The clean way to build a repertoire study
Quick answer: Make one Lichess Study for each side of your repertoire, then create a chapter for each major reply you actually face. Keep every chapter shallow at first, usually six to eight moves, and add a plain language note about the plan at the final position. Hide the moves when you practise so you have to recall them. Don't import a giant opening file and call the job done. A smaller study you can navigate and rehearse is far more useful than a perfect move tree you never open.
Lichess Studies can hold an enormous amount of chess, which is both the attraction and the trap. The tool doesn't stop you adding another branch, another model game, or another engine line. Soon the study has forty chapters with names like Chapter 17, and finding the move you forgot takes longer than playing a game.
Build around questions you need answered. What do I play against the Sicilian? What changes when Black delays castling? Where is my first pawn break? If a chapter can't answer one clear question, it probably contains too much.
What a Lichess Study can actually store
I checked the current Studies area and the official feature post before laying out this method. A study saves moves, variations, comments, symbols, and arrows. It can contain many chapters, and each chapter has its own starting position and move tree. You can begin from the initial board, set a position, load a Lichess game, paste a FEN, or import a PGN with comments and variations.
I read the official description of access controls too. A public study can be viewed by anyone, while invited contributors can edit it. A private study limits viewing to invited members. Studies can also use computer evaluation, the opening explorer, and endgame tablebases when the owner allows them. The whole study can be exported as a PGN with its chapters, tags, comments, variations, and arrows. Lichess explains those options in its original Studies guide.
That makes Study closer to a chess notebook than a course. It gives you the board, storage, analysis tools, and sharing. It doesn't decide what belongs in your repertoire, which branches matter at your rating, or when you should review a line again.
The organization is your job. Get that part right and the free tool is remarkably capable. Get it wrong and you've built a searchable attic.
Start with two studies, not twenty openings
Create one study called White Repertoire and one called Black Repertoire. That's enough. Splitting every opening into its own study feels tidy on day one and becomes irritating when lines transpose or you forget where a position lives.
Inside the White study, organize chapters by the opponent's reply to your first move. If you play `1.e4`, the first group might cover `1...e5`, the Sicilian, the French, and the Caro Kann. Inside the Black study, use a group for your answer to `1.e4`, another for `1.d4`, then add rare first moves only when they appear in your games.
This is where restraint pays. You don't need an answer to every legal move. You need an answer to the moves people near your rating actually play. The guide on how many opening lines you need explains why three repertoire slots cover most club games.
Name chapters so search isn't necessary. Sicilian, early Bc4 is useful. Main Line 4 isn't. Include the side you're playing and the decision point, especially if two chapters begin with the same moves.
Keep the first version embarrassingly small. Four chapters you know are a repertoire. Forty imported chapters are a library, and a library doesn't become knowledge just because your name is on the cover.
Make each chapter answer one practical question
A good chapter begins at a familiar position and ends after one decision becomes clear. It doesn't have to contain an entire opening. In fact, shorter chapters are easier to review because you know what each one is meant to teach.
Take a common branch from one of your games. Put the main move on the board, then add the two replies you see most often. Stop when you reach a stable position where general plans matter more than exact moves. For club players that often happens around move eight, sometimes earlier.
Use the opening database for frequency, not for collecting moves. Filter near your rating, look at the game count, and choose the replies you're likely to meet. Our opening explorer guide covers why the most successful move isn't automatically the right move for your study.
Don't put every engine suggestion into the tree. If the top three moves are all playable, pick the one that produces a structure you understand and add a note about the alternatives. An engine tree with five branches at every turn turns practice into menu navigation.
When two lines reach the same position, choose one chapter as the home and add a comment in the other that names the transposition. Duplicating the full tree seems harmless until you update one copy and forget the second.
Write plans beside moves
Moves expire the moment an opponent deviates. Plans survive. At the end of each main line, add three short notes: where your pieces belong, which pawn break you want, and which trade helps you. That's enough to make the chapter playable after the memorized sequence ends.
Keep the language blunt. Knight goes to e5. Prepare c4. Avoid trading the dark bishop. If your note sounds like a book introduction, you'll skim it. If it sounds like something you'd tell yourself during a game, you'll use it.
Add one warning at the position where the common mistake becomes possible. Don't scatter warnings on every move. A single comment such as `Qb3 now loses time to Na5` is memorable because it sits on the board where the choice occurs.
Arrows and colored squares are helpful when they show a relationship that words make awkward. Mark a bishop diagonal, the square a knight wants, or the pawn break both sides are fighting over. Don't turn the board into a traffic map. Three arrows should make the plan clearer than one arrow, not merely more colorful.
Include a model game only when you can say what it demonstrates. One clean game showing the intended pawn break is useful. Ten famous games you haven't reviewed are decoration. The point of the study is helping you make a decision, not proving you've researched the opening.
Update the study from your own games
The best source for new chapters isn't a grandmaster database. It's the position that confused you last night. After each serious game, replay the opening and find the first move where you didn't know what to do.
If the move already exists in your study, the problem is recall. Mark that chapter for practice rather than adding more theory. If the opponent played a reasonable move that isn't covered, add one small branch and a note. If the position is already outside opening territory, don't force it into the repertoire. Put the lesson in your game notes instead.
This rule keeps the study tied to reality. Popular opening files devote pages to moves you'll never face while ignoring the ordinary club move that has surprised you four times. Your games reveal which gap is costing points.
Review the chapter names once a month. Merge duplicates, rename anything vague, and remove branches you no longer play. Export a PGN before a large cleanup so you can recover a line if you cut too much.
You should be able to open the study and reach any important position in a few clicks. If you can't, stop adding chess and fix the map. Organization isn't clerical work here. It's what lets you practise the right position while you still remember why it mattered.
A weekly rhythm that stays manageable
On Monday, practise the chapters you missed in recent games. Ten focused minutes is enough. Keep the moves hidden, play each line cleanly, and say the plan at the end.
On Wednesday, choose one chapter and check its replies against the opening explorer. Add a branch only if it appears often near your rating or has already appeared in your own games. This prevents theory collecting from taking over the study.
On Friday, play a slow game with the repertoire you studied. Don't switch openings because an opponent might know your line. Familiar positions are the whole reason you built the study. Review the game while the choices are fresh and update one chapter if needed.
Once a month, export the study and clean it. Delete abandoned branches, combine chapters that reach the same position, and rewrite any note you no longer understand at a glance. If a note confuses you now, it won't help under a clock.
That rhythm is deliberately modest. Repertoire work expands to fill every spare hour if you let it, while the returns fall quickly once your common replies are covered. A study should make it easier to play chess. When maintaining it starts replacing games, you've built too much.