Show me the capital.
Then prove the data makes it better.
Sports Alpha is not asking Genius to buy research. It is asking a commercial question: can official data increase accepted value, remove avoidable losses or extend the usable execution window on the same cases?
Prospective Challenger A
Snapshot date unavailable
This is a prospective research scorecard translated into nominal capital at the selected unit size. It is not realised money, does not assume fills or liquidity, and is not a production-authorised strategy.
Match Desk
Find a fixture, inspect its recorded decision and follow the result. Catalog entries and research decisions are shown separately.
Select a fixture
The decision timeline will appear here.
One question:
what happens to the money?
Take the frozen prospective record, choose a reference bank and a flat unit size, and translate the observed units into nominal capital. Then compare the selective filter against the raw policy on the same research universe.
Same universe. Different selectivity.
The delta is descriptive prospective research accounting. It is not a realised trading return and does not prove execution scalability.
What it refused
Make the unit economics explicit
Prospective Challenger A
Challenge Sports Alpha
Choose from the released future-fixture catalog. Registration records your choice; it does not trigger model inference or guarantee that a decision will be produced.
Select an approved fixture
Recorded at registration
A repeated registration of the same fixture and market returns the existing receipt.
Recent evaluation ledger
Sports Alpha Live
Live visibility without live observability. Only safe, high-level events leave the private system.
What is happening now
The model is not the product.
Sports Alpha is the system that decides which research deserves capital — and which research gets killed.
Reported research state
Outcome-led evaluation
External visibility is intentionally restricted. Partners can observe outcomes and evidence state, not model generation, features, thresholds, orchestration, or training internals.
Published research summaries
What is proven.
What is not.
The money view is deliberately decomposed so a partner can audit the prospective research record without mistaking it for an executed portfolio.
The cash claim stops here.
External evidence receipts
Can your data make
this capital case worth more?
We already have a frozen prospective baseline. The joint test is simple: on the same cases and the same clock, does Genius data increase net units, improve capital selectivity or preserve an actionable state for longer?
Genius only wins if the money endpoints move.
No Genius uplift is claimed today. The purpose of this Lab is to make that uplift impossible to fake: same fixtures, common clock, frozen endpoints and future-only evidence.
Experimental R&D: two future-only decision challengers are tracked below as non-authority research. They cannot change the production decision path.
Future-only challengers under governed validation
These programs show how Sports Alpha tests new decision architecture without widening rules after the fact. They are research evidence, not production models, trading signals or commercial uplift claims.
One path from raw state to commercial evidence
The supplied samples can now be opened from Match Desk and replayed here. Capital and uplift remain deliberately unfilled until a paired future evaluation produces admissible evidence.
What is proven — and what still has to earn its way in
Walk one case from source state to commercial evidence
Use a supplied case, inspect its revision-aware replay, freeze the paired protocol, then see exactly which future evidence would unlock decision, execution and capital claims.
Commercial endpoints, not research vanity metrics.
Measurement status & boundaries
Register the paired test
Use the supplied 2026 coverage reference to frame a joint test. This does not call Sports Alpha, activate a feed, or claim a performance uplift.
Choose a scope
Coverage reference and measurement rules will appear here.
Raw event → normalized event state
Replay the supplied sample archive inside the isolated Lab. The normalization shown here is a partner-evaluation projection, not a claim that the Sports Alpha core has ingested this feed.
Timestamp delta = messageTimestampUtc − action.utcTimestamp inside the supplied payload. It is a payload timing/data-quality measurement, not measured transport or end-to-end feed latency.
What we would test together
| Question | Paired measurement | Why it matters |
|---|---|---|
| Does state arrive earlier? | Event, provider and receipt times on aligned clocks | Time available to make a decision |
| Can more decisions be evaluated? | Eligible observations on the same fixtures and markets | Usable coverage rather than headline volume |
| Does the extra information help? | Predefined probability and repricing endpoints | Separate information value from outcome luck |
| Does it survive execution constraints? | Quote age and observation window; fills only if actually observed | Distinguish research results from attainable performance |
Proposed sequence: agree scope → evaluate supplied samples → freeze the paired protocol → observe future cases → review the commercial case. No uplift or trading volume is assumed.
Supplied coverage report · reference
Coverage reference supplied by Genius
—
—
Sports Alpha without Genius vs. Sports Alpha + Genius
Baseline timing, provenance and execution window.
Measured only when a real paired feed is available.