Memory Methods

[Ordos_Koala]

it's much easier to remember images (mostly you remember the funny ones) so it's better to these letters assign words... you can imagine duck and carrot but you can't imagine "BR"... and few people could just like that remember 40 letters (i don't say it's impossible, but surely it's very, very, very hard)

 

[MichaelErskine]

Yes, I find that images stick really well. One well-rehearsed image per colour pair -- except for blue-orange which I recall as "Mystery Object"!!!

 

[Cubenovice]

When I was still using a single image per sticker the Blue-orange and orange blue were my worst images...
I never got comfortable with them and would mix them up too often.

I am now using letter-pair images. My list is far from complete so I throw in some freestyle words and old images to complete my memo.

My corners:

 

[riffz]

except for blue-orange

sad tabby cat

 

[Cubenovice]

I would like to know how people typically memo new cycles, twisted-in-place corners and flipped edges, especially for multi BLD.

What I am currently doing:
- Old Pochmann for both edges and corners
- All stickers have their own letters
- I do a story-type memo with letter-pair images (words) and words made up on the fly (as my list is not complete yet).

- For new cycles I memo:
In case of an odd cycle I will include in my story: " 'name' places buffer in next position"
This tells me the last sticker of the cycle and that I need to move the buffer piece.
In case of an even cycle I memo actually the same but will skip the specific person as the last sticker is included in the last letter pair.
I am actually thinking of using one specific person for this case so I can better work it into the story.
I do not really have a code for the buffer's next position, this is done on plain old memory.
The above works just fine for single solves, I do not have too much trouble in recalling where the buffer needs to be stored for the various cycles.

For twisted corners and flipped edges I use visual memo.
Twisted corners I solve via two Y-perms
Flipped edges with a simple edge flipper or two T-perms (depending on set up moves required)
For a single solve this typically works just fine.

Now if I want to progress to multi BLD should I start "encoding" all the cycles, twist and flips into letter pairs?
For a new cycle
NOT memo" 'Kevin' places buffer in next position" (suppose the next location is the edge just straight across: UL which has letter E)
BUT just plain and simple memo the letter pair KE
This way I can just "forget about cycles" and just follow the story and keep shooting to the locations in the story

For twisters and flippers:
NOT visual
BUT just remember a letter pair for each corner and/OR edge.
Drawback for edges would be that this means "T-perm by default" so 4 T-perms instead of one edge-flipper to solve two flipped edges.
(Unless I would "decode" the letter pairs again, for instance when there are not too many flips and there is an easy set up).

Any feedback is appreciated!

 

[cmhardw]

I would like to know how people typically memo new cycles, twisted-in-place corners and flipped edges, especially for multi BLD.

I would also be interested in hearing what others have to say. To be honest I don't really have a system for flipped pieces other than visual. I've been wondering how to incorporate this into a set system using letter pairs somehow for multi as well. My worry right now is that I may confuse the flipped pieces with an actual cycle. I suppose this could be gotten around by always breaking new cycles (so that you never have more than one cycle) and then just remembering the flipped/twisted pieces as a separate chunk from the actual cycle.

- For new cycles I memo:
In case of an odd cycle I will include in my story: " 'name' places buffer in next position"
This tells me the last sticker of the cycle and that I need to move the buffer piece.
In case of an even cycle I memo actually the same but will skip the specific person as the last sticker is included in the last letter pair.
I am actually thinking of using one specific person for this case so I can better work it into the story.
I do not really have a code for the buffer's next position, this is done on plain old memory.
The above works just fine for single solves, I do not have too much trouble in recalling where the buffer needs to be stored for the various cycles.

I feel this is the only part I can comment on. I just always break new cycles until I'm done basically. The reason I like doing this on 3x3x3 is that it makes sure to balance the orientation of all pieces by the end (except for permuted but disoriented pieces).

Basically let's say I had the corner cycle:
(ABCD)(EFGH)

I would memo (A BC DE FG HE)
Where bolded letters show where I broke the new cycle, and bolded and underlined showed where I close out the cycle.

The key to making this work is to remember where you have broken into a cycle, and once that cycle is complete, you must shoot to that position one last time to close the cycle out.

So if you had the cycle (AB)(CD)(EF)(GH)
I would memo: (A BC DC EF EG HG)
Where bolded letters show where I broke the new cycle, and bolded and underlined showed where I close out the cycle.

As to permuted but disoriented pieces, I would say that once you condense your memo such that you always break into new cycles such as to only have one longer cycle memorized, then you could memorize the permuted but disoriented pieces the same way you would memorize a 2nd cycle. But, since you know that the 2nd cycle does not exist, then those pieces must be permuted but disoriented.

I have no idea if this is how others do things, but that is probably how I would do it. Hope this helps.

 

[Mike Hughey]

For multiBLD, for new cycles, I break them until I'm done just like Chris describes. Actually, about a year ago I switched for multi to including my entire set of edges in a single location, so I might in a bad case have 7 or 8 images in that single location. I find that despite the greater difficulty in originally memorizing that many images in one location, recall for it is faster that way, which is why I do that. I also include my entire set of corners in a single location. Then, I remember images upside down for pairs of edges that are flipped. If I have a single edge that's flipped, I'll remember my image for the person that goes with that single letter (again, upside down - usually suspended from the ceiling at that location), and then know when I get to it that I need to flip it with the buffer piece. For permutations, I always have a single person who represents my buffer piece (which never changes for a given piece type in a given room) acting on all of the images. So my flipped edges or twisted corners don't get confused because: a. they're upside down, and b. they aren't being acted on by the person. For twisted corners, if they're in pairs (one clockwise and one counterclockwise), I do the same as with flipped edges - I memorize the image for the pair upside down, with the first one needing to go clockwise and the second one needing to go counterclockwise. If I have triplets that need to all go one direction or another, I come up with a word on the spot combining the 3 letters, and place that concocted image on a clock (if they need to go clockwise) or on a kitchen counter (if they need to go counterclockwise). Maybe it's a little silly, but it works for me. Anyway, I put all flipped edges at one location together and all twisted corners at one location together, so I never need more than 4 locations for a 3x3x3. Since all of my rooms have 9 locations, that means I can always fit two 3x3x3s comfortably in a single room.

 

[Sakarie]

But, isn't it so simple so that you have a special image of when FR, UR or DF are flipped? For me, FR is A, and RF is B, so if FR would be flipped, I'd memo AB=Abba. So when Abba is standing in my kitchen, I know FR is flipped.

 

[RyanReese09]

That's really an interesting approach Mike, here is how I approach multi (note that I'm not nearly as good as you (yet ))

I just use standard roman rooms with 10 locations in each room, 2 images per location in my room, letter pairing gives me 4 letters per location. So basically 3 locations max for edges, 2 for corners, thus 2 cubes per room.

For flipped edges I do the same as you, I just have the image upside down, depending on whether the first letter is flipped in my image, the top half of the image will be upside down (aka a person will be flipped, top half only, if you can imagine that) and if it's the second letter that's flipped, it'll be the bottom half.

If two edges are flipped, and permuted, I just (since I solve edges first) put on the blindfold and do a pure two flip. Haven't really decided what to do if only one is permuted but flipped, probably new cycle (if I'm doing multi)

If regular 3x3x3 BLD then I'll just visually remember that it needs to be flipped

Corners, I know pure twists but I never apply them to corners, due to me being afraid if I'm twisting the incorrect way . Should really start doing the pure twists, it would save me a lot of memo for those situations.

 

[Cubenovice]

Chris, Mike,
thank you for your detailled replies.

Are you sure we are BLDig the same puzzle?
As I am usig Old Pochman my cycles seem to work quite different from yours;-), my cycles "close themselves"

I have now tried a few solves in which I coded breaking into new cycles, flips and twists into regular letter pair images and it works like a charm.
At first it feels a bit weird that you encounter a given letter twice but that is something I can get used to.
This system also makes it easier to check for parity. Previously I had letter pairs, single letters and breaking into new cycles to count and see if I ended up odd or even.
But now everything is in pairs and so I only have to look at the very last letter: paired or single.

 

[Krible]

Hey! I'm kind of new to blindfolding and i havent solved a cube blindfolded successfully yet, but i think i have come up with a new method for memorizing corners and would like to see what you think.

The first step is to assign each sticker of every corner to a physical movement in your body. So for instance, LUF can be assigned with putting your toungue to the left, FUL can be putting your tounge to the right, ULF can be biting your tounge, FRD can be crossring your toes on the right foot ect ect..

Then, when memorizing, let's say the first corner is FRD, then you can just cross your toes or do whatever that corner is assigned to, and remember the rest of the corners visually. When executing you notice that your toes are crossed so you know where the first corner belongs. This way, you can remember one less corner visually, so i think it will be faster.

It would be difficult to remember all the corners with this method because you don't know what order to do them. That's why i think it would be useful to only do it with the first corner and do the rest visually.

I'm not that good at explaining, but if you understand what I mean, what do you think?

 

[LarsN]

It is possible to use your body to assist your memo. I cross my legs to remember parity.

 

[Zane_C]

Hey! I'm kind of new to blindfolding and i havent solved a cube blindfolded successfully yet, but i think i have come up with a new method for memorizing corners and would like to see what you think.

The first step is to assign each sticker of every corner to a physical movement in your body. So for instance, LUF can be assigned with putting your toungue to the left, FUL can be putting your tounge to the right, ULF can be biting your tounge, FRD can be crossring your toes on the right foot ect ect..

Then, when memorizing, let's say the first corner is FRD, then you can just cross your toes or do whatever that corner is assigned to, and remember the rest of the corners visually. When executing you notice that your toes are crossed so you know where the first corner belongs. This way, you can remember one less corner visually, so i think it will be faster.

It would be difficult to remember all the corners with this method because you don't know what order to do them. That's why i think it would be useful to only do it with the first corner and do the rest visually.

I'm not that good at explaining, but if you understand what I mean, what do you think?

One corner isn't all that much different, but it does sound interesting. Definitely give it a shot and see how is works for you.

 

[Cubenovice]

Then, when memorizing, let's say the first corner is FRD, then you can just cross your toes or do whatever that corner is assigned to, and remember the rest of the corners visually. When executing you notice that your toes are crossed so you know where the first corner belongs. This way, you can remember one less corner visually, so i think it will be faster.

It would be difficult to remember all the corners with this method because you don't know what order to do them. That's why i think it would be useful to only do it with the first corner and do the rest visually.

I'm not that good at explaining, but if you understand what I mean, what do you think?

Uhm... especially the first (few) corner(s) are easy to remember visually so why start with something complicated?
If you really want to use something like this use it for parity or perhaps remembering the locations of flipped edges.

 

[Krible]

Well, you are right one corner isn't that much. But i find that when i have remembered 7 corners and try to remember the last one, then i instantly forget everything. So even if it's just one corner it might help alot, at least for me.


EDIT: I have been thinking a bit about how this method can be used for remembering more than one corner. Maybe this is complicated, but im just brainstorming: To remember two corners you can do more tension in your muscle for the first corner and less for the second. So if the first corner is FRD, then cross your toes very tight. If your second corner is LUF, then put your tounge just slightly to the left. This way you know what order they are because the muscle with more tension is the first one, and the one with less tension is the second one.

 

[Zane_C]

I seem to have missed the little discussion on flipped and twisted pieces in multi. My system probably isn't too good, sorry for the bad explaining:

I have a separate route, away from the cycles. The locations are numbered, i.e. the first location is where I would find any flipped or twisted pieces on the first cube.

Edges:
If I have 1 edge that needs to be flipped (excluding the buffer), I use the 2 letters on that edge to create an image.

Example: If UR is flipped I'll use the letters for UR "C" and RU "F", hence the image is coffin.

If I have 2 edges which need to be flipped, I'll randomly pick 2 of the 4 letters and make an image.
If the number of flipped edges exceed 2 (Excluding the buffer), more than 1 image is obviously needed. For the flipped edges, the images relate to the sky in some way, i.e. floating, falling, hanging, stuck to the ceiling.

Corners:
For twisted corners, I memorise the letter of the sticker which needs to face up or down.

Example: If the corner UFR needs to be twisted clockwise, the upper sticker is at FRU. For me that letter is "I", if there isn't a pair of twisted corners I will use one of the 2 remaining letters on that corner to make an image. The letter of the sticker which needs to face up or down is always in front.
The images for twisted corners are placed on the ground.

 

[Sakarie]

I came up with an idea the other day, and I'm just wondering if anyone uses it, and if it's any good.

It's basically an image method, were you have images for preferably every pair of letters, and the first letters/image DQ/Duck, suddenly transforms into a table/TB, which suddenly gets crushed by Curios George, and so on. (Everything is happening very "sudden", it's meant to be memorized fast, and the "story" is also happening very fast.) When I've done a few tries, I've ALSO put the whole image-chain into a special place, to avoid confusing one chain to the previous one. Anyone uses it?

 

[riffz]

I came up with an idea the other day, and I'm just wondering if anyone uses it, and if it's any good.

It's basically an image method, were you have images for preferably every pair of letters, and the first letters/image DQ/Duck, suddenly transforms into a table/TB, which suddenly gets crushed by Curios George, and so on. (Everything is happening very "sudden", it's meant to be memorized fast, and the "story" is also happening very fast.) When I've done a few tries, I've ALSO put the whole image-chain into a special place, to avoid confusing one chain to the previous one. Anyone uses it?

Isn't that just images/journey?

 

[amostay2004]

It is possible to use your body to assist your memo. I cross my legs to remember parity.

I thought about using your legs to remember flipped edges the other day. Not just whether or not there are flipped edges, but also the position. For example:

Floor
----------UL------------------UR


BL---------------FL--FR---------------BR


----------DL------------------DR


And possible have some area in the middle for M slice edges, you get my point. It's effective when having 1 edge flipped, but might be hard with 2 edges, and can't be used for more than 2 edges, but the majority of scrambles have 2 or less flipped edges anyway

Does anyone use something like this? I personally have not tried it though

 

[cmhardw]

@Using your body to remember things: I used to use a base 5 counting system with my feet to count how many wings I had cycled to