ZZ, and the others
Edge orientation first, F2L without a single rotation, and a fair account of who Petrus and Mehta actually suit.
By the end you will be able to
- Describe EOLine and what it buys the rest of the solve
- Explain why ZZ needs no F or B turns after the first step
- Say what Petrus and Mehta do differently, and where each came from
- Match a method to the way you actually like to solve
ZZ was published by Zbigniew Zborowski in 2006 and rests on one idea: orient every edge on the cube before you build anything. Everything else in the method is a consequence of that decision.
EOLine
The first step orients all twelve edges and places the two bottom edges that run front to back. Six turns is a typical solution and eight is a bad one — but they have to be found during inspection, in a scramble you have never seen, and that is the part that stops people.
F2L without rotating
With the line in place, the first two layers go in as two blocks, left and right, using nothing but R, U and L. There are no awkward back slots, because there is no back — the cube sits still and your hands do the whole solve from one grip.
This is where ZZ feels different rather than faster. The move count is much like CFOP's. What changes is that the turning is continuous, the pieces stay where your eyes left them, and lookahead comes more readily because nothing you are tracking ever swings out of sight.
The last layer, guaranteed
Every ZZ solve arrives at the last layer with the edges already oriented, which is the one-in-eight case CFOP solvers wait for. So the two-look finish is COLL and then one of four edge cases, and the one-look finish is ZBLL — available on every solve rather than one in eight.
Petrus
Lars Petrus — whose name you have already met attached to sune — worked his method out in 1981, and both Roux and ZZ owe it something. It builds a 2×2×2 block, grows it to 2×2×3, orients the edges, finishes the first two layers, then does the last layer.
Petrus is short on algorithms and short on moves, and long on thinking: nearly every step is worked out rather than recalled. It suits people who like the puzzle more than the stopwatch, and it is the origin of the two ideas the other methods took — build blocks rather than layers, and fix edge orientation early.
Mehta
Mehta is the newcomer, published by Yash Mehta in 2020. It builds a first block, then a "belt" of middle-layer edges all the way round, then orients edges and corners, permutes the corners, and finishes with five edges. Each of those is its own small algorithm set.
It is low on moves, keeps the cube still, and asks for a lot of memorisation across several subsets. Its appeal is to people who already like M-slice work and do not mind learning in sets — which is to say, people who enjoyed learning full PLL.
| Method | The distinctive idea | Algorithms | Suits |
|---|---|---|---|
| CFOP | Layers, then a two-algorithm last layer | 78 for the full method | Anyone. Most material, most people to ask, most competition results. |
| Roux | Two blocks and a free middle slice | 42, plus intuition | People who like planning and do not mind slice turns. Strong one-handed. |
| ZZ | Orient every edge before you build | 28 to 493, depending how far you take the last layer | People who hate rotating, and anyone tempted by ZBLL. |
| Petrus | Grow one block into the whole cube | About 16, with a two-look last layer | People who would rather think than memorise. |
| Mehta | A belt of edges, then everything in small sets | Several hundred across the subsets | People who enjoy learning algorithm sets and already turn the slice well. |
Choosing honestly
Between twenty and thirty seconds, the method is not what is making you slow — the pauses are, and they are the same pauses in every method. Below about twelve seconds the differences start to matter, and by then you will have opinions of your own about which part of a solve you enjoy.
So pick for pleasure rather than for theory. The method you find interesting is the one you will practise, and practice beats every argument on this page.
The full comparison Each method broken into its steps, with what is solved after every one. The vocabulary on this page EOLine, belt, block, slice — defined once, with what they are not.