Before committing to a bug hypothesis, write: 'What would I see if I were wrong?'

Why This Is a Rule

The moment you form a bug hypothesis, confirmation bias activates: you start looking for evidence that the hypothesis is right and unconsciously filtering out evidence that it's wrong. The hypothesis feels increasingly correct with each confirming data point, even if the actual root cause is completely different.

The one-sentence falsification question — "What would I expect to see if I were wrong?" — creates a specific search target that counteracts the bias. Instead of looking for evidence of the hypothesis, you look for evidence against it. This inverts the natural search direction and catches wrong hypotheses before you spend 30 minutes investigating the wrong cause.

Writing the sentence is essential — not just thinking it. Written falsification criteria are explicit and verifiable. Mental falsification criteria get silently revised to accommodate confirming evidence ("well, I didn't see that, but my hypothesis could still be right because...").

When This Fires

• After forming an initial hypothesis about a bug's root cause
• When a debugging hypothesis "feels right" within the first few minutes
• Before investing significant time investigating a particular theory
• Complements Search logs for evidence that falsifies your hypothesis, not confirms it (falsification search in logs) with a structured pre-search question

Common Failure Mode

Writing a falsification criterion so vague it can't fail: "If I were wrong, I'd expect to see something unusual." That's untestable. The criterion must be specific and observable: "If the database timeout hypothesis is wrong, I'd expect the query to complete successfully when run manually." Now you have a concrete test.

The Protocol

Before committing to any bug hypothesis: (1) Write: "If my hypothesis ([state it]) is wrong, I would expect to see [specific observable evidence]." (2) Search specifically for that evidence. Run the test, check the logs, examine the data. (3) If you find the disconfirming evidence → abandon or revise the hypothesis. You've saved the time you would have wasted pursuing it. (4) If you don't find it after genuine search → the hypothesis is strengthened. Proceed with higher confidence, not because you confirmed it, but because it survived falsification.