A real case loses five ways. Nobody verifies four.
You cite a real case, the quote is perfect, and you still lose the motion, because it was narrowed last year, or it favors the other side on the point that matters, or the same judge already rejected that argument. No citation checker catches that. You catch it at 11 PM, or opposing counsel catches it for you in open court.
Each one sails past a citation checker’s green light. Each one loses a motion.
Wrong proposition
The case is real, but it does not stand for what you say it does.
Dicta as holding
You are leaning on language that was not the ruling.
Wrong direction
The authority actually favors the other side on the point that matters.
Narrowed rule
The holding was cut back after the case came down.
Rejected argument
The same court already refused this exact argument.
Five citations. All real. All correctly quoted. All green on the citator.
The book is real. Your proposition is not.
The case exists, the quote is accurate, and it simply does not stand for the thing you filed it for. A citator was never looking at that. It confirms the document; it has no field for what you claimed the document says. Substance verification checks the proposition against the source lines, and stops it when the holding is somewhere else.
Real case. Perfect quote. It just doesn’t stand for the thing you filed it for.
You are leaning on a sentence the court was musing out loud.
Opposing counsel lives for this. A proposition-level record distinguishes what the court decided from what it observed, established from applied, and refuses to let an aside carry a motion.
You’re leaning on a sentence the court was musing out loud.
This one is real, it is good law, and it is working for them.
Party favorability is a stored field: which side a holding helps, calibrated to future litigation rather than to who won the source case. You just cited your opponent’s best case, and the system says so before they do.
This one’s real, it’s good law, and it’s working for them.
Real case. Dead bridge. Bad argument.
The holding was cut back after the case came down. A document-level flag may still be green; the case was not overruled, it was hollowed out. Authority health at the proposition level tracks the narrowing, the prerequisites added, the departments that declined to follow. Your citator flags the case. We flag the sentence.
Your citator flags the case. We flag the sentence.
Same argument, same court, already lost.
Standard search finds holdings. Rejected arguments are indexed from the losing side’s perspective, so “has anyone tried this and failed?” has an answer. We show you the thing that is not in your draft.
Same argument, same court, already lost. We show you the thing that isn’t in your draft.
The green check says the case exists. It never said it says what you need it to.
A document-level flag can’t tell you if the line you cited drew the negative treatment. We can.
Citators are a manual case law verification step bolted onto a manual research process. They flag documents, not holdings; they carry no directional stance; they model no binding-authority chains. The empirical record on their agreement with one another is poor enough that a case can be treated three different ways by three products. When authority health is a pre-computed, continuously updated property of every proposition, the separate validation step dissolves into the architecture. Not because validation is unimportant. Because the architecture made a separate validation step unnecessary.
The rigorous version, with the empirical record, is in Good Law for What?
A checker inspects what reached the page. It cannot see what should have been there and never entered the analysis. That is why the MSJ Qualifier builds the summary judgment analysis before anything is drafted.
What a document-level flag tells you
- A citing decision exists
- It was coded as negative, positive or neutral at the case level
- The case, as a whole, has a status
What it cannot tell you
- Whether the line you cited is the line that drew the treatment
- Whom the proposition favors
- Whether it governs at your posture in your forum
- Whether the same court already rejected your argument
Three California authorities that still read as good law, and aren’t.
Anti-SLAPP: language courts still cite from before the 2016 reframing
The special-motion-to-strike standard was recast; briefs still quote the pre-2016 formulation and courts still cite it. A proposition-level record marks the superseded line.
Motion to compel further: the 45-day and separate-statement trap
The timing rule and the separate-statement requirement are where motions to compel further responses die, not on the merits.
Summary judgment: the separate statement and the objection trap
Objections not made in the required form are waived; facts not in the separate statement do not exist for the motion.
Demurrer: the meet-and-confer declaration
The declaration is a condition of the motion. Its absence is a procedural defect that has nothing to do with whether the complaint states a claim.
Jurisdiction variants are published for the courts the product covers. Not legal advice.
Five ways to lose. Only one system in the room saw it.
Bring a motion you’re nervous about and watch the five checks run on your own authorities.