CMLL Algorithms
Every algorithm for all 42 Roux corner cases, ranked by how often top solvers actually use them.
What makes this database different
Ranked by real solves
Ordered by what top solvers actually use, not by popularity votes.
Pro usage counts how many reconstructed Roux solves used an algorithm for their CMLL step. Alongside it sits each algorithm's community vote count from SpeedCubeDB, so you can see where the crowd and the pros disagree.
One row per algorithm
The same algorithm written in different ways collapses into one row.
Alg lists repeat themselves: the same turns show up reordered, behind a different alignment, or just spelled a different way - a slice M written as r'R , a wide l written as x'R . Those all move the cube through exactly the same states, so we fold them into one row and add up their votes.
Mirrors are the exception: a left-handed version keeps its own row, because doing an algorithm on the other side is a genuinely different thing to learn.
Grouped into triggers
Every algorithm is boxed into the triggers your hands already know.
Grouping gives an algorithm a shape you can picture instead of a string you have to spell, and it exposes the family resemblances: algorithms built from the same trigger carry the same colour, so the one you already know gives you a head start on the next. The Rightie RUR'U' gets one colour, the sledgehammer R'FRF' another. Hover any box for its name.
Cubers have always marked triggers by hand, in parentheses: F (R U R' U') (R U R' U') F'. The boxes are that convention drawn instead of typed, and they are worked out from the moves themselves rather than stored per algorithm - so every algorithm in the database has them, including the ones nobody ever bracketed.
CMLL Overview
most-used algorithms for every case
