Blog

Notes on crime data.

How to read a score, where the data comes from, and what to do with it. Written by the SpotCrime team.

• September 13, 2026 · 3 min read

  How to look up crime data by ZIP code, and what it can and can't tell you

  Where ZIP-level crime data actually comes from, the three questions to ask before trusting a number, and a five-minute method for reading any ZIP.

• September 12, 2026 · 3 min read

  ZIP code vs. ZCTA: why your ZIP's boundary isn't what you think

  USPS ZIP codes are delivery routes, not areas. Crime data uses Census ZCTAs instead. Here is the difference, why it matters, and how to avoid the classic mistakes.

• September 11, 2026 · 3 min read

  How to read a crime score: percentile, rank, level, and trend

  The four numbers on every CrimeScore page, what each one means, and the mistakes people make with them.

• September 10, 2026 · 3 min read

  Crime data API vs. crime score API: which one you actually need

  Raw incident feeds and scored summaries solve different problems. A guide for product teams deciding which to integrate, with the questions to ask a vendor.

• September 9, 2026 · 3 min read

  Giving an AI assistant crime data: what MCP changes

  Why bolting a crime dataset onto a chatbot usually goes wrong, and how a read-only MCP connector lets Claude and other assistants answer neighborhood questions with real, versioned numbers.

• September 8, 2026 · 3 min read

  Baltimore, June to July 2026: what moved, and why rank and percentile disagreed

  A month-over-month read of all 30 scored Baltimore ZIP codes: which fell most, which held, and a quirk of relative scoring that anyone using the numbers should understand.

• September 7, 2026 · 3 min read

  Why two sources give different crime scores for the same ZIP

  Same ZIP code, different numbers: the six choices that make crime ratings disagree, and how to tell which one is answering your question.