ZZ Method - Variants (2024)

In ZZ, there are 4 main variants. This page will go through each one and their own sub-variants.

ZZ-A:

• Full ZZ-A with ZBLL- This is where you solve F2L normally, then you use 1 of 493 algorithms to solve the last layer in one look. It is called ZBLL as Zbigniew Zborowski and Ron van Bruchem came up with a method which also had the same finish, and coined the term.
• OCLL/PLL- When you get to last layer, orientate all the corners (Orientate Corners Last Layer) and then permute everything (Permute Last Layer). There are 7 algs and 21 algs for each step respectively and they are very well developed. A brilliant place for beginners to start.
• COLL/EPLL- When you get to last layer, solve the corners while preserving edge orientation (Corners Orientated Last Layer), then permute the edges (Edge Permutation Last Layer). Uses 42 and 4 algs for each step respectively and they are also well developed. Used in conjunction with OCLL/PLL for the bad cases.
• OCELL/CPLL- When you get to last layer, orientate the corners and permute the edges (Orientate Corners Edges of Last Layer) then permute the corners (Corner Permute Last Layer). Uses 40 and 4 algs for each step respectively, and the algorithms for OCELL are all only <RU>, meaning that they are very ergonomic, although the recognition is worse than with COLL/EPLL.
• SIMPLE- Solve the edges and a corner of LL using one 1 of 27 algorithms, then use a commutator to solve the last 3 corners. (This is an improved version of Fish and Chips.)

ZZ-B:

• Full ZZ-B with ZZLL- This is where, when you insert last slot you phase the edges. Phasing is a lightweight technique where you solve 2 opposite edges during last layer so you get a reduced algorithm set. ZZLL is the set of algorithms that solve phased LL. There are 169 algorithms for ZZ-B if you do phasing intuitively and 227 if you do it algorithmically. Can be used as a transition to full ZBLL. Using algorithms for phasing is called ZZ-B+.
• ZZ-R- This is where you phase edges during last slot, then do OCLL/PLL. You have the normal 7 OCLL cases (but the algs must preserve phasing), then you have a reduced set of PLL cases (9 cases). This gives only 16 algs to solve the last layer in 2 looks.
• Corner Phasing- When you are inserting your last slot, you force at least 2 corners orientated, then solve LL with one of 237 ZBLL algorithms from the T, U, L or O (PLL) sets. Could be used as a transition to full ZBLL.
• Blah- Very similar to corner phasing, but instead of forcing 2 corners orientated, you force 0 corners orientated so that you can solve the last layer in one look with one of 112 ZBLL algorithms from the Pi and H sets. Also could be used as a transition to full ZBLL.

ZZ-C:

• Full ZZ-C with OLS- This is where you solve your last slot+OCLL (Orientate Last Slot), giving you just PLL. There are 502 OLS algorithms plus the 21 PLL algorithms.
• WV+SV- These 2 algorithm sets are known as Winter Variation and Summer variation (Winter because the person was called Winter and Summer because it's the opposite to Winter). They are the OLS sets where the pair is already made and where it is R U' R' or R U R' away from being inserted respectively. They are both 27 algorithms each, and require you to set up the pair, insert it plus OCLL, then PLL, essentially forcing an OCLL skip every time. As there are fewer algorithms, it is possible to learn them for all slots, totalling up to 216 algorithms in total, although it is generally only recommended that you just do the easy cases and use another variant for the rest of the time unless you plan on transitioning to full OLS.
• WVCP- This is where you pair up the last F2L pair in last slot (just like in Winter Variation), then you do a WV alg that also solves CP. This means that you always get an EPLL for LL, although this isn't always worth it as the effort to recognise and there are better options out there.
• ZZ-MGLS- This is the ZZ version of the MGLS (Makisumi-Garron Last Slot) method. You insert the FR edge, then insert the DFR corner while orientating the LL corners using one of 104 CLS (Corner Last Slot) algorithms, before ending with PLL. They are the subset of OLS algorithms where the edge is already inserted.
• EJLS- This algorithm set is the subset of CLS in which the corner is in DFR but twisted, meaning that there are only 16 algorithms before PLL. It was created by Erik Johnson.
• CT- This was created by Chris Tran (hence the CT) and is where you insert just the FR edge and do OCLL using one of 104 TSLE semi-intuitive algorithms (Tran Style Last Edge) then permute the last layer plus the DFR corner using one of 93 TTLL algorithms (Tran Thompson Last Layer), of which 21 are just PLL. In contrast to all the other ZZ-C variants, slot neutrality is more plausible with this as you can do intiuitive TSLE into any slot, then ADF so that the D corner is at DFR, then you can do TTLL from there.
• C++- This method marries ZZ-CT and ZZ-C to create arguably the best ZZ-C variant. For every OLS case that is bad, you almost always get a good TSLE case, so TTLL can be used instead.

ZZ-D:

• Solving CP during last slot- There are different ways of doing this, but they all use some form of CPLS (Corner Permutation Last Slot). There are two easy ways: solve the FR edge then do the subset of CPLS where the FR edge is solved or set up so that you have a pair, then do that subset of CPLS. This is called ZZ-Orbit. Full CPLS is 117 algorithms then ends with 2GLL (2 Generator Last Layer), the subset of ZBLL where CP is solved. It is highly ergonomic, and is good for especially one handed and feet solving. There are 84 algorithms in 2GLL.
• Solving CP after LB- There are a couple of ways of doing this, but Porky v1 is the best currently. After you solve the left block, you place the right block corners in D, recognise CP, then do an algorithm to solve CP. It is then <RU> right block then 2GLL. The major drawback of this method is the CP recognition and the time taken to set it up. You can also do ZZ-E which is Porky v1 but without placing the right block corners. This is impractical for speedsolving.
• Solving CP with LB- There are a few methods: Porky v2 and Rainbow are a couple, plus several ideas that have been proposed over time (see this speedsolving thread, especially the last few posts, plus other ideas randomly dotted around). Porky v2 is where you, in last slot, recognise the CP and set up to insert the last pair. Rainbow is using Porky v1, but doing it on left instead and separating the left block pieces from the right block, so that when CP is solved, you can do left block <LU> then right block <RU>, then 2GLL. Again, the recognition and the extra time used to set up is a big disadvantage.
• Solving CP before LB- There are currently no viable methods that do this, but a theoretically viable method (which uses 2GR's CP method) is to solve EOLine and permute the DBL corner while tracking CP, separate left block pieces from the right block while solving CP, then an <LU> left block and an <RU> right block, then 2GLL. This method would require a lot of practice and an extremely efficient use of the 15s of inspection to be able to consistently do this method and it is almost certainly not worth it when compared to full ZZ-A.

Miscellaneous variants:

• Cardan Reduction 1- Reudce the cube in last slot to having the FR edge plus a 1x2x2 block on U solved, then solve the L4C either with algorithms or 2 commutators.
• Cardan Reduction 2- Reduce the cube in last slot to having the FR edge plus a pair on U solved, then use one of 144 algs (72 and mirrored) to solve all the edges plus a corner of the LL. Use a commutator to solve the last 3 corners.
• Cardan Reduction 3- Just reduce the cube in last slot to being last 3 corners, 1 of which is DFR, then use a commutator.
• ZZ-Zipper- In last slot you use 1 of 614 algorithms L5CO (Last 5 Corners with (edge) Orientation) to solve the last 5 corners (they are relatively short algorithms), then use 1 of 12 L5EP (Last 5 Edge Permute) algorithms for the last 5 edges.

Completely different ways of solving the cube ZZ style:

• ZZ-LOL- This is ZZ but you solve the EOLine on left so that F2L is just <RUD>, a very ergonomic movegroup. The lookahead is arguably worse than standard ZZ though.
• ZZ-Portico- Instead of solving EOLine, you solve EODB, then do blocks as normal, COLL, then L5EP with 12 algs. Leaving the DF edge unsolved allows for more efficient blockbuilding during F2L through F2 moves. It is better than COLL/EPLL, but worse than ZBLL.
• WaterZZ- This method takes inspiration from WaterRoux which in turn takes inspiration from Waterman. You solve an EO2x2x2, in the back, then solve a square in the back and a pair in the front so that you end up with DF, FR and DFR unsolved. You then use 1 of 614 L5CO (Last 5 Corners with (edge) Orientation) algorithms to solve the last 5 corners, then 1 of 95 L6EP (Last 6 Edges Permute) algoritms to finish off the cube. This could be seen as an extension to Portico.
• ZZ-4c- This method takes inspiration from the Roux method. You instead of EOLine, you place either UL and UR or UF and UB. You do F2L as normal, then you do COLL. Afterwards, you solve the EOLine pieces with an M2, then you have 4c of the Roux method. It is similar to Portico, as you have fewer moves than COLL for extremely similar ergonomics, but it is definitely worse than ZZ-A.
• ZZ-SP- Snake Pattern is a way to force 2 right (or left) blocks. You solve EOLine plus right block, then you do a z rotation, then another right block on what was U then do LL on what was L.
• ZZ-Top- When doing EOLine, you ignore the orientation of the U layer edges, giving you ZZF2L but CFOP LL.