Lighthouse AdvisorySLED AI Adoption Intelligence
← Back to results

From the College Athletics edition of September 10, 2026

Academic researchMixedUndated source

Division II qualifier case exposes exception-handling errors before corrected results

Wichita State University; Jeff Noble · Collegiate athletics · United States; NCAA Division II

Publisher
Journal of Contemporary Issues in Sport
Original publication
December 2025; exact day unknown
Source retrieved
2026-09-11
Read original source

What happened

ChatGPT initially omitted a diver in the men's qualification exercise; revised prompts produced the reported correct list. This is feasibility evidence, not independent reliability validation.

Why it matters

Direct Division II event-administration evidence adds a smaller-program rules-processing example.

Evidence and measured results

The author compared outputs with manually verified answers using an NCAA example and 2024 men's and women's prequalification data. Women's qualifiers were reportedly correct on the first attempt. No reproducible multi-run error rate, dataset row count or controlled labor baseline is supplied.

Limitations and uncertainty

Practice-driven author has event-official experience; not an independent audit. Prompt tuning on known answers limits generalization. Methods cite 2024 datasets while a reference describes unpublished 2025 results. Exact event date remains unknown.

Put this evidence to work

Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-11; this does not change the original publication date. Labels below come from the analysis itself.

Sales

Role takeaway

The problem is dependable qualification processing under event deadlines. Engage event directors, officials and athletics IT. Ask how ties, duplicates and appeals are handled today and who certifies the final list. A bounded engagement could reconstruct one historical event with an independent rules engine and an optional language interface. The value hypothesis is quicker preparation of reviewable work. Do not promise fewer errors or staffing reductions from this case; include correction time and appeals in the local baseline.

Pre-sales engineering

Role takeaway

Fit is supervised administrative assistance. Prerequisites include authoritative rules, structured results and an independently approved expected answer. Test unseen tie combinations, duplicated names, missing scores and rule revisions, then repeat identical inputs to measure variation. Isolate document instructions as untrusted data and prohibit automatic result publication. Cloud use requires approved retention and identity controls; no source evidence establishes a need for specialized on-premises infrastructure. Compare the language workflow with conventional code, including total review effort.

Delivery

Role takeaway

An event director should sponsor a shadow pilot with an experienced official and a developer. Build the independent test corpus before prompt tuning, train reviewers and rehearse fallback and appeal handling. Dependencies include licensed or authorized results and protected reviewer time. Proposed acceptance criteria are complete agreement with the approved oracle on all critical qualification cases, traceable sign-off and successful manual recovery. These are proposed gates, not observed outcomes. Stop expansion when unanticipated rule exceptions appear.

Implementation considerations

Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.

Architecture and integration

What must connect, and where does the AI sit in the workflow?

Use a deterministic rules implementation as the oracle; keep any language interface outside the authoritative results pipeline.

Governance

Who approves, reviews and stays accountable for outcomes?

Freeze the applicable rule version and require an official to approve every final list.

Security and privacy

What data, permissions and controls need testing?

Use authorized competition records and minimize identifiers in model inputs; do not add medical or eligibility files.

Accessibility and workforce

Who is affected, and what skills or accommodations follow?

Teach officials to detect omissions and explain exceptions; preserve an accessible manual process.

Procurement

What should contracts, pricing and exit terms secure?

Require exportable inputs, version logs and reproducibility tests before buying an automated qualification service.

Operating model

Which teams own the service once it runs?

Event director owns the decision; technical staff maintain tests and a rollback route.

What changed

Absent from 182 archive records scanned at offsets 0 and 100. Historical evidence newly adds a concrete administrative failure mode; no new September release is claimed.

Publication history

  1. 2026-09-10College Athletics · Issue 053 resources
Read preserved resource versions (JSON)

Stable resource ID: noble-division-ii-diving-qualifier-ai-case-2025