Partners · Research institutions
Put the Athlete Twin to the test.
Opti-U wants university partners to own the evaluation of the Athlete Twin: pre-register the analysis, run it independently and publish what they find.
Why it matters
Independent evidence is scarce in this field.
Peer-reviewed reviews rate football injury models as exploratory, with a high risk of bias. Opti-U is building the product so its claims can be checked by people who have no stake in the answer.
<01>
One pre-registered protocol
The same protocol runs at every site, which gives the findings external validity across clubs.
<02>
The academic partner owns the analysis
Analysis plans, the device agreement study and the usability evaluation sit with the university, so validation claims stay independent of Opti-U.
<03>
Consent to publish, built in
Every design partner club agrees to independent evaluation and publication before it joins.
Research questions
Open questions a live squad can answer.
Each one comes with longitudinal data on real players, gathered under agreed governance.
<01>
Forecast calibration
Do prediction intervals hold their nominal coverage on held-out weeks, for individual players?
<02>
Device agreement
How closely do consumer wearables agree with chest-strap reference measures in elite players?
<03>
Individual thresholds
What is the smallest detectable change for each measure, and how does it vary between players and fixture phases?
<04>
Practitioner value
How much staff time does the twin save against a stopwatch baseline, and how usable do practitioners find it?
<05>
Learning from decisions
With alternatives logged for each decision, which interventions work for which players across seasons?
The partnership
What a research partner gets.
<01>
A reproducible engine
Versioned models with a deterministic mode: the same inputs and model version give the same output.
<02>
Rich longitudinal data
GPS load, session RPE, wellness, availability and practitioner decisions on whole squads, under data-sharing agreements, player consent and UK GDPR.
<03>
Cross-disciplinary work
Sports science, sports medicine, machine learning and psychology questions sit in one programme.
<04>
Routes to funding
Opti-U is pursuing a Knowledge Transfer Partnership and co-funded doctoral studentships for questions that run beyond two seasons.
<05>
A clear timetable
Design and retrospective work in 2027/28, then the first live season in 2028/29.
Opti-U is looking for academic partners.
The engine is built and tested on real human data, and the sports layer is a working design prototype on simulated data. The first live season will be evaluated independently.