Why do you use the method you use?

Pretty explanatory title.

I'm currently using Petrus because I didn't want to have to start trying to speedcube by looking up an algorithm in a table and just applying it, but actually solving the cube. (Although keyhole F2L looks interesting)

I like the idea of the topic, but I don't like it when people just say "Pretty explanatory title." There's always more tosay, and it's nice to see the effort of elucidation. Do you mean 3x3x3? Everything? Anything? Approach to cubing in general? Method of getting fast at a new puzzle?

Anyhow, for 3x3x3 I use MGLS. I like to work on cubing ideas, and there is a certain pride in trying to be successful while being different. The only way we're ever going to improve cubing, other than practice, is development. So, I found and learned algs to show that it can be done, and it turned out to be a decently competitive method.
However, I have recently been considering learning more OLLs (I know 28, for various reasons), to have a better understanding of relative merits of OLL and CLS in speedsolving.

I agree with Lucas. I like to develop BLD memo methods. I've found a few that works for me, but none are perfect. I try to avoid long word lists and keep things simple (although that's not working so well for me at the moment.) And I also like to search for OH algs.

For 3FMC, I use a method without a name that many also use.
0. Premove
1. 2x2x3 block
2. solve some edges and corners.
3. Insert at most 2 3-cycles or one parity.

For 3speed:
Slot style F2L.
2 look OLL.
PLL.
I do not like this method that much, but it works for now. I'm considering MGLS, which I might learn in the far future after I adopt Petrus first.

3OH:
same as 3Speed, but use more LUR algs.
3BLD:
3OP or M2 edges.
3OP-- can memo by visualizing clock, more thinking and better translation to freestyle/TuRBo
M2-- can memo by using random letters/straightforward and fast.
Corners are Classic Pochmann or commutators.
Commutators: efficient
Pochmann: straightforward.
4x4: reduction-- very intuitive, only 2 algs for parity.
2x2: Ortega-- easy to learn and decent speed.

I just use Fridrich because I started with it and it works well for me, oh and because I'm not too good with block-building

2x2: Ortega. it was an easy transition from LBL, and I can sub-5 with it so its fast enough for me.

3x3: Fridrich, F2L on left, opposite color neutral. Learned intuitive F2L entirely by myself just from what phenoma I observed while solving LBL. I originally solved with f2l on top, but I switched to left. I know 34 OLLs, learning the rest. I use the same for OH, but sometimes with 2 look PLL.

FMC: my basic philosophy is to try as many solutions as possible and try to get a good one. I also use lots of mirrors and inverse for OLL/PLL to try and get cancellations. Blockbuilding for F2L.

4x4: reduction, 2 at a time edges. standard method for centers, start with either white or yellow. M-slice pairing, UF buffer slot, set up both pairs before I slice. I use one "alg" and its mirror.

5x5: reduction, color nuetral centers, same edges as 4x4.

3x3BLD: classic Pochmann all around. Shoot corners with Stefans alg. Shoot edges with T-perm, unless its going to UF or UB, then I shoot with Jperms. Memo corners with words, dont have a memo for corners yet, will probably use letters.

I use petrus F2L (without edge orientation) and then fridrich LL. I started with the basic method, then started using petrus F2L. I only started learning the fridrich LL a few weeks ago though. I'm not sure I'll ever learn the full OLL - I probably will eventually - but I want to get fast with the petrus edge orientation method first. If I can do that quickly, there shouldn't be much reason to know more than the corner only PLLs I know now.

3x3x3 Roux - Just because I like that its almost all intuitive. With only 1 step that requires a Algorithm.

5x5x5 Reduction With Fridrich f2l, 2 look oll, and 1 look pll - M,U, finger tricks suck on larger cubes.

WTF! where did my post go

Corners first rulez
Why ? Because I learnt to solve the cube that way back in '81, and I still like it. I don't want to switch to anything else, I'm much more interested in trying to improve my CF and to push it to its limits. I'm already twice as fast as I was in 1981 and there's still room for improvement. My corners first is close to the Guimond method, but with a different corners orientation step. I'm trying to optimize the edges as much as possible too.
2^3: see above
Big cubes: my own cage method Starting by corners and two opposite centers (first the corners on 4^3, first the two centers on anything bigger), then all edges, and finally the last four centers. And yes it can go fast.
BLD: still not very clear, but I use 3-cycle for edges, a mixture of Guimond and 3-cycle for corners, and PLL J for parity.
Megaminx: my own method again. The star, then F2L by putting first edges then corners, then five groups of three pieces, again by putting first edges then corners, then the last 5 pairs in Fridrich style, and finally the LL in 4 steps (orient edges, permute edges, permute corners, orient corners), except when everything is oriented after the fist or second step, in which case I use PLLs to finish the solve.
Pyraminx: the usual method, first layer then one-step LL.
Square-1: corners first again , similar to Lars' method but slightly different.

Scigatt said:

I didn't want to have to start trying to speedcube by looking up an algorithm in a table and just applying it

Click to expand...

I use Fridrich, and I didn't have to look up algorithms and just apply them...

i use petrus because it was the first thing i ever tried. i didn't like the notation when i started cubing and was drawn to the moving cubes on lars' site

I just fridrich because f2l was very easy to "learn" just come to the idea of inserting corner and edge at the same time, which I thought about earlier too, I ever used intuitive f2l.
For OLL and PLL it was just the desire to get faster... and the very beginning was after learning 4x4 while using fridrich f2l and real beginners last layer(E-OLL E-PLL (sune) C-PLL (niklas) C-OLL (R' D R D')), ending with PLL parity (2 swapped corners) so the T perm was very useful...
I also looked into petrus, roux and mgls for fun, but I sticked to fridrich, now I want to look into some CLLs and ELLs too see how good the recognition works for me and can improve my times at some situations, also maybe learn some OH algorithms finally... (just H perm, because M turns pisses me off, but still got OLLs with it..)

3x3x3 : F2L, 2-look OLL (because I'm too lazy to learn them all ), PLL.
(same thing for 3x3x3 OH, but I don't know all the PLL OH)
Because I first learned a layer by layer method, so it's natural to continue with F2L.

2x2x2 : Ortega; because easy to learn, not a lot of algorithm

4x4x4 to 7x7x7 : reduction, because I found it natural (except the orientation parity case)

Megaminx : face by face. With a loooot of F2L pairs. and 2-look OLL and 2-look PLL. Because it's the more natural thing with my 333 method. And because it's fast

Square-one : same order of steps than Lars. But I never practise square-one so each competition, I have to improvise each step I don't even know the parity case (when I have parity, I know I can do something to not having parity anymore, but it scrambles my corners too XD)

3x3x3 bld : I use old pochmann, because I found it really easy to understand when I wanted to learn blind.
At the beginning, I wanted to find myself how to blindfold a cube, so I had a personnal method (wich looked liked 3 cycle but were not exactly) to solve corners, but I didn't have time to finish my method for edges (which also looked like 3-cycles)

CharlieCooper said:

i use petrus because it was the first thing i ever tried. i didn't like the notation when i started cubing and was drawn to the moving cubes on lars' site

Click to expand...

same here. I learned all of my algs by remembering where the pieces moved instead of just remembering the moves.

but yeah, There are a few reasons why I use Petrus method. the first being the lack of algorithms. 7 algs for OLL is amazing. and it also leaves space for other sets of algorithms like COLL and CLS. After flipping bad edges the rest of the F2L is 2gen which is very fast/fun. I just find petrus way more interesting than fridrich and roux.

2x2x2_method: color-neutral Layer-by-Layer, Ortega or freestyle depending on the scramble.
2x2x2_reason_for_method: Color-neutral because you get easier starts, Layer-by-Layer because chances of 1 look last layer are high, Ortega because it's only 3 logical algorithms to learn and gives nice short solutions, freestyle because sometimes I just see a FMC solution during inspection

3x3x3_method: white cross, 3 corners, 4 edges, last corner (my keyhole variation), optional edge-control, 2 look OLL, 2 look PLL
3x3x3_reason_for_method: Easy to learn after a beginners method. To busy/lazy to learn full OLL/PLL, still faster for me than CFOP F2L

4x4x4_method: 1st center, opposite center, 3rd center, adjacent center, last two centers, 2-at-a-time edge-pairing, finish like 3x3x3. For OLL-parity I always do a double-parity-fix I found myself.
4x4x4_reason_for_method: By far the fastest method in the world and entirely intuitive

5x5x5_method: Centers like 4x4x4 order using 3+3+3 lines or spiraling, edges with the absolutely brilliant AVG method , finish like 3x3x3
5x5x5_reason_for_method: Also the fastest method in the world and also entirely intuitive (except for a faster edge-parity-fix I learned)

6x6x6_and_bigger_method: Centers like 5x5x5, Edges like 5x5x5 working inside-out, finish like 3x3x3
6x6x6_and_bigger_reason_for_method: By now you should have figured out that I like intuitive, yet fast.

Square-1_method: A=(1,0)/(-1,-1)/(0,1) and B=/(3,0)/(-3,-3)/(0,3)/ + a parity fix (nerdsparadise)
Square-1_reason_for_method: Did I ever mention intuitive, yet (reasonably) fast?

FMC_method: Solve as many pieces as possible in as few moves as possible. Add pre-moves and in-between moves to solve more pieces. Use insertions to fix all those nasty pieces that didn't get solved before
FMC_reason_for_method: Doesn't require any algorithms and gives great results

Megaminx: I suck

I use fridrich because it's faster than the beginner's method and I haven't taken the time to research other methods yet. I chose fridrich because it is what most other cubers use and what people recommended I try next to improve speed.

AvGalen said:

4x4x4_reason_for_method: By far the fastest method in the world

Click to expand...

I definitely don't agree with that statement.
I get better times at big cubes than cubers who have the same basic speed as me and use reduction

TMOY said:

AvGalen said:

4x4x4_reason_for_method: By far the fastest method in the world

Click to expand...

I definitely don't agree with that statement.
I get better times at big cubes than cubers who have the same basic speed as me and use reduction

Click to expand...

Can you name 1 top 10 4x4x4 cuber that doesn't use reduction? Somehow "fastest in the world" and you or me don't seem to have much in common

My turn!

2x2: I don't really have a method, just use simple LBL with OLL & PLL and uses J perm and E perm for the permutation part. Fast algs!

3x3: Fridrich!!! Just a few more OLLs to go, but I don't have the boost to learn them. Found them on google, but was put off by the number of algs. But then i found dan harris' website...
I wanted to learn CLL and ELL but I couldn't find a page with the algs on it untill... it was too late.

4x4: Reduction centers; fridrich finish: Purely Intuitive, i love it. Six pair for edges.

5x5: Reduction centers; fridrich finish: Same, but I use a method i found out myself for the edges, pairing one whole tredge at a time.

AvGalen said:

Can you name 1 top 10 4x4x4 cuber that doesn't use reduction? Somehow "fastest in the world" and you or me don't seem to have much in common

Click to expand...

That doesn't prove anything. Most cubers use reduction because they were told to do so and never seriously tried anything else. And that includes the fastest ones.