Layer by Layer Rubik's Cube Tutorial (LBL)
What is the layer-by-layer method (LBL)
Prerequisite: before starting, work through Fundamentals — recognize the cube's faces, edges, corners, and layers, know how to turn a face, and pick up the minimal notation. This page won't repeat those basics.
The layer-by-layer method (LBL) solves the cube one layer at a time, from the bottom up: finish the bottom layer completely, then the middle layer, then the top. Once a layer is done, later steps only touch the still-unsolved layers above — never a finished one.
It's the method most people meet first when learning the cube, and it's the foundation CFOP (the standard speedcubing method) builds on. Finish the whole flow and you can take a fully scrambled cube and solve it reliably.
Why is this breakdown sensible? Why do later steps leave the earlier layers alone? The next section lays out the core idea.
Core idea: bottom-up, and every step preserves what's done
The whole logic of LBL compresses to two sentences: complete one layer at a time, from the bottom up; and every step leaves the already-finished parts untouched. The first sets the order — bottom before top. The second is the constraint — every turn you make must keep the finished layers intact and only affect what isn't solved yet. That second line is exactly the beginner's biggest worry: “will one turn mess up what I just did?” The answer, if you follow the method, is no. It's also what separates LBL from random twisting: each step carries the constraint “preserve what's done”, rather than turning freely.
Why bottom-up? Because only the top face can be turned freely without reaching the layers below — so you always keep the finished work underneath and the work-in-progress on top, using the top face as your workbench; the bottom face faces down and is hardest to inspect, so finish it first and never look at it again. Go top-down instead and every turn would crash into a layer you've already finished.
Why does following the method never scramble a finished layer? — a concrete demo
Ignore the notation and don't try to follow along — just watch the two finished layers at the bottom: see them shatter mid-sequence, then put themselves back together.
Watch: F cracks the front open and F' snaps it right back; R cracks the right face and R' restores it — each pair "goes out and comes back", so the bottom two layers never budge. This is a standard OLL top-cross algorithm, but the reason it preserves finished layers is exactly the same as every step you'll learn in layer-by-layer.
Complexity reduction: how 4.3×10¹⁹ states become a string of small problems
A 3×3 cube has roughly 4.3×10¹⁹ scrambled states — a number far too large to memorize or exhaust. What LBL does isn't “remember more”; it breaks that enormous problem into a string of tiny local ones: each step only concerns the few pieces of the current layer, treating the rest as already fixed. You never have to watch the whole cube at once — just place the handful of pieces in front of you, one step at a time, and the cube resolves layer by layer.
It's closer to divide-and-conquer than to memorizing one giant map. Each single step is easy, yet strung together they cover every state. Better still, every layer you finish chops a huge chunk off the remaining state space — later steps search an ever-smaller range instead of flailing across the full 4.3×10¹⁹.
Roadmap
Here are the eight steps of LBL, in order from top to bottom. Click any step to jump to its concept below.
The phase tour: eight steps that form one complete story
The eight steps below form one chain: each follows the one before and points to the next. Read this section and you'll understand the whole layer-by-layer method end to end.
Daisy: lift the four edges that carry white up to the top face, white-side up, ringing the yellow center like the petals of a flower. This step watches only the top face and ignores whether the side colors match — nothing on the cube is fixed yet, so you only lift the white edges up with no fear of disturbing anything, which makes it the gentlest entry point. Once the daisy is set, though, those four white edges still aren't matched to the side centers — the next step sends each one home to form the cross on the bottom.
Cross: send each white edge from the daisy to its proper place — white facing down, and its side color lined up with the center beside it. With all four in place, the bottom face is a white cross, and each white edge's side color matches its neighboring center. This is the first step that asks “the side colors must line up too.” With the cross built, the four bottom white edges are down and aligned — next we fill in the four white corners to complete the bottom layer.
Where each white edge belongs is decided by its side color
Every white edge carries, besides white, one side color. That side color tells you which center it belongs to — the white-blue edge, for instance, goes to the blue center's side. So the cross isn't just “white face down”; it's also “side colors matched to their centers.”
First Layer: insert the four white corners into the bottom layer, completing the entire first layer (cross plus corners equals one whole layer). Each white corner carries three colors — white plus two side colors — and it can only sit in the one spot where those three centers meet. Once this step is done, the bottom layer is finished for good: the white layer is never touched again. With the bottom layer whole and fixed, we move up to the middle layer and insert the four edges that have no yellow.
Why a corner has only one home
A corner's three colors correspond exactly to three neighboring centers. The corner where those three centers meet is the only place that piece belongs — a different color combination means a different spot. So once you read a corner's colors, you know where it goes.
Second Layer: insert the four middle-layer edges — the ones with no yellow — into their correct slots between two side centers. This is where “preserve the bottom layer” kicks in: the bottom is already complete, so while inserting you only touch the top and middle layers and leave the layer below intact. Once all four are in, the bottom two layers are done. With two layers complete, only the top face remains — first we turn the four yellow edges on top to face up, forming the yellow cross.
Why only the edges "without yellow"
Among the cube's edges, any edge that carries yellow is bound for the top layer (the yellow face); only the edges with no yellow at all belong in the middle. So this step's "input" is naturally those four yellow-free edges — the yellow ones are left for the top-layer steps later.
OLL Cross: flip the orientation of the four top-layer edges so yellow faces up, forming a yellow cross on top (the four corners may still be wrong at this point). This step fixes only the edges' orientation, not their position, and the bottom two layers stay exactly as they are. Once the yellow cross appears, the top edges face the right way — next we flip the four top corners so yellow faces up too.
Why fix orientation first, not position
The last two things to solve are "orientation" (which way a sticker faces) and "position" (where a piece sits), and they're independent. OLL handles orientation — turning every sticker to face its correct color; PLL then handles position — moving the correctly-oriented pieces into place. So the yellow cross only asks "yellow facing up"; exactly where that edge belongs horizontally is PLL's job.
OLL Corner: flip the orientation of the four top corners so yellow faces up — by the end of this step the entire top face is one solid yellow (this completes OLL). The top layer's edges and corners may still be out of position, but the whole face is the right color, and the bottom two layers still hold. With the whole top face yellow, every sticker faces the right way — what's left is to move the top-layer pieces into their correct positions (PLL).
PLL Corner: from here we enter PLL — move the four top-layer corners into their correct spots (edges aside for this step). The top face stays yellow and the bottom two layers stay put. With the four top corners each in their place, only the four top edges remain out of position — the final step sends them home.
PLL Edge: the final step — move the four top-layer edges into their correct spots. The moment all four land, all six faces are complete and the cube is solved. No orientation work here (OLL already handled that), and the bottom two layers are never touched — you simply move the top edges into their homes. This is the finish line of the layer-by-layer method.
How to study
The learning loop is the same for every phase: first read the concept here, then open the phase's own tutorial to see its structure (how the cases are organized, how to recognize them), then go to the practice area and work that phase's cases by hand until they're solid, and only move on once it's consolidated.
Follow the order strictly — don't skip stages. The eight steps depend on each other layer by layer: each one assumes the ones before it are done. Skip a step and the later steps have nothing solid to stand on.
A soft standard for “I've learned a phase”: without the tutorial open, you can handle most of that phase's cases, fairly reliably. That's enough to move on — “rock-solid on every single case” is a matter of practice volume that comes later.
Finally, keep the three layers' roles straight: the method page (this one) is a map you read once to grasp the whole method end to end; the phase tutorial goes deep on that step's structure; the practice area is where you drill the cases repeatedly. Use all three together.