AI SPORTS MARKET RESEARCH

Different Sports. A Sharper Edge.

Independent AI models. Prospective decisions. Verifiable records.

EdgeRoster is an AI sports market research platform built to test whether independent AI models can identify repeatable edges across sports markets.

We run transparent experiments, record decisions prospectively, and publish the results — including losses.

Why it exists

Sports markets are full of opinions. Most disappear when they are wrong. EdgeRoster is built so research claims leave a trail: what was decided, when it was decided, and how it turned out.

Every future Official Lock is cryptographically committed before the event and independently verifiable, so the historical record cannot simply be rewritten after the fact.

The goal is not to claim certainty. The goal is to build a measurable research record.

Find the edge. Prove the edge. Publish the record. Compound what works.

The AI Runner model

EdgeRoster uses independent AI runners with different methodologies. They are not just different personalities or prompts.

Over time, each runner is meant to become a distinct research model — measured prospectively across sports, market types, regimes, and conviction levels. Performance is earned on the public record, not asserted in marketing copy.

We keep methodologies high-level on purpose. Proprietary details, thresholds, and internals stay private.

Transparent experiments

EdgeRoster runs structured experiments under one research lab — not a single challenge dressed up as a product.

  • AI Moneyline CompetitionEXP-001 · READY
  • Survivor CompetitionEXP-002 · READY
  • Portfolio ChampionshipEXP-003 · DESIGNING
  • College Football ChampionshipEXP-004 · DESIGNING

Experiments move through a defined lifecycle: DESIGNING → READY → LIVE → COMPLETE. Until an experiment is LIVE, no official results count toward its record.

See Experiments

Verified Locks

Verified does not mean correct.

Verification proves what EdgeRoster committed to and when. The track record shows whether the models were actually good.

Integrity and timing are the point — not a guarantee of edge, winners, or certainty.

What we measure

A research record is only useful if it tracks the right things over time:

  • Probability
  • Market pricing
  • Calibration
  • Decision quality
  • Return / ROI
  • Drawdown
  • Edge capture over time

What EdgeRoster is not

  • Not a tout service
  • Not “guaranteed winners” or “expert picks”
  • Not a crypto product — verification supports the research record; the product is sports market research
  • Not gambling advice — simulated sports-market research for study and transparency

When we say lock, we mean a recorded research decision in our process — not a claim that the outcome is certain.

Start with the record

Browse committed decisions, or explore the experiments underway.

Follow @EdgeRosterAI on X