Methodology
How CaseParity builds its reports
Executive summary
CaseParity organizes the public record of how criminal charges have historically been resolved. Statutes are validated against multiple sources and selected by the offense date.
Read the full methodology ↓
CaseParity Public Records Transparency Reports summarize what already happened in publicly reported criminal cases, drawn from government records that are public by law. They are published statistics, not legal advice — they describe patterns across many past cases; they do not evaluate, predict, or recommend anything about any individual case.
Where the data comes from
We use official, publicly released criminal-justice records — the kind maintained by state agencies and courts and made available for public transparency. We do not collect any information about you or your case to produce a report.
How we check it
Public records are imperfect. We cross-check the legal references in our reports against more than one independent source. Where sources disagree, we investigate the discrepancy and apply documented reconciliation rules before publication.
Version-aware
The law changes over time. Statutes are amended. A report reflects the version of the law that was in effect for the time period a case involves, not just today’s text — so the comparison is to cases decided under the same law.
What we show, and how
Statistics are presented as rates with their totals stated plainly (“of N cases like this, X%”), with a measure of statistical uncertainty, and in plain language. “Cases like this” means cases charged under the same statute (and subsection where the data supports it) — the comparison is grouped by the charge, not by the facts of any case.
Sparse data is shown, not hidden
Every published rate must clear a minimum number of cases. When a county doesn’t have enough on its own, the figure isn’t dropped — it rolls up to the next broader geography (county → judicial circuit or district → statewide) until it clears the threshold, labeled with the level it was computed at. A figure is shown as unavailable only when even the broadest level has too few cases. We never fill a gap with an estimated or imputed number.
What we’re honest about
No public dataset is complete. Some categories of records are sealed, removed, or not reported uniformly across counties. In many public datasets the stages of a case aren’t linked together, so reports show patterns at each stage from its own records rather than tracking a single case’s outcome through time. Reports describe these limits where they apply rather than papering over them.
What a report is not
A CaseParity report cannot tell you what will happen in your case, what you should do, or whether a particular outcome applies to you. Those are questions for a licensed attorney. Our job is to make the public record legible — accurately, with its limits stated.
Technical detail · what we do not infer
The public record supports statements about what happened across many past cases. It does not support inferences about an individual case. From our data we do not infer:
Publication floor. A rate is published only when it is computed from at least 30 cases. Below that, the figure rolls up to the next broader geography (county → judicial circuit / district → statewide); only when fewer than 30 cases exist even at the broadest level is the figure shown as unavailable, never imputed.
Finer per-cell thresholds. Some detailed breakdowns use a higher bar before a single cell is displayed: the Florida county-by-year disposition cell and the Cook County per-judge / per-courthouse cell each require 100 cases; the Virginia departure-reason bucket requires 10. A cell below its threshold merges into the broader aggregate, and the merge is disclosed.
Uncertainty. Where the case count supports it, rates are reported with a 95% confidence interval, so the reader can see the precision of the estimate rather than a bare percentage.
Validation sources. Statute references are cross-checked against more than one independent source — the originating state legislature’s official statute text (e.g. flsenate.gov for Florida), OpenStates, and public legislative datasets — with documented reconciliation rules where sources disagree.
Coverage
Currently incorporating publicly available criminal-court records from Florida (statewide, all 67 counties; documented dispositions 2018–2026, refreshed monthly), Cook County, Illinois (2010 through December 30, 2024 — a fixed extract), Virginia (Virginia Criminal Sentencing Commission, FY2020–FY2024), and Harris County, Texas (records back to 1990, refreshed monthly; additional TX counties in progress). Coverage reflects what each jurisdiction publishes and when; some records are sealed, removed, or not reported uniformly, so completeness varies by jurisdiction, county, and time period.
See the Coverage page for source detail and current status per jurisdiction →
The Circuit Court of Cook County describes itself as “one of the nation’s largest unified court systems,” served by nearly 400 judges across 17 divisions and districts. Harris County’s District Clerk’s Office is the record-keeper for over 90 courts and receives roughly 300,000 criminal filings a year.
Version-awareness, demonstrated
A worked example: Fla. Stat. § 893.13 was amended by Chapter 2024-176, Laws of Florida, effective July 1, 2024. A report for a case dated June 15, 2024 cites the statute as it read before that amendment; a report for a case dated July 15, 2024 cites the version that took effect July 1. We also track which part of a statute an amendment actually touched: Chapter 2024-176 amended only a cross-reference in subsection (1)(h), so a report on a charge under subsection (6)(a) does not imply that the charged subsection changed in 2024. Statute text and effective dates are sourced from the state legislature’s published statute text and session laws, and cross-checked against independent legislative datasets.