AlphaAssayThe Signal Validation Benchmark
This page is the public scoreboard for AlphaAssay's own correctness — every row is reproducible by anyone, free, today. A validator you cannot test is just another promise, so we publish the tests: known-answer specimens any agent can replay bit-identically, our live self-grading endpoints, and — as the public ledger accrues — mortality statistics of real strategy families. Methodology audits, not investment advice; no strategy identities are ever published.
Known-answer verification (reproducible now)
Four specimens with planted properties; the battery must catch the flaws, pass the clean one and abstain on thin data. Replay any row: quickstart.
| specimen | planted property | expected verdict | what it proves |
|---|---|---|---|
| golden_clean | genuine persistent edge | ✓ pass | we don't cry wolf |
| golden_lookahead | leaked future information | ✕ fail · edge_collapses_at_lag1 | we catch leakage |
| golden_cherry | best-of-300 cherry-pick | ✕ fail · family_budget_exhausted | we price the search |
| golden_thin | too little data to judge | insufficient_evidence | we admit uncertainty |
Field result: 9 popular public strategies, full battery
We ran nine of the most popular public TradingView strategy classics through the complete battery on real market data with real costs. Aggregate result: 0× pass · 2× conditional · 7× fail — the six 1-hour classics lose money after costs (Sharpe −3 to −12). The per-strategy autopsy series with named setups and failure codes is in preparation and will publish on this URL's changelog.
Live self-grading (pull it yourself)
| what | endpoint |
|---|---|
| daily calibration record — how our past verdicts turned out | GET /v1/public/calibration |
| graveyard digest — anonymised mortality of strategy families | GET /v1/public/graveyard-digest |
How the record works and why it cannot be doctored: how we grade ourselves.
Methodology
Deterministic four-gate battery (net edge → family deflation → placebo vs. 500 matched twins → robustness attacks), verdicts signed ed25519 and bit-identically replayable — full description. Benchmark policy: this URL is permanent; re-runs append to the changelog below; the year lives in the title only.
Changelog
| 2026-07 | benchmark page established: 4 known-answer specimens; 9-classic field aggregate; live endpoints |