How is the data validated?

Every figure on your dashboard runs through FloorRadar's proprietary multi-stage validation engine before it ever reaches you. Records are reconciled across multiple independent feeds, normalized through a tiered consistency-check pipeline, and scored against historical baselines by our anomaly-detection algorithms. Nothing reaches your dashboard on a single source's word.

Multi-source reconciliation

Our reconciliation engine cross-references each record against multiple independent inputs, applies entity-resolution algorithms to merge identifiers across sources, and rejects records that fail the consistency matrix. The validation logic is tuned to the specifics of Illinois VGT economics — graduated tax behavior, seasonality, machine-count constraints, municipality-level reporting quirks — through models we've refined against years of historical data.

Continuous re-validation

New records aren't just appended. The historical re-validation pass re-scores every period against the full corpus, surfacing any record that drifts outside its expected range. Quarantined records are escalated, investigated, and either reconciled or held back from publication. That's why a number you see in FloorRadar today is still the same number when you look six months from now.

Proprietary enrichment layer

On top of the validated baseline figures, FloorRadar's analytical engine derives a layer of proprietary intelligence — efficiency metrics, peer-cohort rankings, momentum indicators, comparable-set inference, trend-decomposition models. These aren't pulled from any upstream source; they're computed by our own algorithms, tuned for the specific signal you actually care about. The Acquirer add-on layers an additional enrichment stack specific to acquisition-grade analysis.

What this means for you

You don't have to second-guess the numbers. Every value shown is the output of a validation chain you don't have to manage, against a historical baseline you don't have to maintain, refined by an algorithmic layer you don't have to build. We do the reconciliation, normalization, anomaly detection, and enrichment work so you can spend your time on the decision, not the data.