Estimate the illustrative impact of a proposed fraud-rule change by deterministically evaluating the current and proposed rule over a labeled transaction sample — block counts, decline-rate delta, and the confusion-matrix shift (caught fraud, missed fraud, false positives) — for a fraud-strategy analyst to review. Use when an analyst asks to "estimate the impact of this fraud rule", "how many transactions would this rule block", "what's the false-positive tradeoff", or "compare the current vs proposed rule on this sample". HARD BOUNDARY: R3 model/decision-support only — it produces an illustrative estimate on a sample; it NEVER deploys, enables, or rolls out a rule, and N…
