How can I improve?

I can improve my skills in terms of things such as:

Lookahead - The ability to decide which moves I'm going to do for the next step, while I'm doing the current step. For example, if I'm solving the cross, I could use lookahead to think about how I'm going to solve the first pair.

Colour-neutrality - The ability to solve the cross on any face, and therefore do any step from any face. This is useful because there can be an easier way to solve the cross from a different face than white, and there might be an opportunity for extra pairs or other advanced F2L techniques. I have tried to learn this in the past but it usually has more negative impact than positive for me.

X-cross - X-cross is a specific situation where 1 or more pairs can be made while solving the cross. This can decrease the amount of time to solve F2L.

Pseudo-Slotting - Solving multiple pairs at once using already half-completed f2l pairs

If I want to get better at solving the cube, I plan on optimising each stage of the process; I can do this by using the following methods:

Cross:

• Practice solving the cross in as few steps as possible.
• Practice memorising how I'm going to solve the cross during inspection.
• Memorise the cross solve during inspection and execute it blindfolded.
• Practice speeding up this process as much as possible.

F2L:

• Learning the quickest solves for specific cases.
• Practice doing these solves without looking

OLL:

• Learning all 1-look algorithms for the OLL stage. This means that I can use 1 algorithm to solve this stage from any combination. I currently know 2-look OLL, which means that I can always solve this stage within 2 algorithms at most. 2-look requires 10 algorithms, whereas 1-look requires 57 algorithms; I currently know 21 of them, so I have to learn 36 more algorithms if I want to cut the time of just one extra algorithm.

PLL:

• Learning all 1-look algorithms for the PLL stage. This has the same benefit as at OLL, but has 21 algorithms so is easier to learn. I currently know 17/21 of these.
• Practicing all algorithms, so I don't waste time trying to remember which to use - they should become muscle-memory.

In summary, if I learn 36 more OLL and 4 more PLL algorithms, I can have a 2-look last layer, meaning I use just 2 algorithms for the whole of the last layer. It may take a while to learn 40 more algorithms, so this is a goal for the end of my project, not a quick fix. I also need to keep up to date with the algorithms I currently know, so that I don't forget them.

I can mix these together and learn separate parts which work well together at the same time. For example, I can practice doing the cross and F2L at the same time, and learn to lookahead to my first pair during inspection; this will allow me to improve both aspects of the solve at the same time, and create a fluid movement from one to the other. Another great match would be the OLL and PLL which make up the last layer.