The case record

No chapter in the book names a case. This is where the matters described in those chapters are identified, so a reader who wants the underlying material can find it.

The book describes 11 matters. Of those, 9 are identified below from the order or opinion itself, read in full. The book's appendix lists only those, because it identifies nothing on the authority of secondary reporting. The rest are here, marked pending, with no facts stated beyond what the court's own records confirm, until the primary document has been read. Every row carries the date it was last checked.

Matters reported in the press but not yet read in the original, and matters outside the book's subject, are recorded separately in AI in the Courtroom. Every entry there states whether the primary document was read, so the two tiers never mix.

The index

Index last reviewed 2026-10-08.

The matters

Each matter is stated in the same order: where the book describes it, the court and docket, what the professional did, what the tribunal did, and where it stands. Where a matter also has a fuller entry in AI in the Courtroom, the link is given.

How entries are written

Each entry states the caption, the court, the date, what the professional did, and what the tribunal did. No entry characterizes a holding. That discipline is the book's and it applies here with more force, not less, because this page will be read by people who have not read the surrounding argument.

Where a matter is on appeal, subject to an unresolved objection, or has been distinguished or superseded, the entry says so and gives the date. An entry with a stale check date is worse than no entry, so every row carries one.

What this page is not

It is not a research database, it does not purport to be complete beyond the matters the book describes, and nothing on it is legal advice. Whether any of it bears on your matter is a question for counsel.

Reading a new decision in this area

Four things are worth extracting, and the rest can be ignored. What the professional actually did, in sequence. What the tribunal did about it — the disposition, not the reasoning as characterized by whoever is describing it. Which durable principle it engaged. And whether it is final, because appeals and objections are routine and a great deal of commentary is written about orders that do not survive.

What to ignore is the characterization in the commentary. A dozen alerts summarizing one order is one source, not twelve.

What the decided matters actually show

The reported matters divide into a small number of recurring fact patterns, and holding them apart is more useful than counting outcomes. In the largest group a filing contained citations to authorities that do not exist. In a second group the analysis underlying an opinion could not be reproduced, because the method was a conversation rather than a procedure. In a third the material submitted to a system was material the submitter was not free to submit. In a fourth — the smallest and the most instructive — nothing was wrong with the work and the dispute was entirely about what the record showed.

Two observations run across all four, and both are in the book. Consequences have fallen overwhelmingly on the lawyers who signed filings rather than on the experts whose work they relied on, for the straightforward reason that certification obligations attach to the signature. And the objection that succeeds is almost never a system was used; it is that a specific step which required the professional's own judgment was performed somewhere the professional cannot account for.

Why so few entries, and why that is deliberate

An entry appears here when the order or opinion has been read in full. That is slower than reading the alerts, and it is the only method under which the page can say what it says. A reference built from secondary summaries inherits every error in them, and the errors in this area are systematic: orders get described by their most quotable line, unresolved objections get reported as holdings, and a magistrate's recommendation gets reported as a decision.

Entries are removed as well as added. A matter that is vacated, reversed, or withdrawn does not stay on the page with a footnote; it comes off, with the removal noted in the errata, because a reference that only accumulates is a reference that slowly becomes wrong.

How to use an entry

Read the underlying document. The entry exists to help you find it and to tell you whether it is still good, and it is not a substitute for the thing itself. Anyone relying on a one-line description of an order — including this one — is relying on a summary written by someone who does not know what your matter needs from it.

Then read it for conduct rather than for rule. What transfers across courts, jurisdictions, and disciplines is the sequence of things a professional did and the point at which that sequence became indefensible. What does not transfer is the disposition, which turns on procedural posture, the record before that tribunal, and what was actually argued — none of which will match your situation.

And check the date on the row. Every entry carries the date it was last checked against the primary document rather than the date it was added, because those are different facts and only one of them tells you whether the entry is currently true.

Most federal orders can be retrieved from PACER, and many federal and state opinions are free on CourtListener, a public archive of court opinions and dockets run by the Free Law Project. Where an entry here gives a docket number, that is the fastest way to the document itself.

Terms used in the entries

Disposition is what the tribunal did — granted, denied, struck, sanctioned, deferred — stated in the tribunal's own terms rather than paraphrased. Posture is where the matter stands now: final, on appeal, under an unresolved objection, remanded, withdrawn, or settled. Checked is the date a person last opened the primary document and confirmed the row against it.

Where a field is blank, it is genuinely unknown rather than not yet typed. A record of this kind is only useful if the difference between no and not established survives onto the page, and blanks are how that difference is carried.