Why does spaced repetition help with law school material at all?
Law school gives you an enormous volume of discrete, retrievable facts sitting underneath the analytical work: the four elements of adverse possession, the FRE 803 exceptions and their numbers, the Erie doctrine's steps, what Rule 12(b)(6) requires after Twombly and Iqbal, the difference between express and implied preemption. You cannot reason your way to these. You either have them or you are reconstructing them badly during a three-hour exam.
Anki schedules each card to reappear just before you would forget it, and it makes you retrieve the answer rather than reread it. Both features are well supported in cognitive psychology — the spacing effect and the testing effect. The practical payoff is that a rule you carded in September is still there in December without a panicked reread of the whole outline, and still there in July when you sit for the bar.
The limit is equally real. A closed-universe recall card cannot teach you to see that a fact pattern presents a Palsgraf foreseeability problem rather than a proximate cause problem, or to decide which of three plausible characterizations of a contract term to argue. That skill comes only from writing practice answers under time and comparing them to model answers. Anki frees up time for that work. It does not substitute for it.
What should you actually put on cards?
Card things that have a single correct answer and that you will be penalized for not knowing cold. Leave off anything whose value comes from arguing it.
Good candidates: elements lists, multi-factor tests, statutory and rule numbers, definitions with terms of art, exceptions and exceptions-to-exceptions, burdens and standards of proof, and jurisdictional splits your professor flagged. Evidence, Civil Procedure, and Criminal Procedure reward this heavily because the rules are numbered and textual. Constitutional Law and Contracts reward it less, because the doctrine is more open-textured.
Poor candidates: case facts in detail, policy arguments, your professor's classroom tangents, and anything you would express in a paragraph. If a card takes more than about eight seconds to answer, it is really three cards or it is an outline entry pretending to be a card.
Case names are a middle case. Card the ones that function as shorthand for a rule — International Shoe Co. v. Washington for minimum contacts, Erie R.R. Co. v. Tompkins for the vertical choice-of-law rule, Palsgraf for duty and foreseeability. Do not card the procedural history of every case in the Civil Procedure reader. Ask whether you would ever write the name in an exam answer; if not, skip it.
- Card: FRE 801(c)'s two-part definition of hearsay, split into the offered-for-truth prong and the out-of-court-statement prong.
- Card: the four elements of adverse possession as your professor states them, one cloze per element.
- Card: what a business records foundation requires under FRE 803(6).
- Skip: whether Iqbal was correctly decided.
- Skip: the three-paragraph balancing discussion your professor gave on dormant Commerce Clause.
How do you write a law card that actually works?
Use cloze deletion for anything list-shaped, and delete one item at a time rather than the whole list. A card reading "Adverse possession requires possession that is {{c1::actual}}, {{c2::open and notorious}}, {{c3::exclusive}}, {{c4::hostile}}, and {{c5::continuous for the statutory period}}" generates five separate retrievals. A single card asking "List the elements of adverse possession" generates one vague retrieval you will pass while actually knowing four of five.
Front-load the context. "Under the FRE, what is the standard for character evidence?" is ambiguous across 404, 405, 608, and 609. Write "FRE 404(b) — prior bad acts are inadmissible to show {{c1::propensity}} but admissible for {{c2::motive, opportunity, intent, preparation, plan, knowledge, identity, absence of mistake}}." The tag on the front should tell you which corner of the doctrine you are in.
Write the card in your professor's formulation, not the casebook's and not a commercial outline's. If your Torts professor uses a three-element negligence framing with breach folded into duty, card that. You are being graded by that person.
Tag by course and by outline heading — "Evidence::Hearsay::Exceptions::803" — so that you can filter to a single topic the night before you write a practice essay on it. Deck-per-course plus hierarchical tags is enough structure. Do not spend an hour building a taxonomy.
When should you start, and how long should reviews take?
Make cards only after you understand the rule, which usually means after you have briefed the cases for a topic and written that section of your outline. Carding first turns Anki into flashcard-shaped confusion; you will memorize a formulation you cannot use. A practical rhythm is to outline a topic on the weekend and card it the same day, roughly one to two weeks behind class.
Aim for 15-25 minutes of reviews per day across all courses. If reviews are running past 40 minutes, you have too many cards, not too little time. Suspend or delete aggressively — a card you have failed six times is a card that is badly written, not a card you need more of.
Volume guidance: 300-600 cards per doctrinal course over a semester is a workable range for most students. Two thousand cards in Civil Procedure means you carded your outline verbatim, and you will abandon the deck in November, which is exactly when it should be paying off.
Do reviews daily, including weekends. Spaced repetition breaks down when you skip four days and return to a 400-card backlog. Ten minutes on a bad day beats zero.
Should you use pre-made decks?
Generally no for doctrinal courses, and sometimes yes for the bar. Making the card is a large share of the learning — you are deciding what the rule is, in whose words, and at what level of generality. A downloaded deck skips that and gives you someone else's professor's formulation, often with errors you cannot detect because you do not know the material yet.
The narrow exceptions: decks of pure lookup material where formulation does not vary. Bluebook Table T6 case-name abbreviations and Table T13 periodical abbreviations are the clearest example — if you are doing a journal write-on or a citation exam, carding those tables saves real time and there is only one right answer. The same logic applies to FRE rule numbers.
For bar prep, most commercial courses now include their own spaced-repetition tools, and their content is calibrated to the tested material. If you use Anki instead, build it from your bar course's outlines during the course, not from your 1L decks. The tested rules are the majority rule and the MBE-specific formulation, which often differ from what your professor taught.
How should you configure Anki for this?
Turn on FSRS, the scheduler built into recent Anki versions, and set a desired retention around 0.90. It will schedule more accurately than the legacy SM-2 algorithm and will reduce your daily load for the same retention. Leave the rest of the algorithm settings alone.
Raise your new-cards-per-day cap deliberately rather than defaulting to 20. If you card a topic on Sunday, you want those cards introduced over the following week, not trickled over a month. Set a per-deck new limit that clears your backlog within a week of creating it.
Anki desktop is free on Windows, Mac, and Linux. AnkiDroid is free on Android. The iOS app is a one-time paid purchase, and the purchase funds the project. AnkiWeb sync is free and is how you review between classes, which is where most of your reviews will actually happen.
Use the built-in image occlusion for anything spatial — a Civil Procedure flowchart for personal jurisdiction, a hearsay decision tree, a UCC 2-207 battle-of-the-forms chart. Hiding one box at a time from a chart you already understand is one of the highest-yield card types in law school.
Where does Anki stop helping?
Your exam grade is mostly a function of issue spotting, organization, and application. Anki touches none of those directly. A student with a perfect deck and no timed practice exams will underperform a student with a mediocre deck and six timed practice exams under the same professor.
Budget accordingly. In the last four weeks before finals, the marginal hour should go to a practice exam or to comparing your answer against a model, not to adding cards. Keep doing reviews — they are cheap and they protect the retention you built — but stop card production around the time reading period starts.
Watch for the productivity illusion. Reviews feel like work, produce a satisfying streak counter, and generate visible statistics. Writing a bad practice essay and reading a model answer feels awful and produces nothing to look at. The second activity is worth more. Use Anki precisely because it makes the memorization cheap enough that you have time for the part that hurts.