{
  "id":"b06-bracket-rewards-v1",
  "declaredBeforeResults":"2026-10-15",
  "question":"How do two double byes versus fourteen single byes change entry burden, title probabilities and modeled travel under the same field and dates?",
  "baseline":"Existing 2024 historical allocation: 2 double, 8 single, 8 opening; games 4,6,4,2,1; redraw before round two and quarterfinals; lock thereafter.",
  "alternative":"Same 18 entrants; 14 single byes including every selected champion and all pre-title top-ten claims; bottom four remaining play 2 campus openers. Games 2,8,4,2,1; same two redraws, same quarterfinal lock. Eight neutral round-two slots require two additions to six-bowl inventory. Removes all double-bye protection as an explicit unadopted policy departure.",
  "dates":"2024 historical scenario 0: December 14, December 21, January 1, January 11, January 18. Same dates and last actual games for both. Illustrative placements, not approvals.",
  "probability":"Illustrative logistic model. Equal strengths, strengths declining with earned rank, and rank 1/rank 8 strength swap. These synthetic strengths are deliberately not empirically calibrated. No NRR Index-to-margin conversion or real-team forecast.",
  "sensitivity":"Home effect 0 or 0.2 logit units. Rest: none; +0.02 per extra day above seven capped at fourteen extra days; or -0.03 per day idle beyond21. Parameters are declared scenarios, not estimated effects.",
  "runsPerSettingPerFormat":20000,
  "seed":6062026,
  "measures":["title probability by earned rank and separate strength rank","earned ranks14–18 title share","strongest-group title share","each entrant first-game idle days and maximum path","expected games and out-and-back trips from campus under a return-after-each-game assumption"],
  "checks":["18 distinct entrants; 17 eliminations; single champion","all five champions skip opening; baseline protected double bye remains rank3 Texas","exactly two redraws and locked semifinals","no team twice per round; dates after last game and intervals at least7days; no Christmas-week game","equal neutral no-rest probabilities reconcile to 2^-wins-needed within Monte Carlo uncertainty","finalists' maximum season totals and trip bounds"],
  "decision":"Publish mechanics and sensitivity; do not adopt alternative based on synthetic probabilities. Actual predictive-strength calibration, athlete-led rest review, eight-slot venue/rights feasibility and priced travel remain open. Demand and coaching effort are outside scope."
}