COALL
| ||||||||||||||||||||||
COALL (Corner Orientation And Last Layer) is an experimental 3x3 method, a ZB variant that orients the last-layer corners when inserting the last F2L pair (COLS). Orienting the corners first, rather than the edges like in normal ZB, has the advantage of requiring less algorithms (111 total compared to ZB's 300).
To solve using COALL, solve the cross and three F2L slots using any method. Next, make the last F2L pair while using COLS to insert it and orient the last-layer corners. Finally, finish the solve using XLL.
The method was re-proposed in 2025 by Jesus Christ and Marvel Lio the latter of whom renamed the method as the Jesus-Marvel method (JM Method). The former proposer was the biggest contributor to the idea and provided comprehensive details on beginner and advanced variants of the JM method, compiled by the latter proposer, which are detailed below.
JM Method details
Similar to original COALL. Do the solve till F2L-1 and then use JMLS and JMLL to finish the solve.
JM Last Slot (JMLS)
With JMLS you solve the last slot while orienting last layer corners.
For ease of learning, this is split into beginner and extended sets. None of the sets care about preserving edge orientation.
Beginner set
The beginner set, also known as insertion algs, of JM last slot involves orienting corners when inserting the last F2L slot.
There are only 57 cases total that are akin to the Winter and Summer variation of VLS/HLS.
It is recommended by Jesus to mirror the algs to the left-hand side rather than learning a new set.
Though you can technically cover all F2L cases using the insertion algs, it is recommended to learn the extended sets especially the extended set A.
Extended set
With the extended set, corners are oriented when solving the F2L pair as opposed to when inserting it. For the extended set Jesus does not recommend mirroring algs, but rather learning them directly.
He recommends learning the extended set A before the beginner sets. This is to avoid an over reliance on the beginner set and putting off learning the extended sets entirely.
He also does not recommend viewing certain sets as being more “worth it” than others. They are all worth learning.
Extended set A
This set consists of all F2L cases that are typically solved in 10 – 11 moves using standard f2l inserts. These cases in this set cover F2L 31, F2L 37, F2L 38, F2L 39, F2L 40 and F2L 41.
For this set, using the insertion algs would be too inefficient.
Extended set B
This set consists of all other F2L cases. You can find an incomplete set of the JMLS algs here. The original generators of the algs are CubeBlazer, Cubeur-Manchot, and "Justin" as part of the COLS method proposed by Chester Lian. I will generate the rest.
JM Last Layer (JMLL)
After orienting corners during last slot, the cube is in 1 of 4 possible states. 1LLL 28, 1LLL 57, Anti-PLL or PLL.
The last layer is solved in 2 looks using the beginner version and 1 look for the advanced version.
Jesus recommends aiming for 1 looking.
Beginner version
If the solver gets either a 1LLL 28 or 1LLL 57, they use 1 algorithm to reduce the cube to an Anti-PLL by misorienting edges. There are only 2 algorithms total needed for this, one for 1LLL 28 cases, and the other for 1LLL 57 cases.
For 1LLL 28 one possible alg is r U R' U' M U R U' R', for 1LLL 57 a possible alg is R' U' R U M U' R' U r. To misorient edges for 1LLL 28 the alg should be done a U2 away from how you would normally do it when solving OLL.For 1LLL 57, it should be done a U or U’away.
If the solver, gets an Anti PLL they do the alg directly.
Alternatively, they could get a PLL and do that alg directly.
The total number of algs for the beginner version is only 24, not including PLL. These are the 22 Anti PLL algs and the 2 algs needed to reduce to an anti PLL. Anti PLL is the main set for the beginner version.
Advanced version
The solver finishes last layer in 1 look. The total number of algs is 164 not counting PLL.
The set consist of 1LLL 28, 1LLL 57, Anti-PLL and PLL.
Note: Because this is a last slot to last layer technique, it can be used across many different solving methods. CFOP users may find it appealing and I highly recommend it to Petrus users (get rid of the EO step though since JM last slot does not care to preserve edge orientation). I propose the names CFJM and Petrus-JM respectively.
Pros (JM Method)
- The advanced version of the JM method has a low move count and is more efficient than the ZB method on average by a lot according to Jesus.
- The last layer has a low algorithm count of only 164 cases (not counting PLL) which is more manageable to learn than ZBLL.
- When compared to a set like ZBLL, it is relatively common to get an 8 or 10 move last layer with JM Last Layer.
- JM last slot has very easy recognition and can be learned quickly.
- Jesus thinks it’s very easy to learn.
Cons (JM Method)
- The totality of JM last slot is a gigantic alg set with over 1000 cases but worth learning according to Jesus.
Pros (Original COALL)
- Much fewer algorithms than ZB and perhaps as fast
- Reasonable recognition
Cons (Original COALL)
- Relatively large number of cases (111, including 21 PLL)
- Has some bad algorithms
- An extra look during Winter Variation/Too many algorithms with COLS
See also
External links
- Speedsolving.com: KLL
- Speedsolving.com: WW New Speedcubing Method ZB Variant
- Discussion on speedcube.de (German)
- Full COALL and COLS algorithm sheet
