Waveform evidence
Peak, clipping-related evidence, dynamics, crest factor, timing and spectral summaries are derived from the selected audio inside the browser.
EVIDENCE STANDARDS
Every result should say what it knows — and what it does not.
HitLytics separates measurement, heuristic interpretation, market context, human judgment and real-world outcomes so a convenient score never becomes fake certainty.
Signal 1.1Confidence labelsMethod versioningRegression gates
Readiness is guidance
Audio Readiness and genre-fit outputs are transparent engineering heuristics. They are not calibrated probabilities of commercial success.
Rank is separate context
Public Top-100 rank, artist and release metadata describe the current market. Rank is never reverse-engineered into imaginary acoustic features.
Creative judgment stays human
Hook quality, lyric meaning, emotional impact and artistic identity cannot be proven by waveform statistics and require human/listener evidence.
Confidence is visible
When browser support, file properties or evidence limits reduce confidence, the interface should say so instead of presenting false precision.
Changes are versioned
Scoring changes should pass a golden regression library across genres and edge cases, be assigned a method version, and preserve the interpretation boundary.
Golden regression library
Before a scoring release replaces the current method, it should be tested against a fixed internal library spanning multiple genres, dynamics, durations, sample rates, loudness profiles and intentionally problematic files. The goal is to detect unexpected regressions before users do.
Correction rule
If a material scoring defect is confirmed, the responsible response is to reproduce it, constrain or disable the affected metric when necessary, correct the method, document the new version, and avoid pretending the original result was infallible.
What HitLytics will never promise
No score guarantees streams, sales, playlist placement, label attention, chart position or profitability. The product exists to reduce avoidable uncertainty before a release decision — not to manufacture certainty that the evidence cannot support.