Hello forum, long time since I posted a new thread here
Today I managed to write something that I have had in mind for quite some time: a tutorial for beginners, aimed at people that want to figure out how to solve a cube by themselves rather than following a step-by-step guide. As such, it explain 3 "basic" concepts:
- How the cube works (centers are fixed, edges have two colors etc...)
- Blockbuilding
- Commutators
Now you might argue that commutators are an advanced topic, but I think it makes sense to teach them to beginners, at least if they want to try and solve a cube without reading a proper tutorial.
Here is a link to my writeup:
Rubik's cube: how to figure it out.
A true legend returns
So, I had a similar though experiment a few years ago. My partner, at the time, refused to let me teach them how to solve a rubik's cube because she, "needed to figure it out herself". Naturally, a question arose, "How do you teach someone to solve a rubik's cube while imparting the least information possible". This would include everything, including things like notation. We're assuming the person we are teaching has no background knowledge whatsoever. The method/approach would need to be completely intuitive, including algs; ones which could be "reasonably" found. This would leave out things like commutators/conjugates. Idealy, our method would be broken down into subpuzzles, ones which could be bruteforced (Not exactly what we want but passes the bar of "reasonability"). What I came up with is a mix of LMCF, Morozov, SSC, HSC, and Guimond
Steps:
(1) Corner Orientation. Same step as in Guimond. All of the CO algs are <R U>. The apporach would be to setup into R U R', R U2 R', R U' R'. Our subpuzzle is 2-Gen Corner Orientation.
(2) Corner Separation. Separation is inuitive - no notes.
(3) Corner Permutation (CPBL). All you need to do is figure out Adj/Adj. Our subspuzzle is solving corners with <R2 U D F2 B2 L2>
(4) E-Slice Edges. <R U D M>. Setups into R U M* U' R' and other variations (Reflections and Inverses).
(5) Edge Orientation. Roux EO with D moves to set up into <M U> EO. Variations of M' U M.
(6) Edge Permutation. Roux 4c with 2 extra edges, <M U D>. Setups into M' U2 M and other variations.
Notes:
You can solve 2-3 E-Slice edges (Psuedo-Pairs) and then solve CO. Example: 3 oriented corners + E-Slice minus FR Edge => <R U> CO => insert FR edge => CPBL
Step (2) and (3) can be combined.
Step (4) could be extended to DL + DR and ending with <M U> L6E.
You could start the solve with EO and continue as normal.
How would it be taught?
This is a wonderful question - but how does it help us? Well, we can reform our original question, "How do you teach someone to solve a rubik's cube while imparting the least information possible" and rather ask "How efficiently can we teach our method". We can quantify this by counting the number of concepts you need to learn and how much we can extrapolate and reapply from one step to another.
Our progression would looksomething like this:
(1) CP: Break down CS and CPBL into the <R2 U D F2 B2 L2> subgroup. Important concept: U/D colour pieces facing the same direction are oriented.
(2) EP: Same pinciple as CPBL with a <M U D> subgroup. Opposite U/D colours recognition transfer between steps. Trigger: M' U2 M. Congrats, you can now solve DR!
(3) EO: <R U M D> subgroup. Opposite U/D colours recognition transfer between steps. Trigger M* U* M*. EO is iterative.
(4) Inserts: Basic principle as EP with a setup. R U M* U' R'. This isn't that hard to figure out extrapolating from our M* U* M and M* U2 M* triggers.
(5) CO: 2-Gen CO. Opposite U/D colours recognition transfer between steps. Triggers: R U R', R U' R', and R U2 R'.
Notes:
Most of these steps are interchangable. Example: teach CP and then CO to give the solver the ability to solve a 2x2.
If we look through our steps we'll notice 2 things. (1) U/D colour recognition extends between steps and (2) our steps are all intuitve or brute forcable. 2-Gen CO and EO can be brute forced (both are iterative). EP and Inserts are intuitive. CPBL is reasonably findable (not 100% brute force but with practice you can "accidentally" solve corners). Another note is that since we break down steps into subgroups we don't need to teach notation. Its also allows the student to practice each step separately (Just need to scramble within the subgroup for each step).
Anyhow, I think that's everything.
