I learned how to solve a Rubik’s cube with the normal beginner method. Now I wanted to know how to get from this to speedcubing. Is speedcubing just executing the steps really fast? I don't think so. What are the fastest ways to solve the cube?
The answer quite honestly is: depends on how much you want to memorize.
I'd personally recommend exploring the algorithms and why they work before delving into speedcubing algorithms. The reason for this is simple: over the use of the cube, you'll see faster ways to do what you're trying to do. My solve time using the beginner's method is between 30 and 40 seconds (PB avg of 5 for me is about 24-6, if I recall?).
Also, do remember that a lot of it does actually have to do with how fast you move. Never underestimate the impact of training yourself to build up sheer speed in execution - but also never forget the benefits of muscle memory.
That being said, the ways to truly speed up your solves are with memorizations of specific steps. And to reiterate, this all depends on how much you intend to memorize.
The Fridrich Method is widely regarded to be one of the fastest modern methods for solving the cube, though it is certainly not the fastest. In essence, it relies on the cross, then F2L, then OLL, then PLL, and takes a significant amount of memorization to execute in full.
- The Cross is what you're familiar with from the basic solve.
- F2L stands for First Two Layers, and indicates a concurrent solve of both of the first two layers. This revolves, in essence, around pairing edges with corners and inserting both at once. There are a decent number of algorithms associated with specific conditions, all designed for the shortest path.
- OLL stands for Orientation of the Last Layer, and is, in essence, a dictionary method that orients all the pieces on the last layer. Lots of documentation is available.
- PLL stands for Permutation of the Last Layer, and is also effectively a dictionary attack on the permutation of pieces. This is slightly less intensive than OLL.
However, these are the extremes. I'm using these to illustrate a point that the more you memorize, the faster you'll get. You can find a list of methods on the speedcubing wiki. There are a lot of intermediate steps, though, which can help you along your path.
- F2L doesn't necessarily need piles of algorithms. You can learn partial solutions, and intuitively guide yourself to cases you know.
- You can split OLL into two steps: one to orient edges, and one for corners.
- You can do the same for PLL.
In essence, the more shortcuts you discover and memorize, the faster you'll go. I definitely recommend looking through the speedcubing methods wiki page, as it contains further useful information.
You need to use better methods like the fidrich method.
Cross - Find a faster way to solve the cross. After you have learnt to solve the cross quicker, try doing it on the bottom to avoid cube rotations
F2L - After you have practiced F2L, learn some advanced F2L techniques. Especially try to look ahead, meaning when you know what to do to create one slot, while you are creating it and inserting it, look for another corner and edge and keep track of where they are, so when you go to that corner, you won't have trouble looking for the ege to pair it up.
OLL - Try doing this in two steps. The first is to orient the edges with an algorithm , and then to orient the corners with an algorithm.Here is a link to a page that will show you the algorithms. http://badmephisto.com/2LookOLL.pdf
PLL - In the beginners method, it will teach you to do the PLL in two steps,only learning 7 algorithms. This is OK, but it is twice as slow as just learning the 21 algorithms (Try only to learn 1-2 a day, or your brain will overfill with algorithms) which you can find here. http://badmephisto.com/pll.php
1 look OLL (This is only optional to get sub 8): When you have a very good average at about 15 to 20, if you want to improve your average to sub 10, learn the ful olls, which are 57 algorithms (Again, to not overfill, only learn one a day) which you can find on the internet.