# Predictable Irrationality

> Human irrationality is not random but systematic and predictable, with people consistently deviating from rational decision-making in specific patterns

- Skill: `lev-os/predictable-irrationality` (Agent Skill)
- Install (CLI): `npx skillmds@latest add lev-os/predictable-irrationality`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lev-os/predictable-irrationality/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: lev-os (https://skillmd.com/u/lev-os)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/lev-os/predictable-irrationality

---


# Predictable Irrationality

## Classification
**Domain:** Cognitive Biases & Behavioral Economics
**Category:** Systematic Behavioral Patterns
**Complexity:** Medium
**Abstraction Level:** Meta-Framework

## Core Principle
A behavioral economics framework by Dan Ariely showing that human irrationality is not random but systematic and predictable. Rather than making occasional mistakes, people consistently deviate from rational decision-making in specific, repeatable patterns driven by hidden psychological forces (relativity, social norms, expectations, arousal states, ownership, etc.). Understanding these predictable patterns allows you to anticipate irrational behavior in yourself and others, design better systems, and avoid systematic mistakes.

## When to Use
- **Product design** → Predict how users will respond to pricing, features, defaults
- **Marketing strategy** → Leverage anchoring, social proof, relativity effects
- **Negotiation** → Understand how reference points, ownership, and fairness norms shape offers
- **Personal decision-making** → Recognize when you're falling into predictable traps
- **Policy design** → Anticipate systematic deviations from rational response
- **Organizational behavior** → Design incentives accounting for motivation crowding, fairness concerns

## When to Avoid
- **Purely rational contexts** → Engineering specs, mathematical proofs, logical systems
- **Unique one-off decisions** → When past patterns don't predict novel situations
- **Manipulative intent** → Using predictable irrationality to exploit rather than inform
- **Perfectly informed experts** → Rare cases where domain mastery overcomes biases

## Execution Steps

### 1. Identify the Decision Context
Map the environment where decisions are made:
- **Who** decides? (Individual, group, expert, novice)
- **What** information is available? (Complete, partial, overwhelming)
- **When** is decision made? (Hot/cold state, time pressure, recurring)
- **Why** does it matter? (High stakes, low stakes, symbolic value)

### 2. Scan for Systematic Bias Patterns
Look for Ariely's documented patterns:
- **Relativity:** Decisions influenced by comparisons, not absolutes
- **Anchoring:** First number shapes all subsequent valuations
- **Social norms vs. market norms:** Money changes motivation
- **Arousal effects:** Hot states predict different choices than cold states
- **Ownership/endowment:** Possessing increases valuation
- **Expectations:** Beliefs shape experience (placebo, branding)

### 3. Predict the Irrational Behavior
Use known patterns to forecast decisions:
- People will overpay for relative gain vs. competitor
- First price seen will anchor all valuations
- Introducing money to social exchange will reduce effort
- People in arousal states (hunger, anger, lust) will make different choices
- Owners will value items ~2x higher than non-owners
- Expectations will alter actual experience

### 4. Design to Mitigate or Leverage
Two approaches:

**Mitigate (help people):**
- Remove misleading anchors
- Separate social and market norms
- Encourage cold-state decisions for hot-state situations
- Make ownership/attachment visible

**Leverage (ethically):**
- Set helpful anchors (high-quality reference point)
- Use social norms for prosocial behavior
- Create commitment devices for future self
- Use expectations to improve actual experience (placebo effect for good)

### 5. Test Empirically
Ariely's work is empirical. Test predictions:
- A/B test control vs. bias-aware design
- Measure actual behavior, not stated intentions
- Look for systematic patterns across cohorts
- Iterate based on results

### 6. Apply Consistently
Predictable irrationality means patterns repeat. Build systems that account for these biases by default, not as one-off fixes.

## Key Insights
- **Systematic, not random** → Same biases, same patterns, predictably wrong
- **Hidden forces** → People unaware of what's influencing them (context, anchors, arousal)
- **Relativity dominates** → We decide by comparison, not absolute value
- **Context shapes choice** → Change environment, change decisions (no stable preferences)
- **Dual selves problem** → Current self makes promises, future self breaks them
- **Money changes motivation** → Social norms → market norms = less effort, more selfishness
- **Expectations create reality** → Beliefs about wine quality literally change taste experience

## Common Pitfalls
- **Assuming awareness prevents bias** → Knowing about anchoring doesn't stop anchoring effects
- **One-time fixes** → Treating predictable patterns as occasional mistakes
- **Pure information** → Believing education overcomes systematic bias
- **Ignoring context** → Focusing on individual rationality, missing environmental influence
- **Manipulation** → Using predictable irrationality unethically to exploit
- **Rational actor models** → Building systems assuming people behave logically

## Practical Examples

### Scenario 1: The Economist Subscription Pricing
**Context:** Dan Ariely's famous experiment with The Economist pricing tiers

**Application:**
1. **Option A:** Web subscription $59
2. **Option B:** Print subscription $125
3. **Option C:** Print + Web subscription $125 (same as B)

**Behavior:**
- Without Option B (decoy): 68% choose Web ($59), 32% choose Print+Web ($125)
- With Option B (decoy): 16% choose Web, 0% choose Print-only, 84% choose Print+Web

**Result:** Decoy option (B) made option C look like incredible deal by comparison (relativity effect)

**Key Takeaway:** People don't know what they want absolutely; they decide by comparison

### Scenario 2: FREE! Shipping vs. $0.10 Shipping
**Context:** Amazon testing shipping fee impact

**Application:**
- **Condition A:** Free shipping on orders >$25 = 60% conversion
- **Condition B:** $0.10 shipping on orders >$25 = 38% conversion
- **Rational difference:** 10 cents ($0.10)
- **Actual impact:** 22 percentage points (37% relative decrease)

**Result:** "FREE" has disproportionate psychological power beyond economic value

**Key Takeaway:** Zero is not just another number; it's a special category (predictably irrational)

### Scenario 3: Honesty and Cash Box Experiment
**Context:** Office coffee honor system - take coffee, leave money in box

**Application:**
1. **Control:** Sign says "Please pay $1 for coffee" = 40% compliance
2. **Treatment:** Add image of watching eyes above sign = 76% compliance
3. **Mechanism:** Social norms activated by cue of observation (even though no one actually watching)

**Result:** Subtle environmental cue (eyes) doubled payment rate

**Key Takeaway:** Hidden forces (social norms, observation cues) shape behavior more than explicit rules

## Related Concepts
- **Prospect Theory** → Theoretical foundation for loss aversion, reference dependence
- **Nudge Theory** → Applied framework using predictable patterns to design choice architecture
- **Anchoring** → First number/value sets reference point for all subsequent judgments
- **Endowment Effect** → Ownership increases valuation (predictable irrationality pattern)
- **System 1/System 2 Thinking** → Automatic (System 1) processes drive predictable biases
- **Mental Accounting** → Money treated differently based on source/category (irrational but predictable)

## Prerequisites
- Basic understanding of behavioral economics
- Familiarity with cognitive biases
- Awareness that stated preferences ≠ revealed preferences
- Willingness to test empirically vs. assume rationality

## Learning Path
1. Start with **Anchoring** and **Framing Effects** to see how context shapes decisions
2. Progress to **Relativity** and **Decoy Effect** to understand comparison-based choice
3. Study **Predictable Irrationality** as meta-framework integrating patterns
4. Apply **Nudge Theory** to design systems accounting for systematic biases
5. Read *Predictably Irrational* by Dan Ariely for comprehensive examples

## Field Expertise
- **Dan Ariely** → Duke professor, author of *Predictably Irrational*, behavioral economist
- **George Loewenstein** → Carnegie Mellon, hot-cold empathy gap, intertemporal choice
- **Richard Thaler** → Nobel laureate, mental accounting, endowment effect
- **Daniel Kahneman** → Nobel laureate, prospect theory, heuristics & biases

## Tags
#predictable-irrationality #behavioral-economics #dan-ariely #systematic-biases #hidden-forces #decision-making #relativity #anchoring #social-norms #expectations

## Visual Cues
```
PREDICTABLE IRRATIONALITY PATTERNS:

┌──────────────────────────────────────┐
│  SYSTEMATIC BIAS CATEGORIES          │
├──────────────────────────────────────┤
│ 1. RELATIVITY ────► Compare, not value│
│ 2. ANCHORING ─────► First # sticks   │
│ 3. SOCIAL NORMS ──► $ crowds out     │
│ 4. AROUSAL STATES ► Hot ≠ Cold       │
│ 5. OWNERSHIP ─────► Endowment        │
│ 6. EXPECTATIONS ──► Belief → Reality │
└──────────────────────────────────────┘

Rational Model ───────X───► Fails to predict
Predictable Irrationality ─► Accurate forecasts
```

## Validation Checklist
- [ ] Identified decision context (who, what, when, why)
- [ ] Scanned for systematic bias patterns (relativity, anchoring, etc.)
- [ ] Predicted specific irrational behavior based on patterns
- [ ] Designed mitigation or leverage strategy
- [ ] Tested empirically with A/B test or controlled experiment
- [ ] Measured actual behavior, not stated intentions
- [ ] Applied learnings systematically, not one-off
- [ ] Verified ethical use (inform, not exploit)

## Success Metrics
- **Relativity effects:** 30-80% choice shift with decoy option
- **Anchoring impact:** 20-50% valuation change based on initial anchor
- **FREE effect:** 15-40% higher conversion with $0 vs. $0.01 cost
- **Social norms:** 20-60% behavior change with observation cues
- **Ownership premium:** 2-3x higher valuation for owned items

## Anti-Patterns
- **Rationality assumption** → Building systems expecting logical behavior
- **Information-only solutions** → Assuming education overcomes systematic bias
- **Ignoring context** → Focusing on individual choice, missing environmental drivers
- **One-off bias fixes** → Treating predictable patterns as occasional errors
- **Exploitation** → Using systematic biases to manipulate rather than help
- **Awareness fallacy** → Believing knowing about bias prevents it (it doesn't)