Not sure how to learn or practice? Navigate through some techniques that I frequently use.

Practicing the first layer is pretty straight-forward. Start with the easiest first layers in my First Layer Trainer (meaning 1 move, just at one colour and non-shuffled). Then, gradually increase the difficulty by adding more moves, become colour-neutral by selecting the "Use all colours" checkbox and shuffle the scrambles.

If you want to find out about efficient first layers, or have no idea how to solve specific first layers, you might be interested in Elias video plus the First Layer tab of the Skewb Resources doc. Try to make sure that you understand what's going on, and immediately try to mirror the layers you just learned.

For the last two layers, there is a lot more to say. In general, the transition from Beginner's to Intermediate to Advanced to eventually learning additional algs takes some time. This is a fact and you should use every step so long until you master it completely and want to step up your game.

For L2L algs, there is a recommended learning order available in the Skewb Resources doc mentioned earlier. I think they really did a good job on that, and I kind of intuitively chose the almost same order without knowing about this document before.

My typical strategy for mastering algs is the following: I choose a set I want to work on, e.g. Pi + Swirl Perm. This means when learning Advanced or Algs, there are eight cases that look similar and can be confused with other types and even among each other. Find out what all cases in one set have in common.

Find out how the cases in this set are related to each other, both by visually studying the scramble image + the Advanced Algs listed with them.

Start with a smaller subgroup of one set and learn e.g. four cases at a time, by trying out the solutions, choosing the ones that fit well to your turning style and repeating them until you have them in your muscly memory. Of course, it is neccesary to understand the Fixed Center Notation & Rubiksskewb notation on top of the Fixed Corner Notation. Watch this video if you have not yet understood the Rubiksskewb notation. Smaller movemount does not automatically mean better solution! Take into account if there are triggers (common three- or four-movers) such that you do not have to memorize each and every move in one alg. A bunch of them are just a combination of e.g. Sledges or Hedges + few other triggers.

Start timing your solutions and keep track of your pbs for each case! Also think about filming the execution such that your future self can compare it to other skewber's exe. That way you will remember how you do your algs and will be able to implement better fingertricks or even change to other algs that are better.

When practicing L2L, sometimes I feel like just practicing a few particular cases where I feel like doing bad at (Select by ID), sometimes I feel like doing one complete set for a few minutes. But don't forget that by selecting only a few cases, your brain already knows what type you are going to receive. So please also always select more than one set if a competition is approaching. In this situation, you will first have to recognize the type and subtype of the case before you can identify what exact one you'll have to apply. This should be included in your practicing routine frequently.

You could conclude that L2L will require you to change the type of practicing strategy a few times during one practicing day, and that you'll have to repeat the rather fresh algs more frequently than the ones you already know for a long time. So far, adaptiveness or spaced repetition is not included in SkewbSkills. But you can guess that this would be the optimal way to memorize and practice the algs, and I plan to include these features in the future.

Recently I added a one-looking trainer for SkewbSkillsWeb. It is a plain combination of both previous trainers, and after the application of a L2L scramble, a FL scramble is added. You can set the difficulty by adding between 1 and 7 moves to "destroy" the first layer. There, you don't have the option to select the cases, because I suppose the ones that use this feature are already familiar with all L2L cases and want to go one step further.

I did not have time to really use this new feature, but so far I feel like just starting out with only one move drastically helps me using inspection wisely. You are forced to solve the Skewb blind you could say, and it requires some concentration. Practice makes perfect, so don't hesitate to start now!

Some general advice now covers competitions: one thing I did last year was to film my solves and analyse them, just privately for me and not uploading anywhere. Even by just summarizing the pros and cons of each solve in a stress situation helped me understand how I think and react during an official average of 5.

Some other interesting advice is that you should only practice when you feel like it. If you just want to spam solves, and you know that you are going back to your normal practicing routine the next day for example, then just spam solves! Noone (not you yourself) should force you to do long learning sessions each and every day. Just don't forget to regularly practice your weaknesses and steadily implement a few new algs or tricks. You might already know, but I'll just repeat that doing really effective practicing just a few minutes a day will in the long term help you more than cramming dozens of cases each day. After all, skewbing should be fun and not something that tires you or makes you feel bad during or after practicing.