What you see: research candidates ranked by quantitative models — not buy orders.

Evidence score (0–100): model evidence strength — not a calibrated probability or return forecast. Equal values are explicit ties; tied rows are not statistically ordered.

Setup state: an internal model state, not a recommendation. No Clean Setup means the model did not detect a research-grade setup right now.

What it is NOT: personalized advice, a performance promise, or a validated track record (see Methodology & Evidence).