Understanding the 3x3

[_D2_]

Hi guys, his thread is not about how to get faster but i'm going to make a comment, I need/want to get faster because i average 1:20minfp) but before that i feel like i'm a robot solving a cube so i would like to know commutators, permutators and things that help me understand the cube, understand the algorithms. I would try to solve it intuitively but i got the algorithms stuck in my head:fp.
I ordered a 2x2, guhong, 4x4, sqr-1 and pyraminx and i'm going to toy around with them to see if it can help me understand puzzles overall too.
So what i'm asking is how do you create algorithms, any tutorials, i have Joels guide here but i need more, about orientation too and 2-cycles, how do you watch a set of cubies(the ones you need to restore later) and the ones to orientate or permutate. If that factorial and permutation math things are good to know, i'm learning them next week. I want to know this stuff to stop feeling like a robot and hopefully create my own algorithms and maybe methods too.
Bonus: So i know practicing is the only way to get faster, but like 100 solves a day sounds a bit boring, any way to spice it up ?
Ps: if this wall of text is too much for you i can organize it
EDIT: Found a theory for creating algorithms (or macros that solve the cube as said in the text) here is a small part, the most important one. I didn't read the whole text yet, stopped at cycles.

3.1 Order
The question is, how do you discover these macros that do very small, specific things and
leave most of the cube unaltered? It turns out (and we shall see why later) that if you have a
macro with a certain order and you apply it for half or a third of that number of steps, the result
is often a usable (although usually very long) macro.
As an example, consider the FFRR macro that we experimented with before. We found (in
three different ways, hopefully) that the order of this macro is 6. The “large” divisors of 6 are
2 and 3, so you may find interesting macros by repeating the FFRR combination twice or three
times.
To do this, reset the cube and type FFRR into the “CurrentMacro” window. If you press the
Apply Macro button twice there’s a sort of a nice pattern, but it moves far too many facelets to
be useful for solving the cube. Press it a third time, however, and you’ll see that the net result is
that two pairs of edge cubies are exchanged and everything else remains exactly as it was before.
Thus FFRRFFRRFFRR might be useful to you as you’re solving a cube.
It’s easy to look for such macros. Simply type in various (usually short) sets of moves and
find the order of that operation. If that order is divisible by a small number like 2 or 3 or perhaps
5, try dividing the order by that number and applying the macro that number of times. There
is a shortcut for doing this. Suppose you find a pattern that repeats every 90 moves (the macro
FFLLBR, for example, has order 90). If you want to see what this does to the cube after 45
moves (for a total of 45×6 = 270 moves, which would be fairly painful to use), you can simply
type the following into the “Current Macro” window: “45(FFLLBR)”. A number in front of a
group of moves in parentheses tells Rubik to repeat the stuff inside that many times. These
groupings can be nested, but this will not be too useful for finding cube-solving macros.

Obtained @ http://www.geometer.org/rubik/group.pdf

 

[irontwig]

You could try to look at Petrus and then Heise:

http://lar5.com/cube/index.html
http://www.ryanheise.com/cube/

Not only does these methods give you more of an understanding of the cube, they also lead to efficient solutions.

 

[_D2_]

Yeah i know those, i was looking more for a guide to create algorithms, i see lots of different algorithms for one case, so i was hoping to make one that suit me better,etc; and also create a crazy method just for fun. Imagine like i didn't know a method, i want to create one or for example do fridrich without knowing it existed, i want to understand algorithms.

 

[Chrish]

Asking for a guide to create algorithms doesn't make sense.

With algorithms that you currently know.. do them repeatedly, slowly, to see what exactly you are doing to the pieces. When you get a better understanding of how to move pieces around without disturbing a certain portion of the cube.. then you can experiment with personal algorithms.

I've came up with a U perm that I've never seen before, but I haven't tried to do anything more.


edit--

Example, OLLs are literally taking out an F2L slot, moving the top layer, then inserting that F2L piece a different way than you inserted it. There are a few exceptions, but that's an algorithm.

 

[_D2_]

Asking for a guide to create algorithms doesn't make sense.

With algorithms that you currently know.. do them repeatedly, slowly, to see what exactly you are doing to the pieces. When you get a better understanding of how to move pieces around without disturbing a certain portion of the cube.. then you can experiment with personal algorithms.

I've came up with a U perm that I've never seen before, but I haven't tried to do anything more.


edit--

Example, OLLs are literally taking out an F2L slot, moving the top layer, then inserting that F2L piece a different way than you inserted it. There are a few exceptions, but that's an algorithm.

I'm not asking for a guide only about creating algorithms, maybe there is a guide to another thing that may teach you what the algorithm does, how he created one, etc; just like you did in that bold paragraph. But does every OLL take the f2l pair and never break it to restore it later?

 

[Radcuber]

You're getting a GuHong when you're averaging 1:20? :fp You won't really get the full benefits of a DIY until you're at least sub -40....

 

[_D2_]

You're getting a GuHong when you're averaging 1:20? :fp You won't really get the full benefits of a DIY until you're at least sub -40....

It's actually only getting here in a month, and i bought it because i was buying other cubes and the store had it. Also you're very rude.

 

[irontwig]

Yeah i know those,

That could mean everything from you simply being aware of the methods' existance to you fully mastered the methods, since you're not even sub-60 I presume you're closer to the first, or you have a really bad cube.

i was looking more for a guide to create algorithms, i see lots of different algorithms for one case, so i was hoping to make one that suit me better,etc; and also create a crazy method just for fun. Imagine like i didn't know a method, i want to create one or for example do fridrich without knowing it existed, i want to understand algorithms.

http://www.ryanheise.com/cube/commutators.html

 

[Stefan]

i know practicing is the only way to get faster, but like 100 solves a day sounds a bit boring, any way to spice it up ?

Listen to CubeCast while you're doing it.

 

[flan]

You're getting a GuHong when you're averaging 1:20? :fp You won't really get the full benefits of a DIY until you're at least sub -40....

DIY's get more necessary the faster you get but they benefit anyone who is using finger tricks and such. Anyway why does it matter. A guhong bought now is a guhong not bought later

EDIT: you have ordered a guhong and you are barely sub40, thats kinda hypocritical since you did say atleast sub 40

 

[some1rational]

like irontwig mentioned

if you want to make up your own algorithms, learn/understand commutators and conjugates (both on heise's site I believe)

 

[Radcuber]

You're getting a GuHong when you're averaging 1:20? :fp You won't really get the full benefits of a DIY until you're at least sub -40....

DIY's get more necessary the faster you get but they benefit anyone who is using finger tricks and such. Anyway why does it matter. A guhong bought now is a guhong not bought later

EDIT: you have ordered a guhong and you are barely sub40, thats kinda hypocritical since you did say atleast sub 40

lolno. Because I AM sub -40? It doesn't matter if you're barely sub -40 or not.

 

[cincyaviation]

You're getting a GuHong when you're averaging 1:20? :fp You won't really get the full benefits of a DIY until you're at least sub -40....

DIY's get more necessary the faster you get but they benefit anyone who is using finger tricks and such. Anyway why does it matter. A guhong bought now is a guhong not bought later

EDIT: you have ordered a guhong and you are barely sub40, thats kinda hypocritical since you did say atleast sub 40

lolno. Because I AM sub -40? It doesn't matter if you're barely sub -40 or not.

Yet you had a DIY before that...

 

[blah]

so i would like to know commutators, permutators and things that help me understand the cube

Permutators are easy. Just, like, yeah.

 

[tx789]

Megaminx pll

I got a megaminx got to the last layer bu my self and i need a good clear corner pllhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gif

 

[maggot]

creating your own algorithms comes from a basic sense of understanding commutators. it helps to understand this mathematically, however a good way to learn them is to practice BLD and get more involved in complex BLD methods. freestyle methods are very user specific and i think this is what you are looking for. basic knowledge of the 3x3 is usually understood when trying to teach BLD methods. even though it may be painful for you to grind PLL and OLL algs, i think it would be easier to understand if you did. if you are ready to immerse yourself in things you probably wont understand, http://www.speedsolving.com/forum/showthread.php?p=310411 is an explaination of BH commutators. With an understanding of this, it would be simple to create your own algorithms, even though by the time you did understand this, you would think of things as setup move, switch, undo setup move, instead of an alg.

 

[Kirjava]

Commutators aren't the only way to create algorithms

I got a megaminx got to the last layer bu my self and i need a good clear corner pllhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gifhttp://www.speedsolving.com/forum/images/smilies/confused.gif

Post of the year.

 

[_D2_]

creating your own algorithms comes from a basic sense of understanding commutators. it helps to understand this mathematically, however a good way to learn them is to practice BLD and get more involved in complex BLD methods. freestyle methods are very user specific and i think this is what you are looking for. basic knowledge of the 3x3 is usually understood when trying to teach BLD methods. even though it may be painful for you to grind PLL and OLL algs, i think it would be easier to understand if you did. if you are ready to immerse yourself in things you probably wont understand, http://www.speedsolving.com/forum/showthread.php?p=310411 is an explaination of BH commutators. With an understanding of this, it would be simple to create your own algorithms, even though by the time you did understand this, you would think of things as setup move, switch, undo setup move, instead of an alg.

Commutators aren't the only way to create algorithms

Thank you very much, i think i can fetch things on my own now, however if anyone would still like to help feel free to do so.

 

[pistelli]

I agree with the first poster, I think that it would be neat for a sort-of algorithm creation guide. It probably wouldn't show you how to create and algorithm but how algorithms were created. By the way, how did the first cubers develop algorithms? Did they do it just by playing around with the cube or did they use mathematics?

Also, talking about speed cubes. I don't think it is necessary to be sub anything to get a speed cube. I'm 50-60 seconds average on a storebought cube and just bought a ghosthand 2. It isn't here yet, but I think diys would help you even more when trying to learn to get faster because you aren't learning on a crap cube.

 

[Brian Kremer]

Although it is hardly a tutorial, you might check out chapter 14, "Magic Cubology", in Metagagical Themas by Douglas Hofstadter. It explains the idea behind making algorithms.