{"resourceId":"sportico-sec-fastbreak-scheduling-ai-2025","versions":[{"version":"external-99bcf26ff87e332bb1fd51fb15d8e0157b7ae0bdc4788e0be1fdcc6e77e324d6","resource":{"id":"sportico-sec-fastbreak-scheduling-ai-2025","title":"SEC Fastbreak AI scheduling posts conference-claimed fairness gains for 2026 slate","organization":"Sportico","sector":"College athletics / conference operations and scheduling","geography":"United States — Southeastern Conference (16 members)","publishedAt":"December 19, 2025","publicationDate":"2025-12-19","eventDate":null,"sourceName":"Sportico","sourceLabel":"Independent reporting — SEC AI scheduling","sourceUrl":"https://www.sportico.com/leagues/college-sports/2025/2026-sec-football-schedule-process-conference-espn-games-ai-1234879594/","evidenceClass":"independent-reporting","outcomeClass":"mixed","topics":["operating-model","infrastructure","knowledge-work","governance-procurement"],"finding":"Jacob Feldman reports the SEC uses Fastbreak AI across football and about 10 sport schedules; associate commissioner Garth Glissman says the software produced men's/women's basketball calendars 25-30% fairer on the conference's own slate metrics and claims analytics prove schedules objectively better, while football assistant commissioner Matt Evans cites five back-to-back away instances for 2026—lowest since 2015—yet stresses dozens of human iterations because mathematical rankings miss contextual equity.","sledRelevance":"Why now: 2026-09-17 left facilities/admin conference-ops AI uncovered after ChatGPT eligibility, Living Lab, and NIL monitoring. The 2026 nine-game SEC football season is live; this still-unarchived Sportico process piece supplies rare quantified conference scheduling outcomes plus human-override constraints transferable to other SLED athletic conferences.","evidence":"Glissman 25-30% fairer quote; Evans five back-to-back away / lowest since 2015; description of loading constraints into Fastbreak for minute-scale scored options; Groer note that second-to-last vs last schedule versions can differ dramatically; ESPN media-partner input acknowledged.","architectureImplications":"Constraint-optimization SaaS with conference-defined scoring, media-partner parameters, and human re-parameterization loops—not one-shot autonomous schedule generation.","governanceImplications":"Document who owns fairness definitions, how media partners influence tweaks, and when human overrides supersede model rankings for student-athlete rest/travel equity.","securityPrivacyImplications":"Schedule platforms ingest competitive and commercial constraints more than biometric PII, but travel/rest optimizations still affect athlete welfare and need transparent criteria.","caveats":"Fairness metric is conference-defined and not third-party audited; Dec 2025 publication predates last run (explicit why-now applied); vendor Series A context disclosed in article.","streamIds":["college-athletics"],"roles":{"sales":"Sales framing: Power conferences facing expansion and nine-game football slates are already buying AI schedule optimizers and quoting internal fairness KPIs, but Sportico shows schedulemakers still need weeks of human iteration and media-partner feedback. Position offerings as constraint-solvers that surface scored alternatives for deputy commissioners—not autopilots that lock TV windows alone. Discovery should map permanent rivals, rest/travel rules, and ESPN-class obligations before promising equity gains. Do not sell a single fairness percentage as transferable without the conference's own metric definition.","engineering":"Engineering framing: Fit is multi-constraint combinatorial optimization with scored schedule candidates, configurable home/away cadence and rest parity, and export into conference calendars. Prerequisites include encoded rivalries, nonconference holds, broadcast windows, and a versioned fairness objective. Constraints: team-controlled nonconference dates, fixed rivalry weekends, and opaque media preferences. Proof of value is conference-agreed fairness deltas plus rest/travel KPIs under holdout seasons—not vendor demos alone.","delivery":"Delivery framing: Pilot one sport's slate with schedulemakers owning accept/reject of AI candidates, log parameter tweaks across dozens of iterations, and brief ADs on student-athlete rest/travel impacts. Dependencies: conference ops staff like Evans/Glissman roles, broadcast partners, and institutional preference intake. Skills adoption teaches interpreting scores versus contextual equity. Governance checkpoints: written fairness definition and media-input policy. Acceptance criteria: agreed fairness/rest KPIs before locking the public schedule."},"roleDetails":{"sales":{"customerProblem":"Expanded conferences struggle to hand-build fair multiport schedules under travel, rest, rivalry, and broadcast constraints.","stakeholders":["Conference deputy/associate commissioners","Member ADs","Broadcast partners","Student-athlete welfare leads"],"discoveryQuestions":["How is schedule fairness scored today?","Which constraints are hard vs soft?","Who overrides an AI-ranked slate?","How do media partners enter the loop?"],"valueHypothesis":"Constraint-scored AI can improve conference-defined fairness and rest metrics if humans still iterate parameters—as Glissman/Evans describe—not if TV windows alone drive acceptance.","potentialEngagement":"One-sport shadowed schedule pilot with published fairness/rest KPI comparison to prior season.","unsupportedClaims":["Guaranteed TV ratings lift","Transferable 25-30% fairness without local metrics","Fully autonomous schedule lock"]},"engineering":{"fit":"Combinatorial schedule optimization SaaS with scored candidates and human re-parameterization.","architectureIntegration":"Ingest rivalries, venues, rest rules, and broadcast windows; emit ranked calendars into conference publishing tools.","prerequisites":["Encoded constraints","Versioned fairness objective","Historical slate baselines","Role-based tweak logs"],"constraints":["Team-controlled nonconference dates","Fixed rivalry traditions","Media-partner soft constraints","Opaque fairness definitions across conferences"],"security":"Protect competitive strategy and unpublished slate drafts; limit external partner access to constraint sets.","proofOfValue":["Conference-agreed fairness delta","Back-to-back road / unequal-rest counts","Iteration count to lock","Override rate after AI draft"]},"delivery":{"work":"Configure constraints, run iteration workshops with schedulemakers, document media feedback, and publish acceptance KPIs before release.","dependencies":["Conference ops ownership","Broadcast liaison","Member school preference process"],"ownership":"Conference scheduling leadership owns the locked slate; vendor supplies solver.","skillsAdoption":"Train staff to read scores and challenge mathematically high but contextually unfair drafts.","governanceCheckpoints":["Written fairness definition","Media-input policy","Student-athlete rest review"],"acceptanceCriteria":["Agreed fairness/rest KPIs met","Rivalry and broadcast obligations preserved","Override log complete"],"risks":["Overfitting to TV preferences","Member school pushback","Opaque vendor scoring","Student-athlete travel inequity"]}},"retrievedAt":"2026-09-19T03:10:20Z","enrichedAt":"2026-09-19T03:10:20Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Conference-level AI scheduling concentrates expertise in ops analytics roles; member schools still need staff to interpret travel/rest impacts on athletes.","procurementImplications":"Buy against conference-owned fairness definitions and human-iteration SLAs; separate broadcast-optimization goals from student-athlete welfare KPIs in contracts.","operatingModelImplications":"Moves schedule drafting from single-exec manual craft to AI-assisted iteration owned by conference ops with media partners as influencers, not sole decision-makers.","sourceVerification":{"openedUrl":"https://www.sportico.com/leagues/college-sports/2025/2026-sec-football-schedule-process-conference-espn-games-ai-1234879594/","referenceExcerpt":"create a calendar that is 25-30% fairer based on the way the conference measures its slates","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}