Anki is a free, open-source spaced-repetition program. Its scheduler implements the same principles described in the memory literature — expanding review intervals derived from the SM-2 algorithm and its successors — and it is effective. Roger Craig prepared for his 2010 Jeopardy! appearances with it, setting what was then the highest single-day total in the programme's history. For anyone revising their own material, Anki is the appropriate tool and this page does not argue otherwise.
What follows concerns a narrower question: whether the scheduler is the binding constraint for someone whose goal is general knowledge of the kind quizzes ask about. The evidence suggests it is not.
Where deck-building fails, and why it is not a scheduling failure
Three failure modes are documented at length in Anki's own user community. None of them is attributable to the algorithm.
Production cost. A general-knowledge deck of useful coverage requires thousands of cards. Reported deck sizes run from roughly two thousand to twenty-five thousand items. Writing them is a sustained transcription task whose duration typically exceeds the period over which the writer remains motivated.
Shared decks degrade predictably. The large downloadable trivia decks are unranked extracts, weighted towards United States material, fixed at the date of scraping, and generally stripped of audio and images. Curated decks are better but are maintained by individual volunteers, and maintenance ends when the volunteer's circumstances change.
Selection is unspecified. A blank deck carries no information about which items are worth adding. This is the substantive problem. Users regularly report large decks accompanied by limited improvement in quiz performance — an outcome consistent with correct scheduling applied to a poorly selected item set. As set out with citations in the guide to getting better at pub quizzes, item selection is the larger term: quiz content concentrates in a stable, countable set that recurs in measurable proportions, and revision weighted by that frequency returns more per hour than even coverage. A scheduler cannot correct a selection error, however well specified it is.
What differs in QuizMaxx
The scheduling does not differ. QuizMaxx uses the SM-2 family, so the memory science underlying it is the same. The differences are in what is scheduled and how it is presented.
The items already exist, written as questions in the form a quizmaster would ask them rather than as front-and-back prompts.
The items are ordered. Frequency of occurrence was measured across archives of real quiz questions, and previously unseen material is introduced in that order, so early study is concentrated on the items most likely to be asked.
Answers are typed and marked with tolerance for form. A misspelling of the right answer is accepted; naming a different member of the same set is not. Production rather than selection is the operation a quiz tests.
Some material cannot be represented as a flashcard. Music rounds play the recording, picture rounds show the image, and there are anagram, connection and weekly-news rounds.
One clarification on what is and is not claimed: the facts are not proprietary and are not claimed to be. The work is selection, ordering and verification — which is precisely the work a blank deck leaves to the user.
When Anki remains the better choice
Anki is the better tool where the material is the user's own: medical training, languages, law, or any syllabus set by an examining body. A curated corpus cannot substitute for a blank slate in those cases. The same applies where fine control of scheduling parameters is wanted, where the plugin ecosystem is in use, or where the subject is one nobody has curated.
QuizMaxx covers one domain: general knowledge of the kind quizzes ask about, with a British centre of gravity. Within that domain the argument for it is that a corpus which does not have to be written is one that is still being reviewed several months later. Outside it, the blank slate is the point.
Availability
QuizMaxx runs in the browser. There is no installation and no account required before use. Open QuizMaxx. The evidence for the underlying method is set out in the guide to getting better at pub quizzes and the general-knowledge case; the 364 short essays covering the same material are free and need no software at all.
The claim here is a narrow one: not that a curated corpus beats a well-built Anki deck, but that building and selecting such a deck is the step at which most attempts stop.