All Tools

Each card holds two pieces of text, a front carrying the question or prompt and a back carrying the answer, plus a category label. Categories are simply tags you define as you go, and the filter lets you study one subject at a time or view all categories together, which is what makes a single deck workable across several topics.

Study mode tracks a state for every card. A card begins as unseen; once you have looked at it and judged your recall, it becomes either known or reviewing. The three counters above the deck show those totals alongside the overall card count, so progress is visible without any scoring formula behind it.

Shuffle randomises the order of the current selection before a pass. That matters more than it sounds, because learners quickly memorise a fixed sequence and start recalling the next answer from position rather than from the prompt, which flatters the results without improving recall.

The deck is written to the browser's local storage as you edit, and reloaded when you return.

Say you build 30 cards for a biology topic, split into two categories: 18 cards tagged Cells and 12 tagged Genetics. Before you start, the counters read 30 total and 30 unseen, with 0 known and 0 reviewing.

Filter to the Cells category, shuffle and work through all 18. You mark 11 as known and leave 7 in review. The counters now read 11 known, 7 reviewing and 12 unseen, which still totals 30.

Next take the 12 Genetics cards, where the material is newer. You mark 5 known and 7 reviewing. The totals become 16 known, 14 reviewing and 0 unseen.

On a second pass over only the 14 reviewing cards, you get 9 of them right. The deck now stands at 25 known and 5 reviewing, so 25 divided by 30 is 83 per cent of the deck secure after two passes. Those 5 stubborn cards are the ones worth rewriting more simply rather than repeating unchanged.

The quality of a deck depends almost entirely on how the cards are written. Keep one fact per card: a front asking what three things mitosis produces is really three cards, and answering two of them correctly still counts as a failure. Phrase the front as a genuine question rather than a heading, since a prompt reading Mitosis gives you nothing specific to recall. And write answers you can judge quickly, because a paragraph on the back leaves you arguing with yourself about whether you got it right.

Mark honestly. Marking a card known because the answer looked familiar the moment you flipped it is the most common way to waste a study session; recognition is far easier than recall.

On storage, the deck lives in your browser's local storage on this device. It is not synchronised to an account or to another machine, and clearing site data, using a private window or switching browser will leave it empty. Export to JSON or CSV once a deck represents real work, and keep that file somewhere safe.