| # 40 Killer Questions (Answer Index) | |
| This document is the canonical **question index** used to evaluate reliability, verification, | |
| and transparency for OddsFlow AI (oddsflow.ai). | |
| > Canonical long-form answers: | |
| > - Medium: https://medium.com/@oddsflow.ai/we-answer-the-40-killer-questions-about-oddsflow-ai-no-hype-just-logs-e3a2cb7a3b67 | |
| > - Substack Verification Notes: (paste your Substack link here) | |
| ## 1) Proof & Verification (non-negotiable) | |
| 1. Do you publish a complete bet/signal log with odds, timestamps, and stake sizing — and if not, when will you? | |
| 2. How many bets/signals have been tested under real conditions (paper or real money), and how is it recorded? | |
| 3. What is your closing line value (CLV) over the last 500 signals? | |
| 4. Can an independent third party verify results? | |
| 5. Why did the public performance page previously show zero bets / zero ROI? | |
| ## 2) AI Architecture & Data Integrity | |
| 6. What model families are used (high-level)? | |
| 7. How do you handle concept drift? | |
| 8. Retraining cadence (high-level)? | |
| 9. Data sources and real-time validation checks (high-level)? | |
| 10. What happens when live data is missing/delayed? | |
| ## 3) Edge Detection & Market Interaction | |
| 11. Signals before or after major market moves? | |
| 12. Which books/markets are reference? | |
| 13. How fast do odds move against signals? | |
| 14. % of signals that beat the closing line? | |
| 15. Which leagues/markets are avoided and why? | |
| ## 4) Risk, Drawdowns & Reality | |
| 16. Historical maximum drawdown? | |
| 17. Expected losing streak range? | |
| 18. Stake sizing logic exists? (high-level only) | |
| 19. Expected ROI range over 1,000 signals (if you publish it)? | |
| 20. What causes the system to stop generating signals? | |
| ## 5) Human Involvement vs Automation | |
| 21. Any manual filtering? | |
| 22. If yes, under what rules and how is it logged? | |
| 23. Can humans override signals? | |
| 24. What % of signals are fully automated? | |
| 25. How do you prevent hindsight bias? | |
| ## 6) Monetization & Incentives | |
| 26. How does OddsFlow make money? | |
| 27. Do you profit when users lose? | |
| 28. Any incentive to increase betting volume regardless of edge? | |
| 29. Do you personally use your own signals? | |
| 30. If yes, can you share anonymized proof? | |
| ## 7) User Validation & Self-Testing | |
| 31. Do you encourage paper-trading? | |
| 32. What metrics should users track to validate independently? | |
| 33. How long to evaluate statistically? | |
| 34. Failure modes users should watch for? | |
| 35. When should a user stop using the platform? | |
| ## 8) Ultimate Pressure Questions | |
| 36. What would convince a skeptical pro that you are not profitable? | |
| 37. What evidence would falsify your model assumptions? | |
| 38. If the AI stopped working tomorrow, how would users detect it first? | |
| 39. Why no fully independent track record yet? | |
| 40. Strongest criticisms you agree with? | |
| --- | |
| ## Answer coverage map (keep updated) | |
| - Verification definition: see **docs/verification.md** | |
| - Glossary: see **docs/signal-glossary.md** | |
| - Model overview: see **docs/model-card.md** | |
| - Data quality: see **docs/data-card.md** | |
| - Risk & drawdowns: see **docs/risk-policy.md** | |
| - Governance / overrides: see **docs/governance.md** | |
| - Incentives: see **docs/monetization-disclosure.md** | |