How it works, and how to get the most from it.

One scan, every signal

What. Connect Lichess and press Scan. The app reads your most recent 100 games of each time control from the last three months (further back if you've played fewer), with Lichess's analysis or with Stockfish on your device, plus your Lichess puzzle history.

Why. Each error is weighed, not just counted. Errors you repeat count most. A slower game counts more than a faster one (a correspondence error says the most about your understanding), an error with seconds left counts less, and a long think that ended badly counts more. Recent games, errors that changed the result and errors early in the game weigh more too. The app also names the tactic behind each error: the one you missed, and the one your move walked into.

Get the most. The first scan takes a while. Keep the screen open; if you stop, running it again picks up where it left off.

A few decks at a time

What. The plan proposes decks in three steps. Start here: your repeat errors, your other errors ranked by what they cost, "Now what?" if quiet positions trip you up, and at most one proactive deck. Next: what to add once your reviews feel easy. Library: anything else, whenever you like.

Why. It's easy to take on too much. Every new card comes back several times over the following weeks, so a big first day becomes a pile of reviews later. You choose a daily number of new cards for all decks together: Gentle (6), Steady (12) or Intense (24).

Get the most. Start gentle. When your daily reviews feel easy, raise the number or add the next deck from your plan. Scan again about once a month: new errors join your decks, your progress stays, and you see what changed since the last scan (your error rate, the positions you used to repeat, the tactics you missed).

Is it carrying over?

What. After every scan, the app looks for positions you've drilled that came up in your games since, and whether you played the trained move.

Why. That's the real test. A trained move played in a real game counts as a successful review, so the card waits longer; an error there brings it back now. Insights keeps the tally.

Get the most. Play, then scan: the cards you keep getting right in real games are the ones you've truly learned.

Repeat errors

What. Positions you misplayed in more than one game, the same position or nearly the same one, usually an opening line you keep meeting.

Why. These are the surest wins in the app: positions you've proven you'll reach again, where you've proven you go wrong. Fix one and it pays off every time it comes up.

Get the most. Study these first. When one feels automatic, notice it in your next games: that's the point.

Your mistakes

What. Every inaccuracy, mistake or blunder (you choose the minimum), the position before it, and the best move. In correspondence games every inaccuracy counts: you had days to think.

Why. They come in order of what they cost you, so a few minutes a day goes to the errors that matter. Obvious slips made with no time left can be skipped: one careful look would have prevented them, so drilling them teaches little.

Get the most. After a card, press "Why it fails" and "Why it's best" to watch both lines play out. If an idea still isn't clear, open the analysis board.

Now what?

What. The moment after the opening when everyone is developed, nothing is forcing, and the plan isn't obvious. Taken from your games, especially where you then went wrong, and from master games that reached a similar position.

Why. This is where a lot of clock goes. There's rarely one right answer, so any move Stockfish rates about as good counts; afterwards you see the strongest options and the plan in plain words.

Get the most. Before moving, name the plan in a few words (a pawn break, a piece to improve). Positions a piece or two apart come later if the same move is best, and right after if it isn't, so you learn to tell them apart.

Puzzles you failed

What. Every Lichess puzzle you got wrong and haven't solved since.

Why. Each comes back once enough time has passed that you have to see the idea again, not remember the answer.

Get the most. In a session, sort by easiest or hardest first to match your focus that day.

Insights

What. Where and why you lose points: errors per time control, what went wrong, the clock and thinking time behind them, your blind spots by type of position, the openings you don't know well, and the habits behind quiet errors.

Why. Patterns you can't see from one game at a time. Each one has a drill.

Get the most. Look after a scan, then drill the top one or two.

Checkmate patterns

What. About thirty named mates (back rank, smothered, Anastasia's, Arabian, Boden's and more) in the order Lichess Practice teaches them. Each starts with Lichess Practice's own positions, then real puzzles of the same pattern from easy to hard.

Why. Lichess Practice shows you each pattern a few times; spaced repetition makes it second nature. The patterns come in as a spiral: the first few, then harder examples of those as new ones join.

Get the most. A new card names its pattern. After that, try to name it yourself before you move.

A tactic, step by step

What. Thirty puzzles of one theme (or as many as you like), from easiest to hardest: zugzwang, deflection, interference, clearance, the quiet move, any Lichess theme.

Why. Each card is a little harder than the last, so the idea builds. Your plan suggests the themes you miss most in your games and fail most in puzzles, compared with what their ratings predict.

Get the most. One theme at a time. When the set feels easy, start the next.

Openings without the slog

What. A repertoire from the Lichess opening explorer (one move for you, every reply common at your level), or any Lichess study. Each move you're drilled on is checked by Stockfish.

Why. The lines you'll meet most come first. A new line plays itself once so you can watch, then you play it from memory. Forgot one? Press "Watch the line" first; it then comes back soon.

Get the most. At the end of a line, open the analysis board to see what's played next, and add cards from there only if you want them. Where two of your repertoires disagree, the app shows which move Stockfish prefers and teaches that line first.

Traps, three ways

What. Seventeen classic opening traps, drilled as dodge it, punish it, and set it. Each has a short lesson with how risky it is to set.

Why. Dodging comes first: it saves the most points. The ones you've fallen for in your games come first.

Get the most. Setting the risky ones is optional and kept in its own deck.

Fundamentals path

What. An ordered path through the patterns coaches teach first: board safety, the core tactics, forcing combinations, mating patterns, calculation, endgames. The order draws on Dan Heisman, the Steps Method and Lichess Practice.

Why. If you're newer to chess, this is the foundation everything else stands on.

Get the most. One step at a time, easier puzzles first: recognising a pattern instantly is worth more than grinding hard ones once.

Tactic or not?

What. Timed sets mixing real puzzles with quiet positions. First decide whether there's a tactic at all, then play your move.

Why. In a puzzle you know there's a shot. In a game you don't. This trains the decision you face every move.

Get the most. Use it as a warm-up before playing. Misses go into your spaced repetition.

Rating

After each card you rate it Again, Hard, Good or Easy. The app suggests the rating from whether you found the move and how your time compares with your own past times on that card: faster than usual is Easy, slower is Hard. Press Space to accept. The FSRS scheduler (the one in Anki) then brings the card back when your recall would drop to about 90%.

When you're stuck

Hint marks the piece to move (the card counts as Hard at best); Show move reveals it. Skip sends a card to the end of the session without touching its schedule, and Previous card undoes your last rating. Pause hides the board so the clock stays honest; it also pauses by itself when you switch away. Forgot to pause? "Time wrong? Adjust it" fixes the time, or leaves it out of your usual pace.

Other good moves

Stockfish runs on your device and accepts any move about as good as the card's (except in opening lines, which follow your repertoire). After you find the move, the app shows up to two others that are just as good, within a tenth of a pawn.

Reading the arrows

Engine arrows are the same everywhere: blue is the best move, orange the second, violet the third. The wider the arrow, the closer it is to the best; clearly worse moves get no arrow. Red marks a move that fails.

Groups and folders

Decks sit in two groups, Corrective and Proactive, with folders inside (Your weakest tactics, Checkmates, Openings & traps). Study everything at once, one group, one folder, or one deck. A deck's switch "In Study everything" leaves it out of the bigger sessions when you only want it on its own.

Study load

How much time you want to spend changes. Study load shows what the next month will ask of you, in minutes a day, and lets you take on more or less in a few taps: every deck at once in steps of 10%, evenly or by the order you put your decks in (the ones at the top are spared when you cut back), or deck by deck. You see the effect before you save. Fewer new cards lightens the load for good; a review limit works at once but postpones the rest.

Study groups

For a specific purpose (say, a warm-up before a blitz session), combine any decks into a study group on the Decks page. Most people never need this.

The archive

Want to start fresh, or try a new plan on a clean slate? Archive your decks. They're kept: browse them, study one on its own, or bring them all back. While they're in the archive, a new card for a position you already know keeps what you'd learned.

Boards and pieces

Six boards, including Forest & crimson and Leaf & bark, and a piece set of our own coloured to match each of the custom boards. Settings has both.

On your phone

Open the app in your phone's browser and add it to your Home Screen. On iPhone: Share, then Add to Home Screen. On Android, in Chrome: the menu (three dots), then Install app (or Add to Home screen). It then opens like an app, and the browser keeps its data safe.

Offline

After you've opened it once with a connection, studying works without one, including Stockfish judging your moves. Scanning games, the opening explorer and new puzzles need the internet.

Sync

Your decks and progress sync between your devices through your Lichess login, automatically: when the app opens, every few minutes, and when you leave it.

Anki

Any deck exports to Anki, AnkiDroid, AnkiMobile or MintDeck as image cards. The timing, the engine and playing out lines only exist in Chess Fluency.