Short Answer
> Quick answer: Three. One first move as White with a plan against the handful of common replies, and two defences as Black, one against `1.e4` and one against `1.d4`. That's a complete repertoire. Depth matters far more than breadth: knowing six to eight moves plus the resulting plan beats knowing fifteen moves of a line nobody plays against you.
The mistake almost everyone makes is collecting openings rather than learning one. A repertoire you can recall under a clock is worth several you can only recognise.
Why Large Repertoires Quietly Fail
The appeal is obvious. More openings feels like more preparation, and studying a new one is more entertaining than drilling the one you already play.
What actually happens is that recall degrades across all of them at once. Opening knowledge isn't stored as facts, it's stored as pattern recognition, and pattern recognition needs repetition to hold. Split your games across five systems and each one gets a fifth of the reinforcement, which is usually below the threshold where anything sticks.
There's a second cost that shows up later in the game. Playing the same structure repeatedly teaches you its middlegame, where the pieces belong, which trades favour you, what the endgames look like. That transfer is the real payoff of an opening, and it only accrues to lines you play often.
I switched from four defences to one against `1.e4` and my results improved before I'd learned a single new move, purely because I finally recognised the positions I was reaching. The theory hadn't changed. The familiarity had.
Breadth also solves a problem most players don't have. Being predictable only hurts when opponents prepare for you specifically, and below master level almost nobody does.
What Your Games Actually Contain
Open your last twenty games and count how many stayed in known theory past move eight. When I counted mine the answer was three, and I had been studying to move twelve.
Club opponents leave book early, and they leave it in ordinary ways rather than clever ones. A natural developing move that isn't the main line, a slightly wrong move order, a pawn push that theory rejects for reasons that take ten moves to become relevant. Fifteen moves of preparation is no help against a deviation on move five.
What helps is understanding why your opening's moves are played. If you know the plan, an early deviation is a gift rather than a crisis, because the opponent has left a path they understood for one neither of you has memorised, and you're the one who knows the structure.
Checking this yourself takes ten minutes with the opening explorer, which shows what real players actually play at each point rather than what the theory books prioritise. The gap between those two is where most wasted study time goes.
Lichess publishes the same kind of data through its opening database, and the picture is consistent: the further you go down a line, the smaller the share of games that ever reach it.
The Three Slots Explained
As White you need one first move and a response to each common reply. That's it, and how big it gets depends entirely on which first move you pick.
Playing `1.e4` means preparing for the Sicilian, the French, the Caro-Kann, `1...e5` and a few sidelines. That's more work than it sounds. Playing `1.d4` or a system opening narrows it considerably, which is why system openings are so often recommended to people short on study time. Neither choice is objectively better, they just spend your effort differently.
As Black you need one answer to `1.e4` and one to `1.d4`. Anything else your opponents open with tends to transpose into something familiar or into a position where general principles are enough. Preparing separately for every rare first move is a classic way to spend a month on positions you'll meet twice a year.
Try to pick the three so they share structures where possible. A repertoire built around similar pawn formations compounds, because middlegame understanding transfers between the slots instead of being learned three times over. Our guide on learning openings by training rather than memorising goes into how to build that understanding.
How Deep Is Deep Enough
Six to eight moves, then plans. That covers the overwhelming majority of what you will face.
The important word is plans. Reaching move eight with no idea what to do next is barely better than reaching move four with the same problem. For every line you learn, you should be able to state where your pieces belong, which pawn break you're aiming for, and which trades help you. Three sentences per line is enough.
Sharp tactical openings are the exception and they're worth knowing about before you choose one. Certain gambits and heavily analysed attacking lines punish imprecision immediately, so if you play those, your first fifteen moves genuinely do need to be accurate. That's a real cost and it should factor into the choice rather than being discovered afterwards.
Everyone else is better served going deep on understanding rather than moves. The position after eight moves of a solid opening is one you can play from general principles if you know the structure, and you'll reach it far more often than any move fifteen position.
A useful test: play the first eight moves of your main line against the practice bot without thinking, then stop and write down your plan. If you can't, the depth isn't the problem.
When It Is Time to Add Something
Expansion is worth it eventually, but the trigger should be a result rather than boredom.
Add a line when you keep reaching a specific position you don't understand. That's a real gap and filling it pays immediately. Adding a whole second defence because you watched an interesting video is how repertoires bloat without getting better.
The other legitimate reason is a structural one. If your current opening keeps producing middlegames that don't suit you, changing the opening is the right fix, and that's a replacement rather than an addition. Dropping the old line matters as much as picking the new one.
Rating is a reasonable rough guide for when breadth starts to earn its keep. Below club level the return on a second defence is close to nothing, while at higher levels opponents genuinely do prepare and variety starts to have value. Most players expand years before they need to.
One thing worth doing regularly regardless: review the openings you already have against your own recent games. Our guide on analysing your games effectively covers how to spot which part of the repertoire is actually costing you points, which is nearly always more useful than adding to it.
Building It This Week
The whole job is smaller than people expect once you stop trying to cover everything.
Pick your three slots first and write them down. Committing on paper stops the quiet drift where you play something different every third game and never build familiarity with any of it.
Then take one slot at a time and learn it to eight moves against the two or three most common replies. Not every reply, just the ones you actually meet, which the explorer will tell you in minutes. Write your three sentences of plan for each.
Play those lines exclusively for a month, including in fast games where the temptation to experiment is strongest. Repetition under time pressure is what converts knowledge into recall, and blitz is genuinely useful for this narrow purpose even though it's poor practice for most other things.
Review at the end of the month by looking only at where you left your preparation and what happened next. That single question tells you whether to go deeper, change a line, or leave it alone, and it's a far better use of study time than starting the whole cycle again with a new opening.