Ranking method:
leading_score = ln(1 + own_citation_count) × √(mentions_in_case) × flag_factor,
where flag_factor = 1.0 green · 0.6 unknown · 0.5 neutral · 0.4 yellow · 0.3 procedural · 0 red.
Red-flagged cases excluded — their interpretation has been undermined and shouldn't lead the list.
Top 20 returned per statute. Open any case for full G Cite, trajectory, and per-circuit treatment.