Detection Verification Orderings
Input: $ARGUMENTS
Overview
When investigating potential deception or anomalies, the order of investigation matters. Check the hardest-to-fake evidence first, establish baselines before looking for deviations, and verify independently before confronting.
Core Principle
Liars optimize for the most obvious checks. Your ordering should prioritize whatโs hardest to fake, most independent, and least expected.
Ordering Rules
Rule 1: Hardest to Fake First
- Check evidence that would be most difficult to fabricate or manipulate
- Physical evidence before testimony, data before narrative, timestamps before content
- Why: if the hardest-to-fake evidence is clean, softer evidence is more trustworthy
Rule 2: Baseline Before Deviation
- Establish what normal looks like before looking for abnormal
- Compare against historical patterns, peer behavior, expected distributions
- Why: you canโt detect anomalies without knowing what normal is
Rule 3: Independent Sources Before Correlated
- Check independently generated evidence before checking sources that could have been coordinated
- Why: correlated sources could tell a consistent false story
Rule 4: Periphery Before Center
- Check surrounding evidence before the main claim
- Verify context, metadata, and secondary details
- Why: deception focuses on the main story; peripheral details are more likely to be authentic or reveal inconsistencies
Rule 5: Pattern Before Instance
- Look for systematic patterns before investigating individual cases
- Statistical anomalies across many instances are harder to explain away than single events
- Why: one anomaly might be coincidence; a pattern is signal
Application Procedure
Step 1: Define What Youโre Checking
- What is the claim or behavior under investigation?
- What would deception/fraud look like?
- What would innocence look like?
Step 2: Identify Evidence Types
- Physical / digital / testimonial / statistical
- Rank by difficulty to fake
- Identify which are independent vs correlated
Step 3: Execute in Order
- Check hardest-to-fake evidence
- Establish baseline patterns
- Check independent sources
- Look at peripheral details
- Examine statistical patterns
- Only then: confront or conclude
Anti-Patterns
- Starting with confrontation (tips off deceiver)
- Relying on testimony alone (easiest to fake)
- Checking only the main claim (missing peripheral inconsistencies)
- Confirmation bias (looking for guilt, ignoring innocence evidence)
When to Use
- Fraud investigation
- Due diligence
- Fact-checking claims
- Security audits
- Any situation where trust needs verification
Verification
- Evidence ranked by difficulty to fake
- Baseline established
- Independent sources checked before correlated
- Peripheral evidence examined
- Both deception and innocence hypotheses considered