[Help Thread] Megaminx/Master Kilominx/Gigaminx BLD discussion

[Roman]

Wow, there is a thread for it! I'm with you guys
I am currently on my way to my first success, and then maybe I will try MK or Gigaminx.
Somewhy I found megaminx BLD very interesting and challenging.

 

[Jacck]

Slowcubers-Blind for Megaminx corners

There is no specific alg, you just solve a corner with an U' and the next with an U. Its a bit like U2 for 4x4x4-centers.

And again: it works perfect with the Gigaminx-corners, too!

Memo (the same idea as with the edges):
Corners from 1 (white-green-purple) to 5 (white-purple-yellow) on Top, next "layer" 6 (green-brightblue-purple) to 10 (purple-orange-yellow), then 11 (green-brightyellow-brightblue) to 15 and 16 (grey-brightblue-brightyellow-brightblue) to 20 on bottomside. If it has to be twisted clockwise, I add 20, for counterclockwise 40.
"Orientated" corners are with white side up or grey side down or on the front of the related front-face (look at the first color in the triples, this is the oriented side).
With new cycles I try to bring in the Top-Corners with the correct orientation, the other Top-corners are orientated at the end.

Example for pair 20,13:
setup20 U' setup20' setup13 U setup13'

Buffer is 1 and the helper is 2 (??? no target??? I think the target is changing, just look at the setups and the examples and you’ll find it out!)

setups (DL is brightblue face, DR is brightyellow face):
6: D+ R' DR' R (green-brightblue-purple)
26: DL R' DR R (brightblue-purple-green)
46: DL DR R' DR' R (purple-green-brightblue)
11: DR R' DR' R (green-brightyellow-brightblue)
31: DL R' DR2 R
51: R' DR R
16: R' DR2 R
36: F DR2 F'
56: D+ R' DR2' R
All the others need only some D+ before and some D- afterwards.

example pair 33,20 (brightgreen-pink-blue,grey-orange-brightblue):
D-- (bring 33 to front, to 31)
DL R' DR2 R (setup for 31)
U'
R' DR2' R DL' (setup' for 31)
D++ (bring back 33 to 33)
D+ (bring 20 to front, to 16)
R' DR2 R (setup for 16)
U
R' DR2 R (setup' for 16)
D- (bring back 20 to 20)

With a top-edge in a pair it is reduced to a normal commutator. Just setup the other corner and take the top-corner as „helper“.

Don’t forget the misorientated top-corners at the end! Just take R F' R' F and so on.

 

[Hssandwich]

I have finished planning my Megabld corners method (kinda OP for megaminx). My memo method is done with one pair per sticker. I use A+A-Z for the first 26 targets, then E + A-Z for the next 26. I use emoticons for the remaining 5.

Method:

Spoiler

Corners:
Memo clockwise round each face and clockwise round the puzzle.
General rule of thumb: buffer is ULB and set up to RFD and do: R U' R' U' R U R' F' R U R' U' R' F R then for your next pair do the alg twice/ inverted.
My orientation: grey top, light green front
A + letter:
Grey: A—see below// B—R2 rfD'// C—F // D— F R'
Orange: E— F2 //F—lfD rfD // G—lfD2 F' // H— blD lfD F2'
L.Green: I—F2' rfD // J— F2 rfD // K— F rfD // L— rfD // M— F' rfD
PANK: N—R' // O—R2' // P— R2 // Q—R // R— done
Cream: S—R' F // T—see below // U— urB rfD' R2 // V—rbD' R2 // W—rbD' R
L.Blue: X—see below // Y—lbD lfD2 F' // Z—blD2 lfD F2'
E+ letter:
L.Blue: A—bD' urD rfD' R2
Yellow: B—F2' // C—lfD' F2' // D—lfD2' F2' // E—lfD2 F2' // F—lfD F2'
D.Blue: G— rfD' F' // H—rfD2' F' // I— rfD2 F' // J—rfD F' // K—F'
Red: L—rbD' rfD' // M—rbD2' rfD' // N— rbD2 rfD' // O—rbD rfD' // P—rfD'
D.Green: Q—dB' rbD2' R//R—B2' rbD2' R//S—dB2 rbD2' R//T—dB rbD2' R// U—rbD2' R
Purple: V—blD lfD2 rfD//W—lfD2 rfD//X—blD' lfD2 rfD//Y—blD2' lfD2 rfD// Z—blD2 lfD2 rfD
White: —lfD' F' // —D' lfD' F' // :/—D2' lfD' F' // o_O—D2 lfD' F' // ^-^—D lfD' F'
Special corners:
xx AA-skip AA // CREATES PARITY
AA xx-setup xx + do alg twice // CREATES PARITY
xx AX- setup xx + U' R' rfD' R U L U2' L lD L' U2 L lD' L2' U' R' rfD R U
AX xx- setup xx + U' R' rfD' R U L2 lD L' U2 L lD' L' U2' L' U' R' rfD R U
xx AT- setup xx + R2 U R U' lB' U R' U' lB R2'
AT xx- setup xx + R2 lB' U R U' lB U R' U' R2'

 

[Jacck]

Nice to see, that there is really going on something! And Roman did it today in his 2nd attempt!

General rule of thumb: buffer is ULB and set up to RFD and do: R U' R' U' R U R' F' R U R' U' R' F R then for your next pair do the alg twice/ inverted.

Just to make it clear: with "next pair" you mean your next letterpair (one corner), so you solve as me only one cubie at the time?

And I would have another idea for your alg: U2 R' rfD' R U R' rfD R U2 (it doesn't affect the edges, but I think it is the same for the corners).

Interesting for me: I guess my method for corners could be more obvious, if I would "declare" RFD as the target. There would be different and a bit more setup-moves, ok. At the moment, I'm working with many targets.

An idea for using letters: If I had to memo the Megaminx-corners with letters, I would take AA to AE for the top-layer, BA to BE for the 2nd layer, CA to CE for the 3rd and so on. Twisted corners would have AI to AM and AQ to AU (the gap of 8 should be familiar for 3x3x3-bld-cubers). Since you will need 2 letters for one cubie, you can pack information about where it is and how it is orientated in that letter pair.

 

[Jacck]

@Roman:
Watching your video I think I could see, that you bring one edge to the top and then make a 3-cycle? So you don't use absolut free 3-cycles (which would be much harder)?
I'll have to think about that especially with my edges in the second layer (6-10 and 36 to 40). Their setups are quite long and bringing them up should be easy and would not disturb very much.

 

[Roman]

@Roman:
Watching your video I think I could see, that you bring one edge to the top and then make a 3-cycle? So you don't use absolut free 3-cycles (which would be much harder)?
I'll have to think about that especially with my edges in the second layer (6-10 and 36 to 40). Their setups are quite long and bringing them up should be easy and would not disturb very much.

My method: every time I need to solve next two edges, I bring one of them (any) to the top and another to one of four helper positions. Then I do a pure commutator and undo setups. All these commutators have this structure: UF buffer and the element on top are interchangable by moving the U layer, and a piece from the helper position can be moved to the buffer by doing (L' R ? L R'), where '?' is F', F, F2' or F2 - thus, a four helper positions. Examples: (U L' R F' R' L U' L' R F R' L); (U2' L' R F2 R' L U2 L' R F2' R' L) and so on

 

[Hssandwich]

Nice to see, that there is really going on something! And Roman did it today in his 2nd attempt!


Just to make it clear: with "next pair" you mean your next letterpair (one corner), so you solve as me only one cubie at the time?

And I would have another idea for your alg: U2 R' rfD' R U R' rfD R U2 (it doesn't affect the edges, but I think it is the same for the corners).

Interesting for me: I guess my method for corners could be more obvious, if I would "declare" RFD as the target. There would be different and a bit more setup-moves, ok. At the moment, I'm working with many targets.

An idea for using letters: If I had to memo the Megaminx-corners with letters, I would take AA to AE for the top-layer, BA to BE for the 2nd layer, CA to CE for the 3rd and so on. Twisted corners would have AI to AM and AQ to AU (the gap of 8 should be familiar for 3x3x3-bld-cubers). Since you will need 2 letters for one cubie, you can pack information about where it is and how it is orientated in that letter pair.

By next pair I mean the next piece, I'll clarify that in my post. I'll probably switch to your alg because it doesn't create 'parity' and it is super simple to reverse.

I think that the memo system that you are talking about was used by Dan Sarnelli in his megaminx blind attempts, so it definitely works.

 

[Jacck]

@Roman: That is a good way, I'll have to try it. I was afraid, that setting up the first edge could disturb the other one. But if you take either the first or the second, the most pairs should go without problems, especially for someone,who is used to commutators (I'm just beginning to lern them). And having 4 helpers can be very fast, too (but it is no slowcuber method anymore).

@Hssandwich: I haven't understood your parity-problem yet, cause there is no real parity on the Megaminx.
But another idea: if you take UFL as buffer and URF as "helper" the alg would be shorter with R' rfD' R U' R' rfD R U and you would have nearly the same as I do it. The difference would only be the way to setup (I prefer the D+ and D-, but yours are a bit shorter) and that I have more targets. But in the end, both are good solutions for only solving an edge at the time: quite easy to learn, a bit more moves, but very intuitiv - that's what I call slowcubers-style

 

[Hssandwich]

@Hssandwich: I haven't understood your parity-problem yet, cause there is no real parity on the Megaminx.

what I meant by parity was a 3 cycle of edges caused by only requiring to do the alg once or the inverse once. I should have just said a U-perm or an edge cycle but my wording was a little unclear.

 

[Jacck]

Hmmm, finally had a success again in 28:06.14, memo 15:25 (unconcentrated memo of the corners >8min).

I solved the edges again with the X-perms, see https://www.speedsolving.com/forum/threads/blindfold-failures-thread.4909/page-287#post-1174512
Only when the top-edges 2 and 5 (the side-buffers) where involved, I solved the pair with the L' R F2' L R' U' L' R F2 L R' U thing.

I don't know exactly why, but I feel more comfortable with the X-perms. Maybe I had more attempts with the X-perms, but I did the other way also quite a few times. So I think the main reason is, that there ist a bit less of thinking - and that fits better to my brain, especially, when I'm a bit tired like now.

 

[Jacck]

After 5 DNF, Megaminx bld in
20:32.94, memo 10 min, pb by more than 1:30 min.
Solved the edges like written in the last thread and was a bit lucky, that the 5 was orientated right and that I could start a new cycle for the 2, so I could bring it in orientated right, too.

video will be posted in the 6x6-blind-rankings soon.

 

[Jacck]

Triple post, but this should be here, too:

Masterkilominx blind:
3:13:21.50, memo 1:20:55 ----- EDIT: It was only 2:13:21.50 :rolleyes:
First success in third attempt - and I think, that ist the number of attempts, too, that ever have been made worldwide = UWR

Spoiler: video

 

[Dom]

Perhaps y'all could help me understand something.

I am a military instructor and one of my students recently just blew my mind. He solves the 3x3 using the beginner's method and it takes him a few minutes to do. He only has a Rubik's brand cube, so he doesn't know any finger tricks, but he can do blindfolded solves faster than he can unblindfolded. But that's not the strange part. He had never seen a Megaminx before, but I left one in the classroom after class one day and he stayed for about an hour and figured out how to solve it. The very next day he did a blindfolded solve on it in about 3 minutes. Is this more common than I think?

 

[Cale S]

The very next day he did a blindfolded solve on it in about 3 minutes. Is this more common than I think?

Well the UWR for Mega BLD is over 16 minutes so...

 

[Dom]

Well the UWR for Mega BLD is over 16 minutes so...

I see. 16 minutes...That includes the inspection, right?

 

[Dom]

Well the UWR for Mega BLD is over 16 minutes so...

My student inspected for 10 minutes and then solved it in about 3. Hopefully I'll post a video soon.

 

[Jacck]

The video would be very interesting!
Because here is the unofficial ranking:
https://www.speedsolving.com/forum/threads/6x6-blindfolded-rankings-thread.41968/
and even Liliya managed only a memotime of 9 min. So 10 min would be extremely good for a beginner, because this needs to know the color-sheme (ok, I think, that our fastest blindfolders could do it in less than 5 min - but only with long practice).
And the best executiontime till now was over 7 min. 3 min would be really awesome:
corners can be done two with about 12 moves and you will have maybe 9 pairs, so these are 108 turns;
edges will need more moves, because you cannot use normal 8-move-commutators, so you will need perhaps 20 moves per pair. 14 pairs will be 280 turns, with the corners 388 turns and this in 3 min?
If your student is Yu Da-Hyun, this might be possible, but she definitely saw a Megaminx before

So we'll have to wait for the video, and if he really has done it, we should have his name and country for the ranking.

But one thing to the 3x3x3: with advanced blindfold technique you have less turns as with the beginners method. So this is possible, but normally, someone who knows this blindfold techniques does know a bit more than the beginners method.
And yes: the Megaminx can be solved with the beginners method (ok, on the last layer, you have to experiment and use the algs a bit different, but it can be done).

 

[Jacck]

https://www.speedsolving.com/forum/threads/blindfold-accomplishment-thread.3582/page-103#post-178728
This happened in May 2009 and I thought it was time for a next
Pyraminx Crystal blind
49:49.17, memo 17:55 min
Did it with a QJ, which was the main challenge.

Spoiler: video

Haven't found any trace, that someone else has done it yet, so I claim place 2 worldwide for me
Memo was:
edges: 38,22 1,27 36,35 37,18 9,41 26,55 10,2 58,33 14,21 13,17 49,60 20,54 23,15 59,16 42,16
corners: 12,13 2,35 37,44 60,19 26,34 8,43 33,11 29,18 56,51 10/30
and I solved it with this system (colors as actual WCA megaminx-sheme, white on top, green in front):

edges:
buffer LB-U (yellow-white), target UF (white-green), helper RB-U (blue-white)
setup1 L R’ L’ R setup1’ setup2 R’ L R L’ setup2’
L is yellow, R is red
OK, the helper (it was flipped, too) needs a bit of thinking, but can be done with half of the alg. And flipping a edge goes with solving it first normal and then flipped.

corners:
Found probably the same as Istvan (8-mover repeated 7 times), but this seemed too long for a QJ.
Then I found a 8-mover, which seemed to be interesting:
R F’ R’ F RD F RD’ F’
it cycles 3 corners and 2*3 edges. My idea was to "eliminate" the edges-cycles after each corners-cycle. Before actual doing this, I found Stefan’s 14-mover in a video
RB L’ UBR’ DFL y L RB’ U’ RB L’ UBR DFL’ y' L RB’ U
which I used in my first attempt, but I couldn’t control the QJ enough when there were lock-ups and stuff.

So I thought again about my 8-mover, put it on top and searched for insertions for the undo-the-edge-cycles. Result:
F R F’ R’ U’ R’ L’ R L U FDR FDL R’ FDL’ FDR’ R
buffer is UFL, target is F-DR-DL, helper is URF
So it is: setup1 alg setup1’ setup2 alg’ setup2’
If it was easy, I put one corner of the next pair in URF and had to do the alg only once for the pair.

 

[bobthegiraffemonkey]

Impressive! Pyraminx crystal BLD seems awful because of the corners. I am curious to try to find my own corner alg(s) now though, but I would never want to actually attempt it.

 

[Jacck]

Stefan's alg is really good, if you have a non-lock-up-and-nearly-pop-out-all-the-time Pyraminx Crystal. The y's in my notation aren't necessary, just look how he performs it:

Don't think, that there could be something better.

OK, the corners are not that nice, but the edges are easy. Opposite to a Megaminx: it has easy corners and harder edges.