Zipper Method Tutorial - pTr

pTr · Apr 23, 2025

The Zipper Method is a speed-solving method created by Justin Taylor in 2017. The method was created as a Two-Look solution for the Last Slot and Last Layer without preorienting edges and maintaining a manageable algorithm count. This allows great versatility in approach for the F2L, along with a smooth transition into LSLL. Additionally, this method has a very fast LS+LL, as it combines the well-established OLLCP step with L5EP, a 2gen, low algorithm step with easy recognition and execution. The method retains every ergonomic advantage of CFOP, while containing one fewer "look" in the solve and saving an average of 9 moves with a CFOP-like approach to F2L.

Zipper Steps

Cross + 1 Corner (Fish): This is the most distinctive part of the Zipper Method. Taking an average of 6 moves and no more than 9 moves, this step solves the Cross on the bottom and any first-layer corner, forming a "fish" on the bottom layer. This slot is referred to as the Zipper Slot. Technically, the Zipper Slot can be solved at any point during the F2L, such as using Multislotting to insert the lone corner during the solving of another slot. This is done whenever it is easiest during F2L execution.

F2L - 1 Edge: There are three remaining F2L slots to be solved. Typically, this is done using pairs as in CFOP. However, any approach can be taken to solve the cube up to F2L-1 Edge. First Block, carried over from Roux, may be used in conjunction with <RrUM> for the rest of F2L to provide an efficient and rotationless option to finish F2L.
OLLCP: This is the first algorithm set of the Zipper Method. There are 331 algorithms to orient the last layer of the cube and permute the remaining corners in an average of 11 moves with as few as 6. Although OLLCP algs are often used as an extension of CFOP, the full set must be used with Zipper in order to guarantee that the corners are permuted. In order to correctly use OLLCP in Zipper, the orientation of the edge in the Zipper Slot must be accounted for. Using a similar recognition style as ZZ, the Zipper Slot is placed in either the FR or BR position. Using this, the edge that belongs in the Zipper Slot is treated as any other LL edge, and the OLLCP alg is executed.
L5EP: This step solves the remaining 5 oriented edges of the cube, containing the LL edges and either the FR or BR edge. This step is executed in an average of 10 moves with as few as 6. There are 12 algs for each slot, as well as the 4 standard EPLL algs. This set can be executed using exclusively the <RU> move group, but many of the fastest algs for each case use other move groups.

Variants

Zipper-b
Instead of OLLCP and L5EP, CFRLL (CLL without preserving the FR edge) and Zipper L5E can be used to finish the solve. This approach is generally thought of being superior to standard Zipper.

Zipper-c
Zipper-c is a more advanced form of Zipper-b which keeps the Zipper L5E step but replaces corner + CFRLL with a form of L5C, where the corners after last slot are solved without regards to the currently unsolved edges. Although it is superior to Zipper-b, it is rarely learned due to the amount of algorithms required. However, all of its subsets (CFRLL, TCFRLL and CFRLS) have already been generated.

Zipper-D
Zipper-D is a lower algorithm count version of Zipper-B. It starts with an algorithm set (called OCFRLL) of 85 algorithms to solve top layer corners while orienting what ever edge is in the FR slot. Then you do a reduced version of L5E with 120 algorithms, this sets is called OFL5E. This variant as a whole saves ≈40 algorithms.

ZZ-Zipper
Zipper can also be applied to the ZZ method by using algorithms that preserve EO.

In an intermediate variant, the last D-layer corner is solved. This is followed by one of 42 COLL (algorithms that don't preserve the FR edge but do preserve EO may be used instead) and L5EP.

The most advanced version of this would be to solve the last five corners in one step using one of 614 L5CO algorithms (Last 5 Corners with (edge) Orientation).

efattah · Apr 23, 2025

OLLCP has 331 algorithms, my understanding is they account for either 0, 2 or 4 oriented edges. But with Zipper, the zipper slot edge can now have 2 orientations, so the LL could have 0, 1, 2, 3, or 4 oriented edges. It seems the new algorithm count is thus 5/3 * 331 = 551, and that still doesn't account for orienting the zipper edge; L5EP would have to orient the zipper edge as well as permute. Did I miss something?

pTr · Apr 23, 2025

Example Solve #1

L2 U2 R2 B R2 B' U2 B2 F' U2 B2 L' F2 U' B F R' F2 R' D R'

z2 // incpection
F' R B' R' D R' // cross
(U' R U R' U R U' R' U B' U B) (R U' R' U' L' U L R' U' R U2' R' U' R) (R U2' R' d R' U' R // F2L-1 edge
(r U r' U) (M' u2 R U) (r' F' U F) // OLLCP DS+L
[U R U', M'] ' // 3 style would be best here

Not that good that this method as I'm still learning it.

eeeeeeeee · Apr 23, 2025

efattah said:
OLLCP has 331 algorithms, my understanding is they account for either 0, 2 or 4 oriented edges. But with Zipper, the zipper slot edge can now have 2 orientations, so the LL could have 0, 1, 2, 3, or 4 oriented edges. It seems the new algorithm count is thus 5/3 * 331 = 551, and that still doesn't account for orienting the zipper edge; L5EP would have to orient the zipper edge as well as permute. Did I miss something?

5/3 x 331 is definitely not 551. its 551⅔??

pTr · Apr 23, 2025

In order to correctly use OLLCP in Zipper, the orientation of the edge in the Zipper Slot must be accounted for. Using a similar recognition style as ZZ, the Zipper Slot is placed in either the FR or BR position. Using this, the edge that belongs in the Zipper Slot is treated as any other LL edge, and the OLLCP alg is executed.

Using this, the edge that belongs in the Zipper Slot is treated as any other LL edge, and the OLLCP alg is executed.

example.
R U R' U' R U R' F' U F R U2 R' U2 F' U2 F U' R U2 R' U R U' R' U2 R U R' U F' U' F U R U2 R' U' F R' F' R

In order to get the U face EO to be all 'good' you need to treat your F2L edge as a LL edge. In this case the green on BU is treated as a U face colour.

First recognition is the OLL case, for OLLCP this is called DT. Next is the CLL case recognition, this is TU. The OLLCP case is DTU

(r U2' R' U' R U') (r2' U2' R U R' U r)

Next is L5E or 3-Style

This is a good 3-style case. If you're not familiar with 3BLD notation I will also use face lettering, the first face being the target side of the piece.

This is BJ(B=UR,J=FR), my memo is Basculegion (Pokemon) [bask-you-lee-shin] the com is E': [U', R' E R] it's done E' U' R' E R U R E' R' E

The goal is move UF to UR to FR back to UF. This alg is the one above.

efattah · Apr 24, 2025

This still does not explain how you have five edges remaining but the OLLCP algorithm can only orient four edges, not five.
How do you orient the F2L zipper slot edge? Or are you doing L5E as orient & permute in 1 step?

pTr · Apr 24, 2025

Last 5 Edges Permute, often abbreviated L5EP, is a 3x3 subset that permutes the 4 edges on the last layer and an edge from another layer while simultaneously preserving all other pieces.

The five affected edges must be correctly oriented for L5EP to work. L5EP is often used in methods like Petrus-W and Portico, when there are only five edges left to solve on the cube.

There are 16 algorithms for L5EP, 10 excluding mirrors and 7 excluding mirrors and EPLL.

efattah · Apr 24, 2025

OLLCP can orient 4 edges but you have 5 edges to orient and if you use L5EP that does not orient any edges.
So how do you orient the 5th edge?

Twisted101 · Apr 24, 2025

efattah said:
OLLCP can orient 4 edges but you have 5 edges to orient and if you use L5EP that does not orient any edges.
So how do you orient the 5th edge?

either doing d or nothing if a U-layer edge is in FR, but idk what happens if the FR edge is flipped in its spot

Filipe Teixeira · Apr 25, 2025

why learn 500+ algs when you can just blockbuild and use only 9 algs for finishing the cube? lol

is not like the more cases you know the faster you will solve you know...

pTr · Apr 25, 2025

Filipe Teixeira said:
why learn 500+ algs when you can just blockbuild and use only 9 algs for finishing the cube? lol

is not like the more cases you know the faster you will solve you know...

That’s a fair point—but I think it comes down to what your goal is. Is it just to solve the cube, or to solve it as fast as possible? Let's be real: most of us aren’t breaking the 3.05 world record anytime soon.

For me, speedcubing is about having fun. I enjoy learning new methods and pushing my understanding of the cube. I share what I learn because it’s exciting, and if that excitement sparks joy for someone else, even better.

efattah · Apr 25, 2025

Technically speaking if your recognition is near instant and your algo recall is perfect, then yes, the more algorithms you know, the faster you will solve, that is why Tymon had to learn 800 ZBLL/ZBLS to improve even a fraction of a second, why would he bother to learn those if it wasn't going to make him faster? Jessica Fridrich was famous for asking struggling solvers, 'how many algorithms do you know?' And that was her main point in teaching a person to get faster. Having said that, with a method like Roux you can be sub-6 with something like 80 algorithms so it doesn't take a lot to get fast.

Filipe Teixeira · Apr 25, 2025

pTr said:
That’s a fair point—but I think it comes down to what your goal is. Is it just to solve the cube, or to solve it as fast as possible? Let's be real: most of us aren’t breaking the 3.05 world record anytime soon.

For me, speedcubing is about having fun. I enjoy learning new methods and pushing my understanding of the cube. I share what I learn because it’s exciting, and if that excitement sparks joy for someone else, even better.

I liked your attitude.

By the way, kirjava was the first person to learn the whole ollcp and he said it wasn't worth it.

Thread 'OLLCP (hax)'

Aug 16, 2011

OLLCP (or CLLEO) is a substep for either the Roux or CFOP methods. It will fully solve the corners and fix the orientation of edges within the LL.

Here's a wiki link to my table of all cases.

I used an expanded version of the SuneOLL technique, using Sune/Niklas/Fruruf/SexyHammer variation combinations instead of just Sune. Multiple combinations have been listed in the table.

There are 10 cases that can't be solved with two of these algs so alternatives have been provided.

Not all the cases are amazing - but I'm sure you can find some that you like ^^

I only...

Kirjava

Replies: 22

Forum: General Speedcubing Discussion

he used a "hack" method to learn the algs that consisted in combining different triggers to achieve all the cases and practiced after learning all of it.

I believe he tried zipper too, but it is harder than ollcp because you have to pay attention to another piece and use setups.

In my opinion if a person would learn 500+ algs they could learn a better set instead.

maybe some good chunks of zbls and zbll and a solid edge influencing technique would be more beneficial than a rigid method like zipper

pTr · Apr 25, 2025

You are probably right, the fact that Zipper isn't as popular as ZB and CFOP means that it probably isn't as good. I like the way the method works, but I'm not going to learn any OLLCP algs, but keep working on ZBLL. But I didn't see a Zipper method in the forums, so I posted one.

Zipper Method Tutorial - pTr

pTr

Member

Zipper Steps

Variants

efattah

Member

pTr

Member

eeeeeeeee

Member

pTr

Member

efattah

Member

pTr

Member

efattah

Member

Twisted101

Member

Filipe Teixeira

Member

pTr

Member

efattah

Member

Filipe Teixeira

Member

Thread 'OLLCP (hax)'

pTr

Member

Similar threads

Forum statistics

Share this page

Zipper Method Tutorial - pTr

Member

Zipper Steps​

Variants​

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Similar threads

Zipper Steps

Variants