playbook/outfitter-agents/plugins/outfitter/skills/find-root-causes/references/pitfalls.md

5.2 KiB

Common Pitfalls

Cognitive biases and resistance patterns that derail root cause investigation.

Resistance Patterns

Rationalizations that prevent finding root cause:

Thought Why It's Wrong Counter
"I already looked at that" Memory is unreliable under pressure Re-examine with fresh evidence
"That can't be the issue" Assumptions block investigation Test anyway, let evidence decide
"We need to fix this quickly" Pressure leads to random changes Methodical investigation is faster
"The logs don't show anything" Absence of evidence != evidence of absence Consider what logs might be missing
"It worked before" Systems change constantly Past behavior doesn't guarantee current
"Let me just try this one thing" Random trial without hypothesis wastes time Form hypothesis first

Warning Signs

You're falling into resistance when:

  • Same thoughts recurring without new evidence
  • Feeling defensive about previous conclusions
  • Avoiding re-testing areas you "already checked"
  • Making changes without understanding why they might work

Recovery

When you catch yourself:

  1. Pause the investigation
  2. Write down current assumptions
  3. Challenge each assumption with "how do I know this?"
  4. Return to methodology

Confirmation Bias

Tendency to see evidence supporting existing beliefs.

Manifestations

  • Seeing only evidence supporting pet hypothesis
  • Dismissing contradictory data as "noise"
  • Stopping investigation once "a" cause is found
  • Interpreting ambiguous evidence favorably

Counter-Strategies

Actively seek disconfirmation:

  • Ask "what would prove me wrong?"
  • Design tests specifically to disprove hypothesis
  • Have someone else review your reasoning

Test alternative hypotheses:

  • Even when confident in one theory
  • Especially when confident in one theory
  • Give each hypothesis fair testing time

Document objectively:

  • Record all evidence, not just supporting evidence
  • Note your confidence level before and after tests
  • Track hypothesis changes over time

Correlation vs Causation

Mistaking timing for cause.

Common Mistakes

Observation Faulty Conclusion
"It started when X changed" X caused it
"Happens at specific time" Time is the cause
"Only affects user Y" User Y is doing something wrong
"Works after restart" Memory/state is the issue

Verification Steps

  1. Test direct causal mechanism — Can you explain HOW X causes the symptom?
  2. Look for confounding variables — What else changed or varies?
  3. Verify by removing supposed cause — Does removing X fix it?
  4. Test in isolation — Does X cause it when nothing else varies?

Example

Observation: "Bug only appears on Mondays"

Bad conclusion: "Something about Monday causes the bug"

Better investigation:

  • What's different on Monday? (traffic patterns, batch jobs, fresh caches)
  • Is it Monday specifically or "first day after weekend"?
  • Does it happen on holidays?
  • What runs over the weekend?

Anchoring

Over-reliance on first piece of information.

Manifestations

  • First hypothesis dominates thinking
  • Initial symptom description defines investigation
  • Early evidence weighted more heavily
  • Difficulty abandoning initial direction

Counter-Strategies

  • Explicitly generate 3+ hypotheses before testing any
  • Weight evidence by quality, not order discovered
  • Periodically re-read original problem statement
  • Ask "what if my first assumption is wrong?"

Availability Heuristic

Over-weighting recent or memorable experiences.

Manifestations

  • "This looks like the bug we had last week"
  • Assuming familiar problems over unfamiliar ones
  • Checking usual suspects first (sometimes good, often biasing)

Counter-Strategies

  • Consider base rates (how often does this actually happen?)
  • Check if "familiar" actually matches evidence
  • Maintain systematic approach even when "obvious"

Premature Closure

Stopping investigation too early.

Warning Signs

  • Relief when finding "a" cause
  • Desire to move to fix stage quickly
  • Skipping verification steps
  • Not testing alternative hypotheses

Prevention

  • Multiple working hypotheses rule
  • Require explicit disconfirmation of alternatives
  • Verification stage mandatory before declaring root cause
  • Ask "what else could cause this symptom?"

Sunk Cost Fallacy

Continuing failed approach due to invested effort.

Manifestations

  • "We've spent hours on this theory, it must be right"
  • Reluctance to abandon promising-but-wrong direction
  • Adding complexity to failing hypothesis instead of reconsidering

Counter-Strategies

  • Time-box hypothesis testing
  • Set explicit abandon criteria before starting
  • Treat investigation time as learning, not investment
  • Ask "if I started fresh, would I pursue this?"

Escalation Protocol

When you recognize you're stuck in a pitfall:

  1. Acknowledge — Name the bias or pattern
  2. Document — Write down current state and reasoning
  3. Reset — Return to discovery stage
  4. Reframe — Look at problem from different angle
  5. Seek outside perspective — Fresh eyes often see what you miss

If stuck for > 2x expected time, mandatory escalation or perspective shift.