23 Geo Court

Method and limits. MSI v0.2-nyc-2026-09 weights twenty-two public signals by how much each one changed the recorded four-month arm’s-length sale rate across all 636,000 privately owned one- to three-family lots in the five boroughs, checked on a holdout the model never saw (the top 0.5% of homes by score sold at 5.7x the citywide rate, the top 1% at 4.1x). Owner mailing addresses are not public in New York City, so absentee and out-of-state ownership are not part of this model.: new-building permit +27, trade permit +18, lien-sale notice (water/sewer) +13, lien-sale notice (property tax) +12, demolition permit +12, bought since 2024 +9, heat / hot-water violation +9, unsanitary / dirty / rodent complaint +8, hpd class b violation +6, construction / plumbing complaint +6, late-stage lien notice (30- or 10-day) +6, general-construction permit +4, star exemption (owner-occupied) +4, maintenance complaint +3, hpd class c violation +2, two or three units -3, structural violation -4, vermin / garbage violation -7, sidewalk shed / scaffold / fence permit -9, solar permit -12, structural / safety complaint -12, permit valued at $50,000 or more -14. The "in 100 homes" figure is the model's fitted probability for this combination of signals. It describes homes like this one in aggregate; it is not a claim about this owner's intentions, and it is not LienSuite's paid deal score.