BETTER COLLEGE FOOTBALL READER SESSION — v3, October 19 2026
Blank protocol. No participants or observations are represented by this file.
Recruit five willing adults: ideally three unfamiliar with the proposal and two familiar with college football. Use anonymous IDs R01–R05; record familiarity separately. Do not request private academic, medical or financial details. Obtain consent to record anonymous answers; no email addresses or names in published observations.

Use the current quick overview and /evidence/nrr-examples. Let the reader navigate without coaching; record the page/version/date and the exact starting route. Explain that this tests the explanation, not the reader. Do not ask for agreement with 18 teams or model weights.

Ask, in order:
1. In your own words, who skips two rounds, who skips one, and who plays first? What is the difference between underlying ranking and a protected benefit?
2. Choose one worked pair. What makes the teams rank differently? Point to the evidence you used.
3. What is the difference between a component score, its weight and its contribution to the final Index? Does an Index of 80 imply an 80% chance of winning?
4. If A beat B, B beat C and C beat A, can the chain alone establish a unique best team? What information does it leave out?
5. Name one unresolved tradeoff in the proposal and one piece of evidence that might change your view.

For each answer record: exact or faithful anonymous wording; unaided/after prompt; navigation path; confusing label; time if actually measured; facilitator intervention. Mark missing answers missing. Do not invent a pass percentage or statistical significance from five convenience participants.

Blank record:
Reader ID:
Familiarity:
Date and page version:
Consent:
Starting route:
Q1 answer / assistance:
Q2 answer / assistance:
Q3 answer / assistance:
Q4 answer / assistance:
Q5 answer / assistance:
Confusing words:
Facilitator notes:

Review: preserve first answers; prioritize any confusion about probability/units. Revise recurring confusion seen in at least two readers, documenting the reason, and retest affected questions. Report individual counts, recruitment and limits. Agreement is not comprehension; a correct prompted answer is not an unaided answer. Anonymous blank forms are not completed sessions.