Memory Methods

[Mike Hughey]

I have one rather odd thing I use for memory that I've never heard of anyone else using, and while it's probably not useful to anyone who isn't already very familiar with hexadecimal, I thought I'd mention it in case it's useful to somebody else out there.

I use 3-cycle (learned from Macky's website). I'm able to memorize and execute orientations very quickly (for me, anyway) by using hexadecimal to memorize them. I use one bit for each position for edges, which requires 3 hexadecimal digits - one digit for top face, one digit for middle layer, and one digit for bottom face. Then I use 4 hexadecimal digits for corner orientation - 2 digits (one for top face, one for bottom face) for those that need to go counterclockwise, and 2 for those that need to go clockwise. So it's just 7 hexadecimal digits for the entire orientation, and I find hexadecimal easier to memorize than decimal because of the letters. Many times the patterns make a word or some other thing that's easy to remember. And it's nice for the corner orientations if you're good with hexadecimal, since you can just count the bits to see if what you've memorized is really legal or not - they either have to be the same or off by a multiple of 3 bits. After a while of using this, you quickly get to where you can just glance at the edges or corners and instantly recognize them as a particular hexadecimal digit, and that's where it gets really fast.

 

[Karthik]

Thats a very cool method!!
Since I memorize CO last and remember it visually I dont even need 4 out of those 7 hex digits.So for orientations all I have to do is remember 3 hexadecimal digits!!
I guess people with a little mathematical inclination will find this awesome!

 

[HelloiamChow]

I use numbers for EP, chinese numbers and tapping the pieces for CP, tapping pieces for EO, and tapping pieces for CO. Yeah, I rely a lot on tapping pieces. I'm working on some pictures for me to remember too, to back up all my numbers.

 

[pjk]

I have switched to a new way for me to memo. I first memo EP just visually. Then I memo CP by repeating the first 4-5 digits of the cycle, and then the rest I memo visually. I memo EO, and finally CO all visually. Memo'ing EO and CO takes me like 5-10 seconds max normally. I memo the entire cube around 90 seconds on avg.

 

[aznblur]

How would you memo 2-cycle where EP and EO are done at the same time (Stefan Pochmann's 2-cycle)?

 

[joey]

You can use the same methods that people are using. You can do visual, or you can use a letter/number scheme. Except each piece has two numbers/letters, one for each orientation.

I would say that a non visual memory for pochmann would lend itself to bigger cubes aswell.

 

[pjk]

How would you memo 2-cycle where EP and EO are done at the same time (Stefan Pochmann's 2-cycle)?

I would use images. I was doing this before I moved to 3 cycle. Assign an image to all 22 shoot locations, and memorize a story with those images to create the cycle. I can be done extremely fast with practice.

 

[tim]

After a few experiments i want to describe my problems/experience with memory methods:

I switched from 3-cycle to M2, because my memorization times were really bad, so i had to learn some kind of memory method. I think a memory system based on images needs a lot of practice for people using 3-cycle to be as fast as memorizing pure visually, because you have to encode 3 images (or whatever you've memorized). So i decided to give M2 a try, where you need only one image at a time.
At first i used a pure journey system. I placed one image at one location. After a few tries, i noticed that it took a long time to link my image and my location together. I also had the problem, that M2 requires you to keep track of the m slice's position. So i tried to use 2 images per location. That was even harder to memorize, because i had to invent a mini story, to memorize the order of the two images. It was all very frustrating and i didn't want to practice that much, to become fast at memorizing two images at one location.
So i gave a 'story system' a try. I link my images together by making a funny story. And it was fast from the beginning (i needed less than 2 minutes for memorizing and executing edges the first time. That's almost as fast as my times with 3-cycle + visually memorization). So i think that's the system to go, but i have the m slice problem again. I have to know if the m slice is "correct" or not. At the moment i try to count (not really count, but saying "true", "false", "true", ... after each piece), but that distracts me too much and i can't look ahead well. Do you guys have any idea how to keep track the status of the m slice? I also thought of counting while memorizing in order to memorize the pieces at the m slice (UF, DB) how i'll have to execute them and not what their "real" target is. So i wouldn't have to think while executing. What do you think?

And thanks Daniel for your advices and sharing your experience .

 

[hait2]

After a few experiments i want to describe my problems/experience with memory methods:

I switched from 3-cycle to M2, because my memorization times were really bad, so i had to learn some kind of memory method. I think a memory system based on images needs a lot of practice for people using 3-cycle to be as fast as memorizing pure visually, because you have to encode 3 images (or whatever you've memorized). So i decided to give M2 a try, where you need only one image at a time.
At first i used a pure journey system. I placed one image at one location. After a few tries, i noticed that it took a long time to link my image and my location together. I also had the problem, that M2 requires you to keep track of the m slice's position. So i tried to use 2 images per location. That was even harder to memorize, because i had to invent a mini story, to memorize the order of the two images. It was all very frustrating and i didn't want to practice that much, to become fast at memorizing two images at one location.
So i gave a 'story system' a try. I link my images together by making a funny story. And it was fast from the beginning (i needed less than 2 minutes for memorizing and executing edges the first time. That's almost as fast as my times with 3-cycle + visually memorization). So i think that's the system to go, but i have the m slice problem again. I have to know if the m slice is "correct" or not. At the moment i try to count (not really count, but saying "true", "false", "true", ... after each piece), but that distracts me too much and i can't look ahead well. Do you guys have any idea how to keep track the status of the m slice? I also thought of counting while memorizing in order to memorize the pieces at the m slice (UF, DB) how i'll have to execute them and not what their "real" target is. So i wouldn't have to think while executing. What do you think?

And thanks Daniel for your advices and sharing your experience .

i don't know much about the m2 method, but this seems like an elegant solution that i would use
when memorizing your images, you could switch between day/night as a background, corresponding to the location of the M slice.
it would be easy i guess if you link your images to the state of time ( i don't know what images you use so i can't really help with specifics)

just an idea =)

 

[Jack]

I have been trying M2 for with 3 cycles for corners recently, and parity was easy to recognise because I would solve the corners first and know whether there was one by seeing if the corners had parity first.

 

[tim]

Thanks . Manipulating my images with something like day-night is an interesting idea, but day, light won't work. Perhaps setting parts of my story on fire? I'll give it a try

I don't have fixed images. It's more like the word groups Matyas mentioned.

For example: UF (nature under water) -> FR (kitchen) -> BL (Musik)

So my story would look like:
A fish cooks a microphone.
or
I'm swimming under water and seeing a kitchen from a sunken ship in which a dj uses a few pots to make music.

I think you see, why day-night won't work . But thanks for your idea

 

[AvGalen]

Can't you just lift/drop your feet to keep track of the M-slice while solving? Is it really something you need to memorize?

 

[tim]

Can't you just lift/drop your feet to keep track of the M-slice while solving? Is it really something you need to memorize?

Also a good idea, but difficult to visualize (you always have to watch at your feet ).
I don't have to memorize it, but for two targets (BD, UF) i need to know it. Perhaps i'll colorize the part of my story if the m slice is correct, otherwise my image is just gray.
The reason why i want to memorize it, is: I want to make the execution as thinking free as possible to avoid errors. It should just be a mechanical procedure. An image pops into my mind => i execute an algorithm and nothing more.

 

[AvGalen]

If you don't need to memorize it, I would seriously give the lift/drop foot (easier than both feet) idea a go.

If you just lift your toes as far as possible, you will feel a certain strain on your feet. I am sure that after a couple of solves you won't have to "think" about the position of the M-slice. You will just feel the strain (wrong) or not and act accordingly without thinking.

If you don't have to memorize it, don't memorize it!

 

[tim]

If you don't need to memorize it, I would seriously give the lift/drop foot (easier than both feet) idea a go.

If you just lift your toes as far as possible, you will feel a certain strain on your feet. I am sure that after a couple of solves you won't have to "think" about the position of the M-slice. You will just feel the strain (wrong) or not and act accordingly without thinking.

If you don't have to memorize it, don't memorize it!

Ah, i see . I thought you meant memorizing the state of my foot while creating the story. I didn't thought of this "real life" stuff . I'll definitely give it a try. This will look funny at a competion .

 

[Mike Hughey]

The lift / drop your feet thing is very clever, Arnaud! I hadn't thought of doing something like that. You could use one foot for parity, and the other for the M2 slice.

One wonders if you could carry that farther by doing various things to "memorize" - touching your tongue to a certain tooth to remember a certain combination, for instance. I wonder if such will ever be banned as "cheating" (although it would be hard to enforce!).

 

[dbeyer]

In my humble opinion, you can not be deemed as a cheat unless you glance, peek, feel, or otherwise make a physical reference to the cube with the assistance of a partner, or your own senses. Yet anything that you can do to make this blindfolded cubing event easier for yourself couldn't be banned.

Seriously though, I wouldn't want to stand upside down on my head just to remember that I had a parity. How you position yourself whilst solving could actually hinder your mobility.

There are much better systems, such as Letter Pairs, Person Object systems, Journey/Roman Room, peg system.

A trick that I use, is to recall in pairs when using a system like M2 or Commutators, even if the recall is visual. If you recall in pairs, everything on the cube is balanced.

 

[AvGalen]

I actually use my feet for something else sometimes. If I have a couple of small cycles (3+4+4 edges for example) I would put my left foot in a 30 degrees angle and my right foot in a 40 degrees angle.

And I don't see why this would be banned. People use their fingers to count and/or tap pieces. What if someone decided to use one hand / 5 fingers to remember 5 pieces and solved with only 1 hand?

Blindfolded doesn't mean you can't use the rest of your body!

 

[LaffyTaffyKidd]

I'm confused on how would you memorize the cube with hexadecimals... I know what they are, but just don't get it. Could someone explain to me please

 

[hait2]

I'm confused on how would you memorize the cube with hexadecimals... I know what they are, but just don't get it. Could someone explain to me please

i used to do this for edge orientation and thought it was clever ( i don't anymore, it takes a little too long for me, but that's personal preference. my current system for EO is just remember a general gist of what it looks like. so like 1, 2, 3, 4, for singly incorrect edges, horizontal/vertical/topleft/topright/bottomleft/bottom right for two incorrect edges, A D C V for 3 incorrect edges based on what shape they look like or "all" for .. all. it's still 3 things to memorize but no nasty conversions to do in your head. )
what you do is, each edge is either correct or incorrect in terms of orientation. 1 or 0. assign whatever to a state. I used 0 for oriented (so i can ignore it) and 1 for unoriented.

so for example, you're looking at edges 1234 and 2 and 3 are unoriented. you would remember 0110 because 1 and 4 are oriented (so they're 0's); 2 and 3 are not (so they're 1's).

now this 0110 number is a binary number =) it's a 6. you'd just remember a 6 and since there's only 1 way to write 6 in binary (110) then you know orientation of the first 4 edges with 1 number

if you have like 1001, that's a 9, 1111 is an f (15 in hex), etc.