# Random Blindfold Cubing Discussion

#### Mike Hughey

Staff member
Then again, it can be bad even without orientation. Try this scramble:
D R2 D F2 L2 D F2 L2 U L2 U' B2 U' R F' R F L' R D2

#### riffz

##### Member
Chris, how do you use your image system for 2-3 twisted corners? I might adopt images for twisted corners.

Currently I'm memorizing and solving in the same order as you, but for twisted corners I use my feet to remember as much as I can.

If there's 1 corner twisted, I take the foot that is on the same side of the cube (L/R face) and angle it right for CW and left for CCW.
If there's 2 twisted, then if one is on R and the other on L, I can use the technique described above. Otherwise, I just kind of remember them visually.

I often stop the timer in a rush and forget to fix the twisted corners. I could definitely fix this bad habit over time, but as you can see, my foot system really only works for 1-2 images and is still shaky even for 2. If I adopted images for twisted corners I might extend my corners to 2 locations along my journey, as I currently just cram them all into one room every time.

#### RyanReese09

Is it bad I just keep placing 2 images per location in my room (if I have to cross between edges/corners mid location I put an invisible barrier in there, it works surprisingly) for multi?

#### Mike Hughey

Staff member
Is it bad I just keep placing 2 images per location in my room (if I have to cross between edges/corners mid location I put an invisible barrier in there, it works surprisingly) for multi?

No; whatever works best for you is fine. There is a tremendous amount of variety as to how it is done among the top people at multi.

However, it might make sense for you to try some other approaches, and see what works best for you. As I said, I found that putting lots of images at a single location worked really well for me, but there are other people for whom it doesn't work at all.

#### cmhardw

Chris, how do you use your image system for 2-3 twisted corners? I might adopt images for twisted corners.

I tried out Mike's idea of having my regular images just upside down. I found this was difficult for me as most of my journey locations are outdoor spaces, so I had a hard time picturing them upside down as it felt "weird" for them to be suspended in thin air. It sort of gave me the impression that the image was flying, and this would lead me to try to recall from my flying images when I got to the flipped piece.

My system now works very well for me, but it might seem a bit weird at first. Using this approach I can memorize 3/7 of all cases with 2 twisted corners using only a single letter image, 4/7 of all 2 twisted corner cases using one letter pair image, 3/8 of all three corner twisted cases with both one letter pair image and one single letter image, and 5/8 of all three corner twisted cases with two letter pair images.

For 2 flipped corners I do the following:
I memorize one edge sticker for an adjacent pair of corners that need to be flipped. I use double sunes to flipped pairs of corners, and pretty much have ever since I began BLD. The names I give the cases are from 2003 when I first started doing BLD. I give each pair a name of either "Same" or "Opposite" and I use the letters "S" and "O" to memorize during a solve.

The following are the algs I use to solve each case:
S = R' U2 R U R' U R L U2 L' U' L U' L'
O = L' U' L U' L' U2 L R U R' U R U2 R'

The only thing I changed is how I call these cases. Rather than calling the case where UFR twists counter-clockwise which UBR twists clockwise "S" I now call it "I", which is my sticker letter for the RU edge location. If I had an "O" case (UFR twists clockwise and UBR twists counter-clockwise) I would call it "C" which is my sticker letter for the UR edge location.

So now I can identify an S or O pair anywhere on the cube. The letter "K" would be an S pair on the RF edge of the cube (UFR twists clockwise and DFR twists counter-clockwise).

This means that I can memorize any adjacent pair of corners that twist opposite directions using only one letter. This accounts for 3/7 of all possible cases with 2 corners flipped.

-------

For these cases I memorize as follows. Let's say that UFL needs to twist counter-clockwise and DFR needs to twist clockwise. If I do the setup turn R, I get an S pair at the FU edge location. To memorize this case I pick any sticker on the DFR corner and I take the letter "F", which is the letter for an S pair at FU and I make a letter pair image out of them. I pick the letter that has the most memorable image when paired with F (and I make sure to avoid double pairs like FF, etc.) I would probably pick XF which is the Phantom of the Opera.

Now when I see XF during memo, I notice that I have only a letter pair image. Having only a letter pair image always means that the first letter is a corner sticker, and the second letter is the edge sticker that I need to move that particular corner sticker adjacent to, and this creates either the O or S pair at that edge sticker location.

So, for example, XF would tell me to take the DFR corner and move it adjacent to the FU location (the setup turn R does this) and then do either the S or O pair that the edge sticker represents. Afterward I would know to undo the setup turn.

This handles the 4/7 of the two corners twisted orientation cases where the two twisted corners are not adjacent to each other.

3 corners, including the buffer

For these cases I memorize one sticker from each of the non-buffer corners, as well as the twist direction. This gives me a letter pair image, followed by a single letter image. The only possible way this could happen is if 3 corners twist the same direction, one of which is the buffer corner.

So let's say that DFR, UFR, and the buffer after solving (UBL) have to all twist clockwise. I would take a sticker from each of the first two corners, X, and D respectively, and I would add a letter for the twist direction. A=clockwise B=counterclockwise. So my image here would be XD A or the Barad Dur tower from Lord of the Rings with the eye of Sauron (in miniature) being heckled by Dan Aykroyd.

When I see this during my memo I notice a letter pair image, and a single pair image. This tells me that the first two letters are stickers on permuted but disoriented corners. The last sticker is always the twist direction, so since only 2 corner stickers are represented then I know to include the buffer as well. The last letter A tells me that all 3 pieces twist clockwise to solve.

Twist three corners not including the buffer

Same idea as above, only I have 3 stickers (one from each corner) and the twist direction letter at the end. This gives me two letter pair images. The only possible way I could have two letter pair images is to have three twisted corners (not including the buffer, since all 3 corners are represented) and the last letter tells me the twist direction.

I can make sure to avoid letter doubles (AA, BB) by intelligent sticker choices during memo.

Currently I'm memorizing and solving in the same order as you, but for twisted corners I use my feet to remember as much as I can.

If there's 1 corner twisted, I take the foot that is on the same side of the cube (L/R face) and angle it right for CW and left for CCW.
If there's 2 twisted, then if one is on R and the other on L, I can use the technique described above. Otherwise, I just kind of remember them visually.

I often stop the timer in a rush and forget to fix the twisted corners. I could definitely fix this bad habit over time, but as you can see, my foot system really only works for 1-2 images and is still shaky even for 2. If I adopted images for twisted corners I might extend my corners to 2 locations along my journey, as I currently just cram them all into one room every time.

I may try to add in a foot element as a double check, but so far my image system is surprisingly easy to use, for me, and it's a relief to not have to try to hold onto the visual corner twist memo throughout the edge auditory loop memo. Auditory loop single syllable word memo is surprisingly taxing on the brain for me, so I think this is why I forget the corner twists so often. I will probably also now use two locations to memo corners on all solves.

Last edited:

#### RyanReese09

Mike, how would you know which image goes first in execution if all are on the same location? I usually determine it by the first executed is the dominant object/person. Aka the first one executed is hte one doing the killing etc...

#### Mike Hughey

Staff member
Mike, how would you know which image goes first in execution if all are on the same location? I usually determine it by the first executed is the dominant object/person. Aka the first one executed is hte one doing the killing etc...

I don't know - I guess it's just a small story for me at each location, so there's an implied order. I actually have a person representing the buffer piece that does the action on the objects, so that doesn't help. For what it's worth, I do sometimes make the mistake of swapping images, but that's far from being my most common problem - it really doesn't happen all that often.

#### cmhardw

For anyone who was following the discussion on how to memorize permuted but disoriented corners I have an update.

Memo'ing pbd corners with images makes me the opposite of a sad panda. This is totally worth making my general method for the future :tu

Then again, it can be bad even without orientation. Try this scramble:
D R2 D F2 L2 D F2 L2 U L2 U' B2 U' R F' R F L' R D2

EEeeeeeeEEEEeeeeewwwwWwww!!!!!!! *blegh*

1:52.32

That solve is horrendous. At a certain point during edge memo (after corner memo was done) I thought to myself "This is completely ridiculous" and memo'd the rest visually.

--edit--
By the way, the probability of a scramble like that is:

$$\frac{2\left( \begin{array}{ccc} 12 \\ 2 \\ \end{array} \right)\left( \begin{array}{ccc} 10 \\ 2 \\ \end{array} \right)\left[\left( \begin{array}{ccc} 8 \\ 2 \\ \end{array} \right)\left( \begin{array}{ccc} 6 \\ 2 \\ \end{array} \right)\left( \begin{array}{ccc} 4 \\ 2 \\ \end{array} \right)\right]^2}{(6!)(4!)(12!)(8!)} = \frac{1}{8847360}$$

Was this scramble purposefully constructed? Or was it generated by a program? I only just now noticed that it is only 20 turns. Plus the first 13 turns fit the {U,D,R2,L2,F2,B2} sub-group that Cube Explorer uses when solving cases in the final stage.

Last edited:

#### Mike Hughey

Staff member
Was this scramble purposefully constructed? Or was it generated by a program? I only just now noticed that it is only 20 turns. Plus the first 13 turns fit the {U,D,R2,L2,F2,B2} sub-group that Cube Explorer uses when solving cases in the final stage.

Yes, I constructed it purposefully and then used Cube Explorer. I just wanted to see you miserable. Sorry for putting you through that.

I think the worst I've ever had with a real scramble was 5 edge cycles and 2 corner cycles. Still awful, but not quite as bad as that fake scramble.

#### cmhardw

Yes, I constructed it purposefully and then used Cube Explorer. I just wanted to see you miserable. Sorry for putting you through that.

I think the worst I've ever had with a real scramble was 5 edge cycles and 2 corner cycles. Still awful, but not quite as bad as that fake scramble.

No it's fine, it gives me some perspective on how not-so-bad most scrambles really are To be honest, I probably would never have noticed that this was a faked scramble until I calculated the probability of a scramble like that. I certainly didn't notice anything fishy with the scramble algorithm at first. I only got suspicious when I saw that it was an approximately 1 in 9 million chance that a scramble like that would come up

I think I've had 4 two cycles in edges before, but I'm not sure if that was on 3x3x3 or on wings on a bigger cube. While memo'ing, when I still thought that the scramble was real, I kept thinking "bet there'll be one more two cycle. Yep... there it is. Bet there'll... yep there's another one"

Last edited:

#### Mike Hughey

Staff member
I think I've had 4 two cycles in edges before, but I'm not sure if that was on 3x3x3 or on wings on a bigger cube. While memo'ing, when I still thought that the scramble was real, I kept thinking "bet there'll be one more two cycle. Yep... there it is. Bet there'll... yep there's another one"

I'm sorry - I guess I'm a really bad person. I shouldn't be laughing this much. But I can't help it.

#### JonnyWhoopes

Then again, it can be bad even without orientation. Try this scramble:
D R2 D F2 L2 D F2 L2 U L2 U' B2 U' R F' R F L' R D2

Haha, I just tried that scramble. It's ridiculous for execution, but if you're using visual memo it's very fast. I got a 1:15 on it, and I average somewhere around 1:45.

#### aronpm

##### Member
My Letter Pair Word List. Used for everything (2-7bld, multi). It uses almost every type of word (adjectives, nouns, verbs, adverbs, etc) but I'm not studying English so maybe that's not true

Memorization examples:
scramble: F' U2 R' U' B2 L2 R D' R' B' R2 F2 L U R' B' F' D U' L F2 R' L F B
Edges: DX (Dexter) QJ ('s QJ) PK (peaked) FM (at my family) SV (so Jesus) G (is good)
Dexter's QJ peaked at my family so Jesus is good.

Corners: BU (boo) PJ (pyjama) QLS (quills)
boo pyjama quills
scramble: D2 B2 U' F' D' F2 R' U2 D B' D L2 R' B F2 U' D2 R L2 D2 L2 B2 L F2 U'
Edges: FO (my foe) GI (is Jai) DQ (because he was disqualified) JL (from jelly) D (dude)
my foe is Jai because he was disqualified from jelly dude.

Corners: SH (shutup) DUBG (doo bug) J (jay)
shut up doo bug jay
etcetc

#### Marcell

##### Member
I got a little confused over A9 and B9 comms. How come D2 M2 D R2 D' M2 D R2 D is listed as a B9? The way I see it, this alg is built up as
A: D R2 D'
B: M2
P: D2
executed as P BAB'A' P'
meaning it is an A9.
Also, the alg D2 M D R2 D' M' D R2 D has got exactly the same structure, but it is listed as an A9.

My other question is: I think that S2 D' S U' S' D S U S is a STM-optimal B9, is that right? It's just that it isn't listed on Chris' webpage.

#### cmhardw

I got a little confused over A9 and B9 comms. How come D2 M2 D R2 D' M2 D R2 D is listed as a B9? The way I see it, this alg is built up as
A: D R2 D'
B: M2
P: D2
executed as P BAB'A' P'
meaning it is an A9.
Also, the alg D2 M D R2 D' M' D R2 D has got exactly the same structure, but it is listed as an A9.

You are correct, that alg is an A9. That is either a typo or mistake when Daniel and I made the page. I'll correct it when I get home.

My other question is: I think that S2 D' S U' S' D S U S is a STM-optimal B9, is that right? It's just that it isn't listed on Chris' webpage.

That case is listed on the site:
(UR BU DF) B' R B' M B R' B' M' B2 (9 STM)** A9

Also, your alg is an A9 alg:
P = S2
A = S U' S'
B = D'

Perform as PBAB'A'P'

#### Marcell

##### Member
You are correct, that alg is an A9. That is either a typo or mistake when Daniel and I made the page. I'll correct it when I get home.
Alright, so I got it right after all. That's good news.

Also, your alg is an A9 alg:
Yeah, I just realized that that was an A9. And I know there's another A9 already in the list, but I thought you collected all the move optimal commutators for each case.

#### cmhardw

5x5x5 BLD
12:07.30 successful double alarm clock solve wearing earmuffs. I set two alarm clocks shortly before starting my solve, each for just a couple of minutes into the future. One alarm clock goes continuously until shut off, and one rings for 1 minute, then sleeps for 4, then rings for 1 , etc. I don't do this often, but it always feels amazing when I get a success while doing this!

#### riffz

##### Member
5x5x5 BLD
12:07.30 successful double alarm clock solve wearing earmuffs. I set two alarm clocks shortly before starting my solve, each for just a couple of minutes into the future. One alarm clock goes continuously until shut off, and one rings for 1 minute, then sleeps for 4, then rings for 1 , etc. I don't do this often, but it always feels amazing when I get a success while doing this!

really good time too!

#### miniGOINGS

##### Member
5x5x5 BLD
12:07.30 successful double alarm clock solve wearing earmuffs. I set two alarm clocks shortly before starting my solve, each for just a couple of minutes into the future. One alarm clock goes continuously until shut off, and one rings for 1 minute, then sleeps for 4, then rings for 1 , etc. I don't do this often, but it always feels amazing when I get a success while doing this!

I assume this is training for solving in noisy/unpredictable (even though I guess this is predictable...) environments? I hadn't really thought about that before, but it's definitely something I would do if I did blind.

#### amostay2004

##### Member
I think noises are much less of a distraction to BLD than people talking in the background. I can BLD in a noisy environment as long as I don't hear any clear conversation. It's really hard to focus when you hear a couple of people talking in the background when you're BLD-ing, because you're automatically distracted about what they are saying.