: "What could happen if [X] occurs in the future?"
+
+**Context**: [What decision, event, or situation are you analyzing?]
+
+**Time frame**: [Past event date / Future time horizon (6 months, 1 year, 5 years)]
+
+**Purpose**: [What do you hope to learn? Understand causality? Identify risks? Test assumptions?]
+
+---
+
+## Counterfactual Analysis Template
+
+**Actual outcome** (what happened):
+- Decision made: [What did we actually do?]
+- Outcome: [What resulted?]
+- Key metrics: [Quantify results]
+
+**Counterfactual** (what if we had done differently):
+- Alternative decision: "What if we had [done X instead]?"
+- Hypothesized outcome: [What would have happened?]
+- Reasoning: [WHY would outcome be different? Specify causal mechanism]
+
+**Evidence for counterfactual**:
+- Analogies: [Similar cases where X led to Y]
+- Data: [Market data, competitor examples, historical patterns]
+- Expert opinion: [What do domain experts say?]
+
+**Causal insight**:
+- What mattered: [Which factor was causal?]
+- What didn't matter: [Which factors were irrelevant?]
+- Lesson learned: [What should we do differently next time?]
+
+**Example**:
+- **Actual**: Launched in US first, 10k users in 6 months
+- **Counterfactual**: "What if we had launched in EU first?"
+- **Hypothesized outcome**: 5k users (smaller market, slower adoption)
+- **Reasoning**: EU market 40% size of US, GDPR compliance slows growth
+- **Insight**: US-first was right call. Market size matters more than competition.
+
+---
+
+## Pre-Mortem Template
+
+**Project/Decision**: [What are you launching or deciding?]
+
+**Future date**: "It is [6 months / 1 year] from now..."
+
+**Assumed outcome**: "...and the [project has failed / decision was disastrous]."
+
+**Individual brainstorm** (5 min, silent):
+Each person writes 3-5 reasons why it failed.
+
+1. [Failure reason 1]
+2. [Failure reason 2]
+3. [Failure reason 3]
+4. [Failure reason 4]
+5. [Failure reason 5]
+
+**Consolidate** (round-robin sharing):
+- [Consolidated failure cause 1]
+- [Consolidated failure cause 2]
+- [Consolidated failure cause 3]
+- [Consolidated failure cause 4]
+- [Consolidated failure cause 5]
+...
+
+**Vote on top risks** (dot voting):
+
+| Risk | Votes | Likelihood | Impact | Priority |
+|------|-------|------------|--------|----------|
+| [Risk 1] | 8 | High | High | ⚠ Critical |
+| [Risk 2] | 6 | Medium | High | ⚠ High |
+| [Risk 3] | 4 | High | Medium | Medium |
+| [Risk 4] | 2 | Low | Low | Low |
+
+**Mitigation actions** (top 3-5 risks):
+
+| Risk | Mitigation | Owner | Deadline |
+|------|------------|-------|----------|
+| [Risk 1] | [Specific action to prevent/reduce] | [Name] | [Date] |
+| [Risk 2] | [Specific action] | [Name] | [Date] |
+| [Risk 3] | [Specific action] | [Name] | [Date] |
+
+---
+
+## Scenario Generation Template
+
+**Time horizon**: [6 months / 1 year / 3 years / 5 years]
+
+**Key uncertainties** (2-3 factors that most shape the future):
+1. [Uncertainty 1, e.g., "Market adoption rate"]
+2. [Uncertainty 2, e.g., "Competitive intensity"]
+3. [Uncertainty 3, e.g., "Regulatory environment"]
+
+### Option A: Three Scenarios
+
+**Optimistic scenario** (Probability: [%]):
+- Name: "[Descriptive name]"
+- Description: [1-2 paragraphs describing this future]
+- Key drivers: [What makes this happen?]
+- Implications: [What does this mean for us?]
+
+**Baseline scenario** (Probability: [%]):
+- Name: "[Descriptive name]"
+- Description: [1-2 paragraphs]
+- Key drivers: [What makes this happen?]
+- Implications: [What does this mean for us?]
+
+**Pessimistic scenario** (Probability: [%]):
+- Name: "[Descriptive name]"
+- Description: [1-2 paragraphs]
+- Key drivers: [What makes this happen?]
+- Implications: [What does this mean for us?]
+
+### Option B: 2×2 Matrix
+
+**Uncertainty 1**: [e.g., Market adoption] - Axes: [Slow / Fast]
+**Uncertainty 2**: [e.g., Regulation] - Axes: [Strict / Loose]
+
+|  | **Slow Adoption** | **Fast Adoption** |
+|---|---|---|
+| **Strict Regulation** | **Scenario 1**: "[Name]"<br>[Description] | **Scenario 2**: "[Name]"<br>[Description] |
+| **Loose Regulation** | **Scenario 3**: "[Name]"<br>[Description] | **Scenario 4**: "[Name]"<br>[Description] |
+
+---
+
+## Scenario Development Template
+
+**Scenario name**: "[Memorable title]"
+
+**Time**: [Future date, e.g., "January 2026"]
+
+**Narrative** (tell the story, make it vivid):
+[2-4 paragraphs describing this world. Use present tense, concrete details, make it feel real.]
+
+**Key assumptions**:
+- [Assumption 1: what had to be true for this scenario?]
+- [Assumption 2]
+- [Assumption 3]
+
+**Metrics in this world**:
+- [Metric 1]: [Value, e.g., "Market size: $500M"]
+- [Metric 2]: [Value, e.g., "Our market share: 15%"]
+- [Metric 3]: [Value, e.g., "Churn rate: 3%/month"]
+
+**Leading indicators** (early signals this scenario is unfolding):
+- [Indicator 1]: [e.g., "If regulation bill passes Q1"]
+- [Indicator 2]: [e.g., "If competitor raises >$50M"]
+- [Indicator 3]: [e.g., "If adoption rate >20% MoM for 3 months"]
+
+**Implications for our strategy**:
+- What should we do in this world? [Strategic response]
+- What should we avoid? [Actions that fail in this scenario]
+- What capabilities do we need? [Org/tech requirements]
+
+---
+
+## Assumption Reversal Template
+
+**Current assumption**: [State the belief we take for granted]
+
+**Reversed assumption**: "What if [opposite] is true?"
+
+**Explore the reversal**:
+- Is it plausible? [Could the reversal actually be true?]
+- Evidence for reversal: [What would suggest our assumption is wrong?]
+- Implications if reversed: [What would we do differently?]
+- New possibilities: [What doors does this open?]
+
+**Example**:
+- **Current**: "Customers want more features"
+- **Reversed**: "What if customers want fewer features?"
+- **Plausible?**: Yes (research shows feature bloat frustrates users)
+- **Implications**: Simplify product, remove rarely-used features, focus on core workflow
+- **New possibility**: "Feature-light" positioning vs. competitors
+
+---
+
+## Stress Test Template
+
+**Decision being tested**: [What are we deciding?]
+
+**Baseline assumptions**:
+- [Assumption 1]: [Current expectation, e.g., "CAC = $100"]
+- [Assumption 2]: [e.g., "Churn = 5%/month"]
+- [Assumption 3]: [e.g., "Market size = $1B"]
+
+**Stress scenario 1: Optimistic**
+- [Assumption 1]: [Best case, e.g., "CAC = $50"]
+- [Assumption 2]: [e.g., "Churn = 2%/month"]
+- [Assumption 3]: [e.g., "Market size = $2B"]
+- **Decision still valid?**: [Yes/No, with explanation]
+
+**Stress scenario 2: Pessimistic**
+- [Assumption 1]: [Worst case, e.g., "CAC = $200"]
+- [Assumption 2]: [e.g., "Churn = 10%/month"]
+- [Assumption 3]: [e.g., "Market size = $500M"]
+- **Decision still valid?**: [Yes/No, with explanation]
+
+**Stress scenario 3: Black swan**
+- [Extreme event]: [e.g., "Major competitor offers product free"]
+- **Decision still valid?**: [Yes/No, with explanation]
+
+**Conclusion**:
+- Decision robust? [Does it hold across scenarios?]
+- Hedges needed? [What can we do to protect downside?]
+- Go/no-go? [Final decision]
+
+---
+
+## Action Extraction Template
+
+**Scenarios analyzed**: [List 2-4 scenarios explored]
+
+**Common actions** (work across all scenarios):
+- [Action 1]: [What should we do regardless of which future unfolds?]
+- [Action 2]
+- [Action 3]
+
+**Hedges** (protect against downside scenarios):
+- [Hedge 1]: [What reduces risk if pessimistic scenario happens?]
+- [Hedge 2]
+
+**Options** (prepare for upside scenarios):
+- [Option 1]: [What positions us to capture value if optimistic scenario happens?]
+- [Option 2]
+
+**Monitoring** (track which scenario unfolding):
+- [Indicator 1]: [What to watch, e.g., "Track regulation votes monthly"]
+- [Indicator 2]: [e.g., "Monitor competitor funding rounds"]
+- [Indicator 3]: [e.g., "Measure adoption rate vs. baseline"]
+
+**Decision points** (when to adjust):
+- If [indicator crosses threshold], then [action]
+- If [indicator crosses threshold], then [action]
+
+**Example**:
+- **Common**: Build core product, hire team, launch beta
+- **Hedge**: Keep burn low, maintain 18-month runway for slow-growth scenario
+- **Option**: Prepare enterprise sales motion if early adoption strong
+- **Monitor**: Track adoption rate monthly; if >15% MoM for 3 months, trigger enterprise hiring
+
+---
+
+## Quick Examples
+
+### Example 1: Product Launch Pre-Mortem
+
+**Project**: Launch new mobile app, target 50k downloads in 6 months
+
+**Pre-mortem** (failure causes):
+1. App crashes on Android (not tested thoroughly)
+2. Marketing budget too small (couldn't acquire users at scale)
+3. Onboarding too complex (80% drop-off after signup)
+4. Competitor launched free version (undercut pricing)
+5. App Store rejection (didn't follow guidelines)
+
+**Mitigation**:
+- Comprehensive Android testing before launch
+- Double marketing budget or lower target
+- Simplify onboarding to 3 steps max
+- Monitor competitor activity, prepare pricing flex
+- Review App Store guidelines, get pre-approval
+
+### Example 2: Counterfactual Learning
+
+**Actual**: Raised $5M Series A, 18-month runway, hired 15 people
+
+**Outcome**: Burned through runway in 14 months, failed to reach next milestone
+
+**Counterfactual**: "What if we had raised $3M instead?"
+- **Hypothesized outcome**: 12-month runway, hired 8 people, reached profitability
+- **Reasoning**: Smaller team = lower burn, forced focus on revenue, faster decisions
+- **Insight**: Raising more money led to premature scaling. Constraint is good early-stage.
+
+### Example 3: Strategic Scenarios (3 Futures)
+
+**Time**: 2026 (2 years out)
+
+**Optimistic ("Market Leader")**:
+- 40% market share, $10M ARR, profitability
+- Drivers: Product-market fit strong, viral growth, weak competition
+
+**Baseline ("Steady Climb")**:
+- 15% market share, $3M ARR, break-even
+- Drivers: Expected growth, moderate competition, steady execution
+
+**Pessimistic ("Survival Mode")**:
+- 5% market share, $500k ARR, burning cash
+- Drivers: Strong competitor launches, slow adoption, pivot needed
+
+**Implications**: Build for "Steady Climb", hedge for "Survival" (low burn), prepare for "Leader" (scale infrastructure).