Higher Education & ResearchIssue 284 records
College Athletics
Vendor claims outpacing evidence is today’s college-athletics AI signal. Memphis Athletics added two AI tools in two weeks: Evolv Express Gen2 AI weapons screening at its football stadium (vendor release Sep 23, 2026) and Atavus film-based tackling analytics for coaching and transfer-portal evaluation (Oct 2, 2026). Neither announcement reports detection, accuracy, or outcome data. The FTC’s 2024 action against Evolv, included because of the new stadium deployment, alleged unsupported claims about detecting all weapons, false alarms, and 70% labor savings, and cites a missed knife and a 50% false-alarm rate after sensitivity increases. On athlete health, an East Texas A&M conference abstract using DXA and DARI screening on 1,258 NCAA DI records reports AUC 0.74 for 180-day injury risk and 50% top-1 body-region accuracy, a modest benchmark as Ohio State announces predictive human-performance AI. Gaps: no venue-level Evolv Gen2 performance data at college stadiums; no validation of Atavus metrics; no external or prospective validation of the injury model.
- Evidence records
- 4
- Cross-source patterns
- 2
- Evidence classes
- 2 vendor claim1 government evaluation1 academic research
- Outcomes
- 2 emerging1 cautionary1 mixed
- Source freshness
- 2 new this fortnight1 older, newly relevant1 undated
- Research completed
- 2026-10-05
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
Athletics AI purchases are arriving with vendor claims but no venue-level performance data
Memphis adopted both Evolv AI screening and Atavus film analytics within two weeks and both announcements rely on vendor and staff statements, with no published detection, accuracy, or outcome metrics. The FTC Evolv action shows AI performance claims in safety screening can be unsupported and that sensitivity trade-offs raise false alarms, so departments need substantiation, testing rights, and measured local results.
Operating questionBefore an AI tool informs safety or roster decisions, has the department obtained substantiation for vendor claims and scheduled its own measured test?
Supporting evidenceUniversity of Memphis Athletics / Evolv TechnologiesU.S. Federal Trade CommissionUniversity of Memphis Athletics / Atavus
Athlete-scoring models should be validated locally before they affect roster or medical decisions
Memphis plans to use Atavus tackling metrics in transfer evaluation without disclosed validation, while the East Texas A&M abstract shows screening-based injury prediction reaches only moderate discrimination (AUC 0.74) and 50% top-1 body-region accuracy on retrospective single-site data. Both point to local validation and human review before scores drive decisions about athletes.
Operating questionHas the department measured how well an athlete score agrees with expert judgment or outcomes on its own data, and who reviews each decision it informs?
Supporting evidenceUniversity of Memphis Athletics / AtavusEast Texas A&M University ROARHP Lab (Strogalev, Bernhardt, Oldham et al.)
Evidence ledger
Full record · every source keeps its link and limitations
Memphis football adds Atavus film-based tackling analytics for coaching and transfer-portal evaluation
Memphis football partnered with Atavus to evaluate game film and turn tackling performance into position-specific coaching insights, and to add tackling consistency as "another source of information" when evaluating transfer-portal prospects. Staff quotes describe understanding why missed tackles occur and looking "beyond the highlight reel." The release reports no accuracy, validation, injury, or tackling-outcome metrics; Atavus describes its own methodology as "proven" without citing evidence.
Why it matters, evidence and limitations
- Why it matters
- Shows athletics departments extending film-analytics vendors from player development into roster-acquisition decisions, where a vendor score can influence scholarships and roster spots without disclosed validation.
- Evidence and measured results
- Department announcement dated Oct 2, 2026 with quotes from head coach Charles Huff and Director of Athletics Performance Ben Ashford. No study, sample, baseline, or measured result. The "proven training methodology" phrase is an Atavus self-description in the boilerplate. The release does not say whether analysis is automated, human-coded, or hybrid.
- Limitations and uncertainty
- Vendor-partner announcement only; no independent validation or outcome data; analysis method (AI, manual coding, or hybrid) not specified. Memphis-specific; limited transferability.
Memphis Athletics deploys Evolv Express Gen2 AI weapons screening at Simmons Bank Liberty Stadium
Evolv announced that Memphis selected Evolv Express Gen2 as Official Patron Screening Provider for its football stadium, replacing the existing screening solution; installation finished late August 2026 ahead of the season. Memphis cites faster, more efficient entry; Evolv cites "technical performance." No detection rates, false-alarm rates, throughput, staffing, or cost figures are given.
Why it matters, evidence and limitations
- Why it matters
- Stadium entry screening is one of the most consequential AI deployments in athletics operations because detection failures are safety failures; this is a fresh Division I adoption of a product whose marketing claims were the subject of FTC action.
- Evidence and measured results
- Vendor press release dated Sep 23, 2026 with quotes from Evolv and Memphis Associate AD Tabrian Arnold. States roughly 200,000 fans per season and a replacement of the prior system. Reports DHS SAFETY Act QATT designation and "more than 4.5 billion people" scanned since 2019 as company-level claims. No venue-level performance data.
- Limitations and uncertainty
- Vendor press release; outcome claims are qualitative. The 4.5 billion and QATT statements are company-level, not Memphis results.
FTC action against Evolv: AI weapons-screening claims unsupported; order bars misrepresenting detection, false alarms, and labor savings
The FTC alleged Evolv deceptively claimed Evolv Express would detect all weapons, ignore harmless items, outperform metal detectors on accuracy and speed, reduce false alarms, and cut labor costs by 70%. The complaint cites a school case where a seven-inch knife was missed and higher sensitivity then produced a 50% false-alarm rate. The proposed order bans unsupported claims about detection, false alarms, speed, labor costs, testing, and AI performance, and let certain K-12 customers cancel contracts.
Why it matters, evidence and limitations
- Why it matters
- Why now: Memphis Athletics deployed Evolv Express Gen2 at its football stadium for the 2026 season (vendor release Sep 23, 2026), and the FTC page notes the scanners are used at sports stadiums. The record defines what athletics venue buyers should require before relying on AI screening claims.
- Evidence and measured results
- FTC press release, Nov 26, 2024; Commission vote 5-0 to file complaint and stipulated order. Facts are FTC allegations and settlement terms, not an adjudicated finding of fact. The release says Express scanners were in over 800 schools across 40 states per the complaint. It does not evaluate the Gen2 model specifically.
- Limitations and uncertainty
- 2024 action concerning claims mostly in K-12 marketing; allegations settled without adjudication; Gen2 performance at college stadiums is not assessed. Older source included because of the current Memphis deployment.
East Texas A&M abstract: DXA and DARI screening ML predicts 180-day injury risk in 1,258 NCAA DI athlete records with AUC 0.74
A retrospective two-stage CatBoost model using DXA body composition, DARI markerless motion analysis, and Y-Balance data from 1,258 NCAA DI athlete records (246 injuries within 180 days; scans Aug 2022 to Dec 2025) reached AUC 0.74 on a hold-out set for any injury. Body-region prediction among injured athletes had 50.0% top-1, 62.5% top-2, and 77.1% top-3 accuracy. The authors note single-site prediction remains challenging given class imbalance and limited samples.
Why it matters, evidence and limitations
- Why it matters
- Why now: Ohio State announced Project SCARLET (Sep 29 to Oct 1, 2026), using predictive models across 36 sports; this abstract gives a concrete, modest benchmark for what screening-based injury prediction achieves in Division I data.
- Evidence and measured results
- Master’s-category conference abstract; retrospective single-institution dataset; hold-out testing but no external validation, calibration, or prospective test reported. AUC 0.74 for binary injury risk; body-region top-k accuracies as stated. No sport-by-sport breakdown or intervention outcome.
- Limitations and uncertainty
- Conference abstract, not peer-reviewed full paper; single institution; retrospective; publication date not stated beyond 2026; AUC 0.74 is moderate discrimination.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Vendor claim
- A supplier-provided assertion that has not been upgraded to independent evidence.
- Government evaluation
- A public body’s measured evaluation or documented pilot.
- Academic research
- Research produced through an academic institution or peer-reviewed venue.