Full CFOP means 78 OLL and PLL algorithms, plus around forty F2L cases that you mostly pick up through practice. Nobody memorises all that in a month, and trying to is the surest way to give up. Here is a realistic plan, to adjust to your own pace, and above all the principles that make it sustainable.
The principles
- Few algorithms at a time. Two or three new algorithms a week, no more. Beyond that, last week’s ones start to fade.
- Always a full solve. At every stage of the plan, you can solve the cube. You never drop a step until its replacement is ready: see Moving from the beginner method to CFOP.
- Practice first, algorithms second. A session is mostly solves, with a little new material. Not the other way round.
- Review is part of the plan. Every week, the algorithms from previous weeks go back into review: see How to learn an algorithm.
- Measure something other than the timer. At first, count your pauses, your glances and the algorithms you know; the timer comes later, see Fluency.
The plan, in phases
The durations assume a twenty to thirty minute session a day, five days a week. With half that, allow twice as long: it works just as well.
Phase 1: the foundations (2 to 3 weeks)
- The white cross on the bottom, without peeking underneath.
- Intuitive F2L, with the four basic inserts and the “white facing up” case.
- The last layer the way you already do it (beginner method).
End of phase: you finish F2L on your own, even if slowly.
Phase 2: the last layer in two looks (3 to 4 weeks)
- Weeks 1 and 2: 2-look OLL, the three edge algorithms then the seven corner ones. You probably already know a few of them from the beginner method.
- Weeks 3 and 4: 2-look PLL,
TandYfor the corners,Ua,Ub,H,Zfor the edges.
End of phase: you solve with CFOP from start to finish, with 16 algorithms.
The sixteen algorithms of phase 2: three for the yellow cross, seven for the yellow face, two for the corners, four for the edges. They all stay exactly as they are in full CFOP.
Open the “Yellow cross” step in the appPhase 3: the 21 PLLs (5 to 8 weeks)
Two PLLs a week, in this order: Aa, Ab, then Ja, Jb, then F, Ra, Rb, V, then Ga, Gc, then Gb, Gd, and finally E, Na, Nb. Alongside this, work on PLL recognition: learn each new case together with its tell.
End of phase: one PLL, one look, one algorithm.
Phase 4: F2L fluency (ongoing from phase 3)
No algorithms to learn here, but it is real work all the same: looking for the next pair during the insert, cutting down rotations, cleaning up your fingertricks. See Fluency. This is the phase where you start timing yourself regularly.
Phase 5: the 57 OLLs (4 to 6 months)
Two or three OLLs a week, by family, in the order given on the OLL page: the short algorithms you half know already, then T, P and C, then Squares and Small lightning bolts, then Fish, W, Knight move, L and Awkward, then Lines and Big lightning bolts, and the Dots last. You already know the seven Cross cases.
End of phase: one OLL, one look, one algorithm.
Phase 6: refining F2L (ongoing)
Compare the way you solve each pair with the 41 cases in the app, and learn the ones you handle badly: see Recognising F2L cases. That usually means about ten algorithms, since the rest are already what you do.
A typical session
| Time | What |
|---|---|
| 5 min | Review: the algorithms from the last two weeks, two or three times each, starting from the real case |
| 10 to 15 min | Solves, applying whatever you are working on (looking for the next pair, naming the PLL before the algorithm…) |
| 5 to 10 min | New material: one algorithm, using the method in How to learn an algorithm; or pure recognition if that is what is holding you back |
On low-energy days: just the review and a few solves. A short session every day beats a long one on Sunday: muscle memory builds with sleep between repetitions.
Tracking your progress
You don’t need a complicated app. A sheet of paper, or a printout of the app’s Formulas sheet, and three states per algorithm:
- to learn;
- in progress (I know it if I think about it);
- learnt (it comes out on its own, three times in a row on a scrambled cube).
A “learnt” algorithm that makes you hesitate goes back to “in progress”. Look at the sheet once a week to pick the next ones: that is all it takes to avoid forgetting rare cases (some OLLs only come up once every fifty solves).
When you lose steam
- You have been stuck for weeks: it is almost never about algorithms. Time each step separately (cross, F2L, OLL, PLL); the one that takes longer than it should (see the benchmarks in Fluency) is the one to work on.
- You forget faster than you learn: too much new material, not enough review. Stop learning for two weeks and only review and solve.
- You are bored: change what you practise, not your method. Some pure recognition, a blindfolded cross challenge, a “zero cube rotations” F2L session.
- You want all 57 OLLs right now: reread the order of priorities in the overview. Full OLL saves you two seconds; a fluent F2L saves you twenty.
The full plan takes six to twelve months. That is a long time, but every phase improves your solves from the very first week, and nothing you learn will need unlearning later.