Memory Methods

**dbeyer** · 04-12-2007

I use many memorization methods.
This is a small list:
I will provide a description of the methods, plus examples of how to use them.

Visual Memory Methods:

1. Some of the top notch cubist, such as Leyan Lo, use shapes to remember his blindfolded solves. Leyan also orients pieces. What the BLDist does is memorizes a cycle in the shape of a triangle connecting the pieces in the cycle.

2. This is the very basic method that I use, I think it's effective, yet not efficient;
I count as I solve piece by piece. I will alway count consecutively from 1-11.

Some signals I use: "Skip" and "O"
Skip tells me to start a new cycle. "O" (pronounced "oh") designates a piece that is permuted with a twisted orientation.

I do not memorize a string of numbers.
I have 11 variables for edges. (more counting Skips)

These variables are named
"One", "Two", "Three", "Four", "Five", "Six", "Seven", "Eight", "Nine", "Ten", "Eleven"

As I am counting I am tapping pieces, or at least scanning and remembering the piece that I permute as I count each number.

I remember which piece goes to each variable.
This is similar to Roman Rooms (Which I'll describe later)

Imagery Systems:
PAO (or some breakdown of that)
Person, Action, Objects.

I know that Chris Brownlee uses a PO system
Joel van Noort is working on a PAO system as well.

Each location as a predefined set of images.
With a PAO system (the biggest system), there would be about 72 images. You would then use about 20 of those images to memorize the cube. These 20 or so images would be broken down into 7 key points.

A person (Michael Jordan)
An action (slam dunks)
An object (basketball) This could mean something to somebody with a PAO system.

Let's say that UR->UF->UB gives this story.
However ...
UF->UB->UR is the same net effect of the cycle, just memorized from the UF, this would actually result in a completely different blurb.

A person (Chris Hardwick)
An action (solves)
An object (rubik's cube)

Each of the 24 locations have a specific person, action, and object.

You memorize in this order.

P->A->O->P->A->O->P->A->O

You will never get the Action image of the UF if it's a 3n (where n is a whole number) piece in the cycle.

Letter Pairs:
Chris Hardwick and I use letter pairs to memorize big cubes blindfolded. We have a lettering scheme for each piece type. Chris and I don't even use the same lettering scheme, personalize it so that the system makes sense to you, make sure it's something you won't forget.

We have images for every possible combination of 2 letters that can occur, with 24 locations, you get over 500 images! Do not fear the big 500, it is a very powerful system, I'm glad to have learned it and I will probably use it indefinitely for big cubes.

http://dbeyer.110mb.com/images.html

http://speedcubing.com/chris/memorizing.html

They are very similar mind you.

As memorize the order that pieces are solved we think of the letters associated with the locations, every letter pair is made into an image, then stored in a location, you then progress until all pieces of that type have been solved.

Storing Images:

Chris uses a Journey Method, I use Roman Rooms.
What Chris does is he goes to well known locations on a journey. He describes it on his site I believe, take a look at it.

I use Roman Rooms, the way that I use roman rooms, I place one image per location.

I create detailed rooms in my head that I normally am in.

Some of those rooms would include where I work, and the different unique rooms at my house and my grandparents, homes as well.

http://img258.imageshack.us/img258/8...manroomaz9.jpg

**dbeyer** · 04-12-2007

Quote Mike Carroll:

"I saw your post on the yahoo sports memory chat. Here's some insight on what I think.

I HIGHLY recommend you use a roman-room type journey method, as I think it has some advantages over the standard journey method.

If you've not heard of the roman room system, I'll describe it in full.

Basically, it's like a wrapped up journey method, but instead of using a long snake-like path from some point to point, you use points of a familiar physical room.

Almost all rooms are cubic; they have 4 walls, 4 corners, a ceiling, and a floor. To best illustrate a room, this image is given from a top birds eye view (looking from the top, down onto the floor.)

http://img258.imageshack.us/img258/8...manroomaz9.jpg

Now, the numbers designate the numbers of your journey method. This is more effective than the journey method, because it has more order and organisation.

I like this more than the journey method because it always follows the same pattern. Pretend you're standing in this roman room, imagine yourself positioned in the middle, looking at position 4.

Location 2 will ALWAYS be to your left, 6 will ALWAYS be to your right, 8 will ALWAYS be behind you, position 5 will ALWAYS be to your front-right, ect.

It's nice that all walls have even numbers, and corners have odds. It makes it quick and easy to navigate through. You don't "have" to start location 1 at the back-left corner, it can be anywhere. This is just what I use, and I don't see starting somewhere else better or worse in terms of efficiency.

As for position 9 and 10, well, originally that was the floor (9) and ceiling (10), but I've found that all ceilings are too much alike, and it's difficult to distinguish one wall from another (can screw up stories.) So instead, position 9/10 are respectively "under" and "above" the object located in that location. If in the center of the room was a piano, position 9 would be inside (and have the image interact with the location, echos maybe), and position 10 would be on top of the piano, or something similar like that.

Here's one of my rooms:

0 - My bedroom Object/Detail
1 Light switch
2 Game chest
3 Fuzzy white bear
4 Safe
5 Clock
6 Bed
7 Juggling ball
8 Pyramid magnetic
9 Clothes bucket
10 Inside clothes bucket

(hopefully that shows up). Now, what I like to do (and I find it effective), is first go around the room and note a specific detail about each location in preparation. I have a excel file that stores all my locations. When you're making points, it's good to have them distinct and very different. If there's nothing there, it's helpful to try to find something there, or to place some object there. Luckily, my room is messy enough for it easily to have 10 things. If there's a blank location, just take some object from that room and put it there.

Make sure you've got them memorised well, as you'll always use them no matter what you're memorising.

Then, when I am actually memorising and encoding information, this is what goes on in my brain. In location one, I would imagine my "mini-story" happening... INSIDE the lightswitch. Yes, inside. I pretend to shirk myself down and imagine a world of wires and electrical currents, and the scene happening there. It makes it much more visual, and links the story more easily to it. You can even pretend parts get "electrified".

Location 2, I think of just opening my game chest and seeing what's going on there (maybe the scene happens to be on the monopoly land.)

Location 3 I imagine a big fuzzy land (because it's a fuzzy bear, where the scene happens.) again, interaction generally occurs subtly.

It makes my locations stick with my images much more clearly, and it's easy to refer back to them.

It's also very easy to get locations in a house. You can organise your locations by a hierarchical system. So it would be like, Location-Room-House in terms of organizing.

You can use as many houses as you like, or rooms. They're reusable, and easily re-visitable. What's the BEST feature of the roman room system, is say, you want to find your 25th story.

Well, easy. 3rd room, 5th location. (*note, it is not the 2nd room, because the order goes like this)

1 - My bedroom
1
2
3
4
5
6
7
8
9
10

2 Patrick's Room
11
12
13
14
15
16
17
18
19
20
........
Make sense?

It's because I didn't start out with my room being room "0", I did one. You can do both, but I tend to like this better.

I currently have about 4 houses that I've been to and gotten in my mind. It's really not that hard. Each house has an average of 10 rooms, with 10 locations in each. It makes in nicer to use larger and more expensive houses, because then you'll get an even 100 locations per room. So if you have 10 houses, 1000 locations.

Say you need the 524 room. Well, that'd be VERY hard to find on a journey (and tedious), unless you had labeled each journey. Mine, easy. 6th house, 3rd room, 8 location. This would be the story that happened behind you, in the relevant object (when number 4 is in front of you when you're standing in the middle)

For rooms, you can use museums, shops, school, library, friend's houses, anything. I just find it very useful to make them as ordered as possible, as it makes recall so much easier."

End Quote:

**AvGalen** · 04-12-2007

I only just started learning blindfolded (2x2x2 and 3x3x3 only, sofar) using Macky's guide and adjusted numbering and algs to my style.

For orientation I LEARNED numbers, but I convert them to shapes automatically in my head. Untill I read this post, I didn't know this was called shapes. Also, I don't limit this to triangles
For corner-permutation, I try to group numbers so they are easier to remember. 1 3 2 4 becomes 13 24 becomes "unlucky 4!"
For edge-permutation, I use a story based on numbers.

If I use the first scramble of 2007-08 competition (B F R' D2 B F R' U' L' R' B2 F L R D2 U' R2 D U' R' F2 L2 D' U2 F), memory becomes this

(I scramble with white on top, green on front. I solve with yellow on top, red in front)

Edge orientation: UL, UR, RF, RB, DL, DF, DR and DB need to be flipped.
Remember as T-shape and entire bottom.

Corner orientation: UFR, UFB, DFL, DFR have to be turned counter clockwise (backwards), DRB has to be turned clockwise (forwards).
Remember as "Right-bottom back forwards" (so naturally the other piece has to go backwards). All other pieces of the front-right faces, except the first (UFL=1 for me) have to go backwards.

Corner permutation: 12345678 = ULF, URF, URB, ULB, DLF, DRF, DRB, DLB. 2->5->3->7->8->4->6.
Remember as 25 37 846. Easy to remember 5^2, 6^2+1, second highest even number possible

Edge permutation: 123456789101112 = UF, UR, RB, RL, LF, RF, RB, LB, DF, DR, DB, DL = 1->10->5->6->9->2->8->7->3->12(->1), 4->11(->4).
Remember as a Dutch scoring system. 1=lowest, 10=highest, 5=just not good enough, 6=just good enough. This is the story I made:
I got one 10 which was perfect (1->10)
But then I got mediocre and had to work my way up to perfection again (5->6->9)
I went from the lowest limits to the highest and back again (2->8->7->3)
Untill it ended (12)
Then all I had to do was switch the last remaining two (4->11)

Reading this back, I think I am very creative with numbers :)

**pjk** · 04-13-2007

Well, I will explain how I memorize right now. So I will do an example so you can see. This is the same scramble as Arnaud posted:
B F R' D2 B F R' U' L' R' B2 F L R D2 U' R2 D U' R' F2 L2 D' U2 F
Scramble with yellow on top, red front.

Okay, so I hold white on top, red on front now to memorize. First I remember CP. I do this strictly by numbers:
5 1 8 2 6 7
Since I use a very easy method of solving (as described on my site), I use the T-perm to cycle. In this case, that was a 6 corner cycle, and the CP is done for memo.

Next I remember EP. I do this all visually. So I first see it goes to the lower back, then top left, lower front, middle back left, middle front left, lower left, front upper, middle back right, middle front right, then I see the cycle is complete, but still have two un-cycled edges, so I see back upper and lower left and back up to back upper. It is hard to explain the technique, but it works fine for now.

Next, I memorize EO using mostly numbers, unless it is something I think I can visualize. In this case, I can definitely see it visually. I see the upper back needs to be flipped, and the 3 edges towards the back of the down face. So I know in my mind I will do an x2 rotation, then D2 B2, do the 4 edge flip, and undo it.

Finally, I visually memorize CO, very quickly. I see 1 and 2 corners need rotating. I see with a simple commutator, I will rotate 1 to the correct spot, and 2 will be 1 "alg" away from correct. I see in 5 6 (the two corners on the lower front face) need to do an "anti" commutator alg (that is what I call that design). I see that corner 8 is 1 alg from being flipped. So I will remember to do the commutator alg once for 1 and 2, then do F2 and fix 5 6, and finally do a L2 to setup and fix 8 2 together, and finally undo it.

Okay, so this method is kind of odd. I am working on an object system to modify Pochmann. I will assign an object to 22 locations, and basically remember a sentence for EP. My current CP and CO is working well, and for EO I will work on doing it all visually. Once I get that all good, I should be able to memorize sub-60 seconds pretty easily. Then I will work towards speeding the solve phase.

Have fun,
Pat

**AvGalen** · 04-13-2007

PJK, I tried to follow your explanation, but I need some more info:
Do you also solve with yellow on top, red front? Or white on top, red on front?
What is your numbering scheme for CP?
What is the first piece in your EP (I am guessing UR)
Is their any way you could make a video of this and upload it? (I know you are on dial-up)

**Erik** · 04-14-2007

I don't use any system...
Maybe you could call it slightly visual..

**pjk** · 04-14-2007

PJK, I tried to follow your explanation, but I need some more info:
Do you also solve with yellow on top, red front? Or white on top, red on front?
What is your numbering scheme for CP?
What is the first piece in your EP (I am guessing UR)
Is their any way you could make a video of this and upload it? (I know you are on dial-up)

I solve with white on top, red on front. For CP, I start with UFL as 1, UFR as 2, etc. The first piece for EP is UR. As far as a video, I really can't. The only video camera I have is a real nice digital photography camera, and it makes videos huge. A 20 seconds video will be 10 MB or so. However, I will try to make one, compress it, and upload it at the college. My method is very similar to McGaughs, as of right now at least.

Check out my site in my sig and click on BLD cubing for more about the method.

Pat

**AvGalen** · 04-14-2007

Erik: If you have no system, do you really think you can >1 cube at once at the German Open?

PJK: Thanks for explaining. I will try to understand it tomorrow. Today was speedcubing day, tomorrow is blindfolded day :)