Multiply confidence at each transmission hop — multi-step information degrades multiplicatively

Why This Is a Rule

Information degrades at every transmission step. A study's finding (90% confidence) gets summarized in a review paper (85% fidelity), cited in a blog post (75% fidelity), quoted in a tweet (60% fidelity), and forwarded to you with commentary (50% fidelity). The endpoint confidence is not 90% — it's 0.90 × 0.85 × 0.75 × 0.60 × 0.50 ≈ 17%. You're treating 17%-confidence information with the weight of 90%-confidence source material.

Each transmission step introduces degradation through compression (details lost), interpretation (meaning shifted), selection (context removed), and error (simple mistakes). These degradations multiply because each step operates on the already-degraded output of the previous step, not on the original source.

The rule prescribes multiplicative confidence tracking: at each known transmission hop, reduce your confidence estimate. The practical effect is that multi-hop information (the most common kind in social media, organizational communication, and secondary sources) gets appropriately lower confidence than primary sources — which changes how much weight you give it in decisions.

When This Fires

• Reading a summary of a study rather than the study itself
• Receiving information through organizational chains (your manager's manager heard that...)
• Consuming social media commentary about research, events, or claims
• Any time information reached you through 2+ intermediaries

Common Failure Mode

Treating endpoint confidence as source confidence: "Studies show that..." when what you actually read was a blog post summarizing a news article about a study. The confidence you have in "studies show" should reflect the full transmission chain, not just the authority of the word "studies."

The Protocol

When information arrives through multiple hops: (1) Trace the transmission chain: who said it → who cited it → who summarized it → how it reached you. (2) At each hop, estimate fidelity (0.5-0.95 depending on the quality of the transmission). (3) Multiply: source confidence × hop1 fidelity × hop2 fidelity × ... = your actual confidence. (4) Treat the information at its endpoint confidence, not its source confidence. If the resulting confidence is below your decision threshold, go to the primary source before acting on it.