How to Turn a Case Brief Into Flashcards

Only three things reliably earn a place: the rule of law, the trigger that tells you the rule applies, and the case name as a label for that rule. Your issue statement, procedural posture, and the two-paragraph facts section were there to help you follow class discussion. They do not appear on the exam.

Test each line of the brief with one question: could my professor build a hypothetical that turns on this? The rule from International Shoe Co. v. Washington, 326 U.S. 310 (1945) — a defendant must have minimum contacts with the forum such that suit does not offend traditional notions of fair play and substantial justice — passes. The fact that International Shoe sold shoes through salesmen who displayed samples in Washington does not, unless your professor uses those facts as a recurring reference point.

The exception is the case where facts are the rule. Palsgraf v. Long Island Railroad Co., 248 N.Y. 339 (1928) is unusable as an abstract proposition; you need the scales, the fireworks, and the distance to know what Cardozo meant by duty running only to foreseeable plaintiffs within the zone of danger. For those cases, write a compressed fact card — one sentence, not five.

The failure mode is a card that reads "Erie Railroad Co. v. Tompkins" on the front and four hundred words on the back. You cannot grade yourself on it, so you flip it, feel a flash of familiarity, and mark it correct. Familiarity is not retrieval.

Break each case into atomic cards, each with a single correct answer you can say out loud in under fifteen seconds. From Erie, 304 U.S. 64 (1938), you might build: (1) What did Erie overrule? (2) In diversity cases, what law governs substantive issues? (3) What constitutional and policy concerns drove the holding? Three cards, three distinct retrievals.

Use cloze deletion for multi-element rules. Write the rule once, then blank out one element per card: "Under Miranda v. Arizona, 384 U.S. 436 (1966), warnings are required when a suspect is subject to {{c1::custodial}} {{c2::interrogation}}." You now drill both prongs separately, which matters because the exam issue is almost always whether one prong is satisfied.

Build reverse cards deliberately. Front: "Which case established that a defendant's unilateral placement of a product into the stream of commerce, without more, does not create minimum contacts?" Back: World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286 (1980). Recognizing a case name is easy; producing it from the doctrine is what you need under time pressure.

Three to six for a workhorse case, one or two for a case your professor mentioned in passing. A twelve-card deck for a single case usually means you have converted your brief verbatim instead of editing it.

Add one card type most students skip: the distinction card. Personal jurisdiction is not learned by memorizing World-Wide Volkswagen, Burger King Corp. v. Rudzewicz, 471 U.S. 462 (1985), and Asahi Metal Industry Co. v. Superior Court separately. It is learned by answering "What fact distinguishes Burger King from World-Wide Volkswagen on purposeful availment?" That card forces the comparison your exam answer will require.

Also write the dissent or concurrence only when your professor spent real class time on it or when it later became law. Otherwise it is noise competing for review minutes with rules you will actually be tested on.

Paste your own brief into the model and ask it to convert that text. Do not ask a model to tell you what a case holds from memory — that is where fabricated holdings, invented dissents, and wrong reporter citations come from. When the input is your brief, the model is doing formatting and decomposition, which it does well, rather than recall, which it does unreliably.

A prompt that works: "Here is my case brief. Generate 5 flashcards in tab-separated format for Anki import. Column 1 is the front, column 2 is the back. Include one card asking for the rule given the case name, one asking for the case name given the rule, one cloze-style card on the elements, and one comparing this case to [prior case]. Use only information present in my brief. Do not add outside facts."

Then read every card before importing. Two errors recur: the model converts your shorthand into a confident but subtly wrong rule statement, and it smooths over a limitation your professor emphasized (a jurisdictional split, a footnote, a majority-versus-plurality distinction). Editing ten AI-drafted cards takes about as long as writing six yourself, and the editing pass is itself a review session. If you never touch the cards, you get the worst of both worlds: an unfamiliar deck you did not encode.

One deck per course, with tags rather than subdecks for topics. In Anki, tag by doctrinal area ("personal-jurisdiction," "erie," "pleading") and by source ("case," "black-letter," "prof-hypo"). Tags let you build a filtered deck for the two topics you are weak on the week before finals; nested subdecks make that harder.

Turn on Anki's FSRS scheduler and set a realistic daily review cap. A 1L who adds cards after every reading and reviews fifteen minutes a day will hold roughly a semester's worth of doctrine. A student who makes 600 cards in reading week is doing massed practice with extra steps.

Quizlet and Brainscape are fine substitutes if the friction of Anki keeps you from starting. Quizlet's learn mode is weaker at long-interval scheduling but its shared decks and phone interface are better. The tool matters far less than whether the cards are yours and whether you review daily.

On citations: put the case name and year on the card. Add the full reporter citation only if your professor requires citations on exams, which most do not. If you do include them, follow Bluebook Rule 10 for the full form and Rule 10.9 for short forms — "Int'l Shoe, 326 U.S. at 316" — because that is the form you would use in a memo anyway.

Yes, substantially. For a closed-book exam, case-name-to-rule cards are essential because you must reproduce the rule from memory before you can apply it. For an open-book or open-outline exam, memorizing rule text is lower value; what pays is knowing which doctrine a fact pattern triggers, since you cannot look up an issue you failed to spot.

For open-book courses, weight your deck toward trigger cards: "A plaintiff sues a foreign manufacturer in the state where the product injured her but the manufacturer sells nationally through a distributor. What doctrine and which line of cases?" The answer is a doctrinal address, not a rule recitation.

For a multiple-choice component — the MPRE, a 1L midterm, or the MBE later — add distractor-aware cards on the distinctions the testers love: negligence versus recklessness, general versus specific jurisdiction, common law versus Model Penal Code mens rea.

Four habits, all common. Cards copied from a commercial deck you never edited, so the phrasing does not match your professor's. Cards with backs longer than a tweet, which you cannot self-grade. Decks built entirely in the last two weeks, which removes the spacing that makes the method work. And decks that contain only cases, with no cards for the black-letter rules, statutory text, and multi-factor tests your professor actually gave you in lecture.

Flashcards also cannot do everything. They build the recall layer — rules, elements, case labels. They do not build the application layer, which comes only from writing practice answers under time. If you have limited hours in the last two weeks, one timed practice exam beats three hours of review. Cards are the daily maintenance work that makes practice exams productive.