[Help Thread] Algorithm Memorization Discussion

[Nilsibert]

I usually learn them in sets based on how they look. and I give them attributes in my head for when they look similiar. For example the knight cases in OLL. There's four, and the way I learned them was I grouped them into "good and bad". Good just being the easier alg. So I have right-good, left-good, right-bad and left-bad. Hope that wasn't too confusing.

Generally, I learn a set of algs, for example CLL Pi, until I can kinda do them. Then I just do solves and force myself to recognise the case. Over time, I'll get used to all the cases. So rather than learning one alg until it's second nature and easy muscle memory, I "half-learn" a bunch and they come with time. Not sure if I'd recommend that strategy though; It took me quiet some time to get good at full OLL, because there's many algs, thus every single cases comes up less.

 

[scottishcuber]

isn't EG like 254 algs? counting reflections?

It's 126 algs. EG1 is restricted to an adjacent swap in the front (of the bottom face), so there's still only 42 algs for that angle...and of course CLL and EG2 are 42 each, sort of.

 

[thatboyahcubah]

I usually learn them in sets based on how they look. and I give them attributes in my head for when they look similiar. For example the knight cases in OLL. There's four, and the way I learned them was I grouped them into "good and bad". Good just being the easier alg. So I have right-good, left-good, right-bad and left-bad. Hope that wasn't too confusing.

Generally, I learn a set of algs, for example CLL Pi, until I can kinda do them. Then I just do solves and force myself to recognise the case. Over time, I'll get used to all the cases. So rather than learning one alg until it's second nature and easy muscle memory, I "half-learn" a bunch and they come with time. Not sure if I'd recommend that strategy though; It took me quiet some time to get good at full OLL, because there's many algs, thus every single cases comes up less.

that actually makes a lot of sense, thanks!

It's 126 algs. EG1 is restricted to an adjacent swap in the front (of the bottom face), so there's still only 42 algs for that angle...and of course CLL and EG2 are 42 each, sort of.

Oh, my bad. What do you mean by "sort of"?

 

[Rocky0701]

I reduce to triggers, then learn like half of the alg, then the next half, then i just do it over and over like TDM said and get it into my muscle memory. The most important part is to review though. If you do it 100 times, then don't do it for a couple days you won't remember it and then hard work is wasted, do it a couple times like every hour until you know it.

 

[scottishcuber]

Oh, my bad. What do you mean by "sort of"?

In the case of a LL skip you can have an adj or diag swap on bottom which can technically belong to either EG1 or 2, but they aren't necessarily considered as extra cases/algs.

 

[PJKCuber]

Learning which trigger belongs where. Then repeating the other stuff over and over again. Most algs are similar. eg T perm and J perm are similar.

 

[TDM]

The most important part is to review though. If you do it 100 times, then don't do it for a couple days you won't remember it and then hard work is wasted, do it a couple times like every hour until you know it.

Yeah I forgot that, that is definitely important. If you don't do this, you're likely to forget the alg. I sometimes learn algs in the evening and try to do them again in the morning, then maybe a few hours later too.

 

[Alan50]

Tips to learn OLL/PLL algorithms more quickly?

Should I just memorise them and write them down?

It's a lot of algorithms to learn. I have recently become able to do F2L in about 50 seconds (average). I have just started OLL/PLL half an hour ago.

I was thinking I should just draw out the OLL/PLL diagram and write down its algorithm next to it several times on a piece of paper.
I would do this until I have memorised every single OLL & PLL algorithm. I do this for exams and it has worked well.

After memorising them by writing & drawing diagrams, I would then practice cubing a lot to gain the muscle memory.

What other methods do you think will work? Do you like my method? I have a strong feeling that my method would work nicely.

 

[Hypocrism]

No, algorithm learning is just muscle memory. Writing it down is not useful. One alg at a time, perform it over and over, look at the case, etc etc. But you don't need them yet, your F2L is far too slow to need it yet, if you learn anything you should learn 2-look OLL and 2-look PLL which requires far fewer algorithms.

 

[slinky773]

I have never heard of anyone people successful with memorizing algorithms by memorizing the exact moves in notation. I doubt many people have memorized Sune by repeatedly going, "R U R' U R U2 R'! R U R' U R U2 R'!" That might be what works for you, but for most people, it doesn't work that way because 1. if you're going for speed, the only thing you want to do is to use muscle memory, not using your conscious memory (if that makes sense) and 2. it's going to take a long time to memorize 70-odd algorithms that way.

I would recommend learning 4-Look-Last-Layer first. I am assuming that you haven't learned this, given that I have seen your other posts saying that you haven't learned a full cube solve method. 4LLL has only about 10 or so algorithms, I believe. I would strongly urge you to only memorize the 4LLL algorithms through repeatedly seeing a case on the cube and repeatedly solving the case using the 4LLL algorithm, instead of memorizing the algorithms through paper and pencil memorization. This will help you to grasp the idea of how most people learn these algorithms: one at a time, putting them into muscle memory. Only them would I recommend that you learn full OLL/PLL.

Now, it may be that paper and pencil memorization works much better for you than only muscle memorization. If that's the case, well… I would recommend that you at least try it with 4LLL first so that you can see what works for you.

 

[kingi2001]

if you do 50 seconds on f2l, you should not learn full oll. mabye full pll, but not oll.
learn 4LLL instead.

- Yuval

 

[Alan50]

Thanks, I'll store it all in muscle memory. I'll take it slow and do 2-look, thanks for the advice.

 

[uberCuber]

No, algorithm learning is just muscle memory.

The ultimate goal is to have things in muscle memory, but algorithm learning is not just muscle memory if you want to learn a lot of algorithms efficiently.

 

[IamWEB]

OLL

I would recommend you get 1 algorithm into your muscle memory before moving on to the next one, so that you've mastered one case before complicating things with another one. Also, learn similar groups consecutively, because if helps you understand how an algorithm works if it shares triggers (see badmephisto.com's OLL page) with the previous alg you learned. That also helps you with learning more algorithms down the line. It gets easier as it goes on.
Badmephisto.com and the Speedsolving Wiki have their OLL lists categorized. Check out both pages.

BUT for now, 2-Look OLL and 2-Look PLL are how you should begin your Last Layer education. Focus on increasing your understanding of F2L first, because it's the most important step.

 

[Chree]

I agree with previous posts... for where you're at now, you shouldn't worry too much about learning full OLL. Instead you should be concentrating on F2L and 4LLL.

But you can basically start learning these things whenever you want. We're not the boss of you. So when you begin to learn a new algorithm, just do the alg over and over again. Eventually you'll be able to do the first few moves without looking at the notation... then the next few, then next few, then the whole thing. Then just keep doing it. The way I learned my N perms was by doing them over and over again while watching a movie (which is why, to this day, doing Nb still reminds me of the movie Looper).

Writing down the notation will not help you speed up your solves. Being more notation literate does help when learning new algorithms, but actually doing the moves over and over and over again until you aren't even thinking about the notation anymore will do you much more good.

Edit:

OLL

... because it helps you understand how an algorithm works if it shares triggers (see badmephisto.com's OLL page) with the previous alg you learned.

I was about to make a big long rant about how badmephisto is the best site in existence to learn OLL/PLL from... but you hit the nail on the head. Thinking of algorithms in terms of common triggers is hugely beneficial to learning new algs quickly. And it just gets easier and easier over time.

It might not seem obvious at first, but already knowing tons of algs makes it far easier to learn new ones. The hard part is getting that initial library in your head in the first place.

 

[ICanSolveA3By3]

i need help memerizing pll algorithms please help!

Lately i have been trying to learn full pll and i need help memerizing algs. If you have any suggestions please comment them to help me and thank you in advanced.

 

[Rocky0701]

Write each one out on note cards. Break each alg down into triggers and then learn the triggers. Memorize the alg in parts, and then when you are able to remember the entire alg, do it over and over. Do it a couple times before you go to sleep, and then a few times after you wake up if you still remember it. Once you have memorized it, don't focus on remembering it by the notation, but by your muscle memory of repetition. Look at how certain pieces move around during the alg, such as F2L pairs. Do one alg per day, or more if you are able. Incorporate it into your solving as soon as you can, so do some regular 3x3 practice as well. Good luck!

 

[ICanSolveA3By3]

Thanks man, I will try that

 

[nalralz]

Do just the PLL algorithm over and over for about 5 minutes straight looking at it. After that, try to do the algorithm without looking. Then wait about a day or so and try to do the algorithm again. If you can do it, try to time it and see how fast you can do it.

 

[qqwref]

http://i.imgur.com/AsNEM0n.png