In terms of "quick progress", PLLs are your best bet. Grind them though. I never did and mine remain sloppy as hell. Those are the big algs. Once you have those down, OLL becomes a mere afterthought. You need to follow through with both if you can spare the time though. OLL is the edge in CFOP IMO...and rotationless F2L, multi-slotting, "CN", x-cross and co. So much to learn and master.
 
Last edited:
Hi Zarxrax,
I read this message and I am quite interested in using this ANKI program to aid with my memorization. Do you have any cards you developed that I could get from you? I found a little bit of stuff regarding PLL on a Rubik's cube but not much else. I really want to learn full OLL and PLL and whatever else I can work out. I am really interested in starting speed solving and I think this could really help. Anything you could help with would be appreciated. Thanks.

I only started learning the 2-look OLL and PLLs yesterday, but it doesn't look so hard. I consider myself to have a very poor memory, but I enjoy learning languages and have picked up some handy memory tricks and techniques over time. Because I have only started learning the algorithms, I'm not sure how much some of these will apply, but these are the areas which I am exploring.

1. Learn to group common elements.
Look for sequences of moves that appear in multiple algorithms. For instance if there is a sequence of four moves that appears in 4 algorithms, then you can either learn all 4 moves 4 times... or you can learn the sequence one time and then apply the whole sequence for the other algorithms. By looking at a sequence of moves as a single group rather than individual moves, it makes things much easier on you when you come across that same sequence again. I managed to memorize how to write 2000 Japanese kanji in only 3 months using similar techniques.

2. Develop visual mnemonics to help memorize some moves.
For instance, one of the algorithms I use starts out with "R2, D2..."
Hey, R2D2 is that robot from star wars! I can think of that when I do this move to help me remember. You can get creative.

3. For similar moves, develop mnemonics to help you keep them distinct.
For instance when doing PLL, I use the algorithm F2 U R' L F2 R L' U F2 to cycle 3 pieces clockwise, and F2 U' R' L F2 R L' U' F2 to cycle them counter clockwise. How do I keep it straight as to which is which? Well notice that the only difference in the two algorithms is that one contains U and the other contains U'. It just so happens that when I want to do the move which cycles the pieces clockwise, then I simply have to turn the U face clockwise! This makes it extremely easy for me to keep those two algorithms separate.

4. Spaced Repetition
Spaced repetition is an incredible technique that has been the number one most effective thing for helping me to learn and retain information. I'll spare you the gory details, but it basically goes that if you learn a new piece of information today, whether you study it 1 time or 100 times, you are still likely to forget it a week from now. However, if you can study it today, then tomorrow, and then 3 days after that, and then a week after that... you can retain the information with minimal time spent studying. Using spaced repetition software such as Anki, we can make this process very simple. It works essentially like flash cards. You put a question on one side and an answer on the back. For the question side, you can put an image showing a certain situation, and the answer would be the algorithm that should be used to solve it. Or perhaps the question side can be an algorithm which will set up the situation for you to solve, and then you can physically solve it on your cube.
Spaced repetition is a technique for remembering large amounts of information over a long period of time (ultimately, forever). Some algorithms are pretty common and you may not need to do this with them. However some algorithms apply to situations that might only come up under very rare circumstances. Are you going to simply remember those cases when they happen to come up without refreshing it regularly?

5. Initial Memorization
While spaced repetition is useful for memorizing data over a longer period of time, the software isn't so good for initially learning. However the same techniques can be applied by yourself. When learning an algorithm, do it a few times while reading the algorithm to get a feel for it. When try to perform it from memory. If you cant get the whole thing, try to get as many as you can, and keep trying until you have the whole thing. Then wait a minute and try it again. Wait 5 minutes and try it again. Wait an hour and try it again. Wait until tomorrow and try it again.

6. Muscle memory vs regular memory
Your ultimate goal is to get the moves into muscle memory. This requires doing the moves hundreds of times. This is a slow and boring process so I do not recommend trying to learn algorithms this way. If you commit the algorithms to your normal memory, then over time, as you perform them while solving the cube, they will eventually move to your muscle memory without any additional effort on your part.

7. Consistency
When learning any large amount of material, the best way to make it all the way through is to develop good habits. Try to learn at least 1 new algorithm every day. If you can't learn one, try to at least review your old ones. Don't get into a situation where you do nothing. Once you go one day without doing something, its very easy to skip it the next day as well. Before you know it, you haven't reviewed or learned anything in a month, and you forgot everything. If you have trouble with this, it is useful to set aside a specific time each day, such as right after you wake up, or before you go to sleep. This helps you to develop a routine.
 
Last edited by a moderator:
OLL & PLL help?

Hey guys I've got a sub-40 (not very impressive) average (25.10 personal best), and I'm trying to find strategies to memorize PLL algorithms, as my multiple uses of corner swapping and edge rotations is too slow for a quick solve. Also I am looking for some of the OLL algorithms that you find come up a lot because I have a pretty quick OLL so far. Any look-ahead strategies would also be greatly appreciated, as I know that's important for a sub-20 average.
Thanks!
 
look ahead strategy #1 : go slow
Alg memorization strategy : pick one, do it a lot, do it some more, after that, do it even more

what helps is just doing it alot so you get a feel of how the algorithm is executed and so it becomes muscle memory.

edit: inb4 "or you can just learn roux"
 
Yeah, just do the alg many times.
Almost all OLL's have same chances of coming up, just start with easiest.
And you should start lookahead in F2L at about sub-30, now you should work with recognizing cases and doing them without looking/thinking with as small movecount as you can.
 
Thanks, appreciate the help, I currently know all but 1 OLL with the cross, so I usually solve the cross before OLL, want to change that though because it takes time.
 
Once you learn two look OLL there are a few OLL's which are fairly easy to learn straight off the bat which involve the R U R'U' trigger. Learn the T-shaped OLL's first and then go for some of these ones which use the above trigger:

http://badmephisto.com/oll.php

or go on the speedsolving wiki for alternative OLL's.
 
OLL to PLL transition (look-ahead)?

I have watched Feliks' and other worldclass cubers solves and i realized something... It looks like they know which PLL will emerge when they solving OLL..
It's impossible! Whats the tip of Recognize PLL during OLL ? if i know it , it will help me to reduce my time 1-2 seconds... BTW my ao5 is 14 seconds
 
You can cut time from your PLL recognition by familiarizing yourself with CP. It's really useful to be able to recognize OLLCP cases (you can do this easily if you already know CLL for 2x2x2, btw), and to know how each of your OLL algs affect the CP. By being able to know which CP case your PLL will have (either: 1. all corners solved, 2. two adjacent corners solved, or 3. two diagonal corners solved), you can narrow down the possible cases you can have before you even get to it.
Other than that you can watch blocks, headlights and other things form during the last few moves of the OLL.
 
Last edited:
When to learn PLL (and OLL) algorithms?

I have read about this many times. I actually do not understand the reasoning behind putting off learning PLLs (or OLLs). Yes, I know F2L is going to be the biggest difference so most of the time you spend practicing should be invested there. But if you consistently learn LL algs, it will be better in the long run. Imagine that you take a short time each day of your practice time to learn algs. Once you learn full PLL for example, all the solves you do after that will help you practice recognition and execution of those PLLs. If you decide to put off learning algs, it gets easier and easier to keep putting it off.

So let's say you start out with CFOP and only do 4LLL. You put off learning PLL and OLL because F2L is important, right? 10,000 solves later, your F2L is better but you are still only doing 4LLL. Then you finally decide to learn all the algs. Only after that, the solves you do will help you practice those algs that you never practiced before. If you started off learning the algs or at least did them slowly but consistently, you could have been practicing already during those 10,000 solves from before.

That said, I think it would partly depend on how easy it is for each person to learn the algs. Personally, I started off my early CFOP days by learning all 21 PLL in the space of several days. (I have pretty good muscle memory and I have my own method of learning and practicing them.) Then I learned all 57 OLL not too long after that. It took longer of course but all solves since then have let me practice the full thing. F2L takes time to master anyway. I can learn algs while I get used to F2L. If you put off learning LL algs to when your F2L is decently fast, think of all the LL you solved that were a waste because you were not practicing those algs then.

Any thoughts?
 
Yes, but if beginners learn algs early, they will put emphasis on algs making them faster. Learning to do F2L efficiently is quite an effort, and if someone can learn F2L and new LL algs at the same rate, good for them, but imo and experience a few common PLLs (H, Z, T and Y) are enough to solve the cube with CFOP, regardless on how many look the P takes. Your point about doing 10000 solves and practising the LL at the same time is a good one, but doing LL quickly is just recognition/prediction and tps, whereas F2L is all about lookahead. If someone early in their cubing career tries to do both at the same time, they might try to apply tps to F2L, which obviously isn't going to work. Besides, a fast F2L will allow plenty of time for LL to be dealt with in a solve, which will still allow for relatively faster times. Just my two cents, I didn't know full OLL/PLL until I was averaging ~17s.
 
Well.. as soon as I was sub 50 or so I started to learn full PLL. It took me about a month to learn all of them Now, a few months down the road I have all PLL's in muscle memory. No problems there.. I started to lean OLL's now. But man.. pfew that is tough. For example there are 6 lightening bolt cases.. It's hard to see which case you have, and then it takes for ever to think what alg to execute since they all look alike. In PLL most algs are very unlike each other (except the freaking G-perms). Learning all OLL's is something I really want, but the fact that I keep forgetting even the most simple cases like r U2 R' U' R U' r', r U R' U R U2 r', r' U' R U' R' U2 r and r' U2 R U R' U r every day again and again makes it so discouraging. I think you do have a great point there that if I was to learn OLL a few months ago, I would have been past this problem and probably have all of them in muscle memory.. I would be much faster than I am now..

Btw, SneaklyFox, when where your 'early CFOP days'? I want to be as fast as you ;)
 
I also started learning Full LL relatively early after I knew 4LLL. First I learnd all PLLs (except the G-Perms) and then started with OLL. Fot that I used a german Youtube series which teaches you full OLL in 30 days.

Then I took a 10 month cubing break and had to start over again. Currently I'm learning full OLL again with the mentioned youtube videos, but only one alg or a group (like the two L-cases) per one or two weeks, because I try not to speed through all 57 algs and then forget 40, but to keep the ones that I've learned.

What I do is to use an already learned OLL in the solve, even if it takes 10 or more seconds to recognise, remember and execute. That was something that I didn't do before the break because I wanted do make good times, but I then forgot far more algs than now.

So what I recommend ist to start early but learn only one or two at the time and not rush through all of them in a short period of time.

Here's the videos I use for learning full OLL: http://www.youtube.com/playlist?list=PL38A6C0D22DF2BB0C
 
Totally agree with OP, if you're planning on knowing all of OLL/PLL in the long run, then the time you spend practicing a method you're not planning to stick with is almost wasted time. I usually recommend people learn them as they feel comfortable, if you force yourself to learn lots of algs and aren't having fun, it will probably make you less likely to stick with it. But yeah, there is certainly no need to wait.
 
There are a couple of reasons why they are often put off. Primarily its so that someone learning the full method doesn't get put off by the sheer amount to learn to do it "properly". 21 PLLs, 57 OLLs, and 41 F2L cases from 4 different angles (taking full advantage of empty slots where possible and minimising rotations) and an efficient cross is the bare minimum for mastery. Add in learning some PLLs/OLLs from different angles, and a bit of X-Cross/WV/Magic Wondeful/VHLS/RLS/COLL/Jeff2L/etc. This is a huge amount and is just off putting if there is not some stepping stones on the way. Hence learning all f2l algs in the FR slot, and 2-look OLL/PLL, as a stepping stone to further learning and practicing.

It is often recommended to put off OLL and PLL learning, instead working on F2L as a main focus. This is because this is simply a more efficient use of the beginners time. Full PLL will take maybe 2-4 seconds off a 40 second solve, whilst concentrated work on F2L can take 20 seconds off. Thus time spent on F2L is more useful.

When you get into cubing, and are perhaps at stage when you become aware of whats out there, you start making decisions as to what you want to learn, based on an informed choice, and you learn it at your own pace. If you are just starting out, the recommendation to learn a 4-look last layer at first is a good one, as it keeps people in the game because its achievable, and keeps early focus on F2L where its most useful.

We have all seen vids of sub 15 times with 4-look last layer. I have never seen a sub-15 solve with "no lookahead, always solve FR, rotate 8 times a solve" F2L.
 
Any recommendations on a website/video where you can learn easy or fast OLL and PLL cases (Since there are some algorithms which can be ridiculously long and difficult)?? Thanks! :)
 
Learnt full PLL since sub 30. (4-look LL before sub 30).
Then start full OLL since sub around 23.
Currently avg-ing around 16/17. Still, I can easily sub 20 with 2L-OLL. I find that improving the look ahead seems to be a higher priority before learning more PLL or OLL. Though learning them along the way do have their advantages.
Since F2L is intuitive for me, it gives me better understand of cross-building too. Which improve both my cross, F2L/look ahead at the same time. Then the rest would be LL, which is tps/algs reliant but I wouldn't say the LL take a bigger part in a solve than cross and F2L, due to that there are chances of PLL or OLL skip. Which leaves you with CFP or CFO. Where its more unlikely to have skipped cross or F2L.

I'm still happy I learnt PLL along the way to sub 30 and to sub 25, but it makes me struggle to get sub 20 (Maybe just me). I think I've been relying on the algs too much back then.
 
Back
Top