NoiseFloor — Statistical Evidence Checks

Use ten research tools for A/B comparisons, change detection, forecasting, calibration, sampling effects, spectral and market assessments. Interpret uncertainty under each method’s stated assumptions; statistical evidence is not proof of causality or trading authority. Send only public, synthetic or explicitly authorized observations to the hosted service. Preserve unavailable and insufficient-evidence results. Hosted MCP: https://api.seiche.info/noisefloor/mcp (runtime 0.4.0). The separately p