WANTED: New Dedge Flip Algorithm!

[Stefan]

F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' F2 Lw U2 Lw'

Is that supposed to scramble the cube?
=> http://alg.garron.us/

 

[deadalnix]

Or do you know how to solve parity using a commutator? (I do).

???????

EDIT : F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' U2 Lw U2 Lw'

 

[Kenneth]

Or do you know how to solve parity using a commutator? (I do).

???????

EDIT : F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' U2 Lw U2 Lw'

Thank's, don't now what happen there

I showed Joey in the Swedish #rubik at the Quakenet just a while ago, after he saw my earlier post.

[15:47] <+j`ey> Kenta: parity in a commutator?
[15:48] <Kenta> yes
[15:48] <+j`ey> show it?
[15:48] <SimonWestlund> -7 here
[15:48] <Kenta> with little setup add on
[15:48] <+j`ey> conjugated commutator
[15:48] <+j`ey> (like a A perm!)
[15:48] <Kenta> you can commutate a "J-PLL" in the second layer
[15:49] <+j`ey> hm
[15:49] <+j`ey> can you write it out?
[15:49] <+j`ey> (i know that's how lucasparity is done
[15:49] <Kenta> then it is just a few more turns to get double parity
[15:49] <+j`ey> it's 1C here
[15:49] <Kenta> sure, but it takes some time
[15:51] <Kenta> for "J-PLL" : B2 sets up a triple of two centres and one edge at B side, r slice, edge in D
[15:51] <Kenta> a second triple is from start at the F side l slice, edge in D
[15:52] <Kenta> so you do Dww2 to swap
[15:52] <Kenta> three bottom layers
[15:52] <Kenta> then you shift the triple that nos is on B for one in U using r' U2 r
[15:52] <Kenta> now*
[15:53] <Kenta> then another Dww
[15:53] <Kenta> and restore
[15:54] <Kenta> the whole alg for that will be : B2 Dww2 r' U2 r Dww2 r' U2 r B2 ... done
[15:54] <Kenta> now you got a J in the l slice
[15:54] <Kenta> easy to understand
[15:55] <Kenta> then if you work with diffrent setups and resoration, before, afer or even within the alg you can solve most cases of two swapped edges
[15:56] <Kenta> Double parity is the case that is nearest the turns in the basic commutator so I write that one too..
[15:56] <+j`ey> hm
[15:57] <+j`ey> following so far
[15:58] <+j`ey> i gotta go now, but i will read back when i get home
[15:58] <Kenta> (x') B2 Dww2 r' U2 r Dww2 r' U2 r B2 (x) l' U2 l' U2 l
[15:59] <+j`ey> but but
[15:59] <Kenta> there it is
[15:59] <+j`ey> nice
[15:59] <Kenta>
[15:59] <+j`ey> not a pure comm tho
[15:59] <Kenta> no, it needs a litte more
[15:59] <+j`ey> hm
[16:00] <Kenta> that's the 5 turns in the end
[16:00] <+j`ey> that messed up my centres :S
[16:00] <Kenta> hehe
[16:00] <Kenta> you must hve done wrong ot I wrote it wrong
[16:00] <Kenta> I check it...
[16:01] <+j`ey> it does a 3cycle of centres :S
[16:01] <Kenta> not for me
[16:01] <Kenta> ??
[16:02] <Kenta> it works fine here
[16:02] <Kenta> did you performe the x's
[16:03] <Kenta> Works for me but... I'm a confused person, sometimes I'm sure I did all right but I did not
[16:04] <Kenta> Anyone else who can test it to check?
[16:05] <KConny> j`ey You will like this: http://www.youtube.com/watch?v=UHH_5EClbAU&feature=related
[16:06] <KConny> An old swedish comedy show.
[16:06] <GunnarKrig> KConny: haha
[16:06] <+j`ey> 14:07:35 <Swordsman_Kirby> [09:06:21]["T. Sun"] Are you going to school anytime this week?
[16:06] <+j`ey> oops
[16:06] <+j`ey> KConny: can you check this
[16:06] <+j`ey> 14:59:02 < Kenta> (x') B2 Dww2 r' U2 r Dww2 r' U2 r B2 (x) l' U2 l' U2 l
[16:06] <Kenta> wrong channel ^^
[16:07] <KConny> Don't have a cube here.
[16:07] <+j`ey> O_O
[16:07] <+j`ey> ill use alg.garron.us
[16:07] <KConny> "Vem fan är Selma Lagerlöf"
[16:07] <GunnarKrig> ;D
[16:08] <Kenta> Remember Dww are 3 layers, not 2
[16:08] <KConny> Så jävla grym hon brunetten är.
[16:08] <+j`ey> oh
[16:08] <+j`ey> KConny: i see
[16:08] <+j`ey> erm
[16:08] <Kenta> you can write it diffrenty and use U2 but it is not that pedagogic then
[16:08] <+j`ey> Kenta: i eee
[16:08] <KConny> "Kan de ha varit så att hon hade en del vänner"
[16:08] <GunnarKrig> "Ska inte du komma med en hypotes nu också??"
[16:08] <Kenta> it makes it hardet too see what's happening
[16:09] <KConny> Hahaha
[16:10] <+j`ey> Kenta: it works
[16:10] <GunnarKrig> KConny: jag älskar Herman Lindqvist-parodierna
[16:10] <Kenta> nice

Why on earth a J in the second layer you may ask... I read at Akimotos page back in 2006 that you need to swap blocks of centres togehter with the edges to make it possible to solve a parity. So I tried a few diffrent commutators that did exactly that and soon found this, that is useful for all cases of two swapped edges from two opposite deges in the last layer (if you use diffrent setups).

 

[Stefan]

EDIT : F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' U2 Lw U2 Lw'

Is that supposed to scramble the cube?
=> http://alg.garron.us/

 

[Kenneth]

EDIT : F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' U2 Lw U2 Lw'

Is that supposed to scramble the cube?
=> http://alg.garron.us/

Man!, I try again

F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Rw' F2 Rw U2 Lw'

Tried it three times and it worked fine so I hope it does not scramble your cube anymore

 

[deadalnix]

F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Lw' U2 Lw U2 Lw'

Why do you put R move in it ? (I use the reverse one and mirrored to get R insted of L).

 

[Kenneth]

F2 Lw2 F2 U2 Lw' U2 Lw U2 Lw2 F2 Lw' U2 Lw U2 Lw'

Why do you put R move in it ? (I use the reverse one and mirrored to get R insted of L).

Yes yes, you can do it in several ways, that was just the basic turns that are easy to see. For fast execution you will of course do it a bit diffrently, like x - x' between the U and F turns so they all become U2's

This is the forum about why we turn, not how we turn

Edit: the answer to your question is for real that I don't use it like that, I do direct solving and this is a case in my wery last ELL step so I use it with single slices to preserve corners, your version does not work then.

 

[Christopher Mowla]

Do you know what you are doing or do you just go by chance, test some tuns until you find something that works?

I actually think I have been having too much "luck" to be "lucky" anymore. Yes, the first turns in the algs I have given out so far were tests...tests to set up a certain scenario that I wanted. Once I got that scenario, I finished it. (Hence, yes, I definitely know what I am doing). Have you seen my 21q odd permutation break (not double parity)? (It is not pure, but it reduces a 4X4X4 with odd parity to a 3X3X3 cube).

r U2 r' U2 r D2 r D2 r' F L' F' U' r' D2 r
If, instead you do it with single r slice turns followed by the following moves: D2 U F L F' , it restores the cube significantly.


In addition I have found several brief pure double parity algs without computers as well as finding, if you having heard the news, a 19 BQTM pure edge flip alg for the 4X4X4. (And, based on how that alg looks, it is more likely, in my opinion, that I win the lottery, than to test some moves to find that one because the entire cube is scrambled immensely during the alg...and then solved back in very few moves--using many different slices).
================================================
reThinker,

I have revised my technique to not mess up the back bottom cross-edge and still keep the move count the same:

I found a 30q alg that uses the front-left F2L slot and U-layer
z2
r U2 l U2 r'
F R F' U l' F2 l
U B2 l B2 l'
U2 F' R' F D' F2
z2

I have also found a 29q alg that uses the front-right F2L slot, induces double parity, and leaves the U layer free.
x' r U2 l U2 r' F R F' U x l' U2 l U2 x' E2 x l' U2 l U2 F2 x' U L' x' y2
But adding R U R' U' makes a pure double parity….:fp

This was actually a common occurence when I was trying to find algs that use the U layer and 1 F2L slot.

For example, here is a 27 BQTM regular pure double parity that I stumbled on trying to find an alg that uses the F2L layer and U layer:
r U2 l U2 r' F R F' U l' F2 l F2 E2 y2 l' D2 l U F R' F'

Here's a weird pure double parity:
r U2 l U2 r' F2 l' F2 l F' R F' U l' F2 l B2 l B2 l' F2 U' F R' F

And here is another weird pure double parity fix:
r U2 l U2 r' F R F' U l' F2 l U B2 l B2 l' U' F2 U' F R' F'

The 4 algorithms which I have just given above could suggest that double parity algorithms that use a F2L slot are very related to pure double parities, for I used the same method and found several pure double parities (as you have seen). If, in the future, someone comes up with a "proper proof" for this, then, I would say that the alg you really seek is brief pure double parity algs.

Here's another weird pure double parity: (23q)
r U2 r' F2 l' B2 l B2 l' D2 l D2 l' F2 U2

And, as far as me getting lower move algs (less than 29q) that kind of meets your description, here's a 28q that uses the back two F2L slots and the U layer:
x2 r U2 l U2 r' F R F' U l' F2 l U B2 l B2 l' U' F2 U' R' F' x2

And, for your possible interest, here's a non-pure 19q double parity. I didn't find it by "chance", as far as the last 12 moves are concerned. The first 7 moves provided a good set-up for a brief double parity...I sensed this and finished it (I was going to finish it at first because it wasn't a case that needed to use a F2L slot--it appears that all 19q double parity algs don't need to use a F2L slot--I finished it for fun).
r U2 r' U2 r' D2 r D2 r' B2 l U2 l'

Edit:
Here is a 21q, edge flip alg. It flips the front-bottom edge, but it does not preseve the front and bottom centers. It could be useful for solving supercubes or using the cage method:
r U2 l' U2 r' x2 l' U2 r' U2 r l U2 l' U2 l x2

Note: I found all of these algs in the last 2 days (as well as the other ones I have posted recently).

Date: 1/10/10

 

[fanwuq]

I have also found a 29q alg that uses the front-right F2L slot, induces double parity, and leaves the U layer free.
x' r U2 l U2 r' F R F' U x l' U2 l U2 x' E2 x l' U2 l U2 F2 x' U L' x' y2
But adding R U R' U' makes a pure double parity….:fp

This was actually a common occurence when I was trying to find algs that use the U layer and 1 F2L slot.

For example, here is a 27 BQTM regular pure double parity that I stumbled on trying to find an alg that uses the F2L layer and U layer:
r U2 l U2 r' F R F' U l' F2 l F2 E2 y2 l' D2 l U F R' F'

Here's a weird pure double parity:
r U2 l U2 r' F2 l' F2 l F' R F' U l' F2 l B2 l B2 l' F2 U' F R' F

And here is another weird pure double parity fix:
r U2 l U2 r' F R F' U l' F2 l U B2 l B2 l' U' F2 U' F R' F'

The 4 algorithms which I have just given above could suggest that double parity algorithms that use a F2L slot are very related to pure double parities, for I used the same method and found several pure double parities (as you have seen). If, in the future, someone comes up with a "proper proof" for this, then, I would say that the alg you really seek is brief pure double parity algs.

I noticed that you are just inserting F R F' U in the middle of the algs to replace one paired up dedge with another paired up edge that is part of the F2L while flipping another dedge in the M-ring. Then you undo the F R F' U at the end of the alg to restore the cube.
If you take out the quarter moves in the middle of your first alg, it reduces to
x' r U2 l U2 r' x l' U2 l U2 x' E2 x l' U2 l
So you are not really using the quarter turns to induce parity, you are just using them to convert another parity alg to the dedge flip form. This can be done more efficiently without involving a F2L dedge.

r U2 l' U2 r' x2 l' U2 r' U2 r l U2 l' U2 l x2 looks very cool!
http://alg.garron.us/?alg=r_U2_l-_U...U2_r_l_U2_l-_U2_l_x2 &cube=4x4x4&notation=WCA

 

[Christopher Mowla]

I noticed that you are just inserting F R F' U in the middle of the algs to replace one paired up dedge with another paired up edge that is part of the F2L while flipping another dedge in the M-ring. Then you undo the F R F' U at the end of the alg to restore the cube.
If you take out the quarter moves in the middle of your first alg, it reduces to
x' r U2 l U2 r' x l' U2 l U2 x' E2 x l' U2 l
So you are not really using the quarter turns to induce parity, you are just using them to convert another parity alg to the dedge flip form. This can be done more efficiently without involving a F2L dedge.

Well, I understand why you would think so due to the appearance of preliminary moves and reverse of those preliminary moves. However, I do use those moves to set up the cube for parity fix. I undo them later to restore the cube back to a reasonable state (and, I was careful what I did after those outer-layer quarter turns, so that, when I reversed them, they worked positively for restoring the cube). And by doing this, I managed to leave one F2L slot unsolved (and the U layer).

As a matter of fact, I used a very similar technique for the 21q non-pure OLL parity alg (not double parity) which I have listed in this thread and in my thread:
r U2 r' U2 r D2 r D2 r' F L' F' U' r' D2 r
This alg is two 4 cycles and a 2 cycle, thus yielding just an OLL parity fix.

That should definitely proove to you that I do use the quarter turns to induce parity. The first moves in the alg above prepare for only an OLL parity fix, while the first moves in the alg for rethinker is for double parity.

r U2 l' U2 r' x2 l' U2 r' U2 r l U2 l' U2 l x2 looks very cool!

Thanks. I am glad you like it.

 

[reThinking the Cube]

Have you seen my 21q odd permutation break (not double parity)? (It is not pure, but it reduces a 4X4X4 with odd parity to a 3X3X3 cube).

(Rr) U2 (Rr)' U2 (Rr) D2 (Rr) D2 (Rr)' F L' F' U' (Rr)' D2 (Rr)
If, instead you do it with single r slice turns followed by the following moves: D2 U F L F' , it restores the cube significantly.

Yep. I appreciate the effort, but isn't this more about proving to everyone that YOU can actually find/create some algs without using a computer. To ME - any algorithm that unsolves what has already been solved is just an ANTI-ALGORITHM. This means anything that unpairs dedges, undoes the cross, tampers with any part of a previously solved F2L pair, or unsolves any of the centers. I am really not seeking algs that do any of that. I don't think that there will be a practical use for these types of anti-algorithms any time soon.

The 4 algorithms which I have just given above could suggest that double parity algorithms that use a F2L slot are very related to pure double parities, for I used the same method and found several pure double parities (as you have seen). If, in the future, someone comes up with a "proper proof" for this, then, I would say that the alg you really seek is brief pure double parity algs.

For right now, here is MY newest and best algorithm for OLL/PLL parity:

DoubleTrouble™ - (r U2 r' U2 r' D2 r D2) x' (r' U2 r U2 r' U2) 2s2

Notation is SiGN, but "r" can be done as single slice and then U2 2s2 can be taken out. You do the math.

If you REALLY want to provide me with what I really seek, then see if you can find a "better" alg than this.

 

[reThinking the Cube]

[For right now, here is MY newest and best algorithm for OLL/PLL parity:

DoubleTrouble™ - (r U2 r' U2 r' D2 r D2) x' (r' U2 r U2 r' U2) 2s2

Notation is SiGN, but "r" can be done as single slice and then U2 2s2 can be taken out. You do the math.

If you REALLY want to provide me with what I really seek, then see if you can find a "better" alg than this.

DoubleTrouble™ - r U2 r' U2 r' D2 r D2 x' r' U2 r U2 r' U2 2s2

Here is the inner slice version:2R U2 2R' U2 2R' D2 2R D2 x' 2R' U2 2R U2 2R'

Inner slice version: You will see if you look at this carefully, that it is not (as it appears) just a simple 4-cycle dedge commutator, but really is an 8-cycle of edges (odd OLL permutation), that remakes the M-slice dedges as a 4-cycle of dedges (odd PLL permutation).

 

[rachmaninovian]

r U2 l' U2 r' x2 l' U2 r' U2 r l U2 l' U2 l x2 indeed is useful for me

can you possibly find even better cage parities?

 

[Lucas Garron]

until somebody can tell me the best way to do linking

1. Click the textarea on the upper right. The whole field should select.
2. Copy.
3. Paste into post.

 

[reThinking the Cube]

until somebody can tell me the best way to do linking

1. Click the textarea on the upper right. The whole field should select.
2. Copy.
3. Paste into post.

Thanks Lucas. I fixed it. That page is sweet.

 

[Christopher Mowla]

If you REALLY want to provide me with what I really seek, then see if you can find a "better" alg than this.

DoubleTrouble™ - (r U2 r' U2 r' D2 r D2) x' (r' U2 r U2 r' U2) 2s2

How about this alg (21q instead of 23q): This is in SiGN notation too (so you can copy and paste to Lucas' site).

(r U2 r' U2) x' (r U2 r' U2) x2 (r U2 r U2 l') 2s2

It is 21q (as opposed to 23q) only because I don't need to rotate the U Layer Twice at the end (which takes advantage of a "free" LL, just as you like).

It is very similar to yours (however you got it). It just "flips" the back top edge instead of the top-front edge.

I am no expert at memorization or speed, but it appears to me that my alg is easier to memorize and faster to execute than your "DoubleTrouble". I hope you like it (if so, then, none of my research was in vain...I found this 19q double parity fixer in the process of looking for the algs that you thought were ANTI-Algorithms).

I hope you like it...let me know what you think.

 

[Kenneth]

I actually think I have been having too much "luck" to be "lucky" anymore. Yes, the first turns in the algs I have given out so far were tests...tests to set up a certain scenario that I wanted. Once I got that scenario, I finished it. (Hence, yes, I definitely know what I am doing). Have you seen my 21q odd permutation break (not double parity)? (It is not pure, but it reduces a 4X4X4 with odd parity to a 3X3X3 cube).

(Rr) U2 (Rr)' U2 (Rr) D2 (Rr) D2 (Rr)' F L' F' U' (Rr)' D2 (Rr)
If, instead you do it with single r slice turns followed by the following moves: D2 U F L F' , it restores the cube significantly.

LOL, who would use something like that??? to me it is a scramble

And it is not smooth execution eiter... most, not to say all cubers prefer smoothness instead of one or two q-turns lesser. No one need your algs I'm afraid.

 

[Christopher Mowla]

LOL, who would use something like that??? to me it is a scramble

And it is not smooth execution eit[h]er... most, not to say all cubers prefer smoothness instead of one or two q-turns lesser. No one need your algs I'm afraid.

What don't you understand about not bringing tension? This is not healthy for the community. I never said that my alg for that case is good for speedsolving.

This is the theory section of this site. Obviously you do not like me because you consistently write in language that is very sarcastic (and I don’t see any reason for you to do so, especially since I made it clear to you before that I don’t want to fight anymore, anywhere). Whether an alg is fast or not, it is still an alg. Just because I didn’t find a fast alg to solely break OLL parity doesn’t mean anything…I found a 21q that is interesting.

Yes it does scramble the cube. I never said it preserved anything (except for the centers and edges). In fact, I think I made it very clear what is does and does not do (maybe you should read my entire post and not just the algorithms I post).

And I think the alg I just posted for reThinker is pretty good for speedsolving...

 

[reThinking the Cube]

If you REALLY want to provide me with what I really seek, then see if you can find a "better" alg than this.

DoubleTrouble™ - (r U2 r' U2 r' D2 r D2) x' (r' U2 r U2 r' U2) 2s2

How about my DoubleTrouble Fixer: (21q instead of 23q): This is in SiGN notation too (so you can copy and paste to Lucas' site).

(r U2 r' U2) x' (r U2 r' U2) x2 (r U2 r U2 l') 2s2

It is 21q (as opposed to 23q) only because I don't need to rotate the U Layer Twice at the end (which takes advantage of a "free" LL, just as you like).

"Just as you LIKE"? LL unsolved just provides an added degree of freedom. I would actually prefer less change, but usually that restriction produces longer/messier algs. To be kind towards finding something, YES, I am recommending that you take advantage of LL freedom. But here you just take my alg, reorder it, and then "take advantage" of not having to technically restore LL, by leaving out "U2" at the end, and claim a much better alg. Of course, I could have done the same thing, but then the finished position wouldn't have been as recognizably useful. But to be fair, I have to admit that you "CAN" leave out the LL fixing move "U2", and applause........save 2q. Even though it is technically correct, the resultant location of the flipped edge at UB is not practical for me, and would result in an added cube rotation. So the artificial gain of 2q is also mute for that reason. Here is a link to see it, with the finishing move "U2" added.

r U2 r' U2 x' r U2 r' U2 x2 r U2 r U2 l' 2s2 U2

Here is what it would look like with just inner slice turns:

2R U2 2R' U2 x' 2R U2 2R' U2 x2 2R U2 2R U2 2L' 2s2 U2

Compare this to:

DoubleTrouble™ - r U2 r' U2 r' D2 r D2 x' r' U2 r U2 r' U2 2s2

and the inner slice version:2R U2 2R' U2 2R' D2 2R D2 x' 2R' U2 2R U2 2R'

it is very similar to yours (however you got it). It just "flips" the back top edge instead of the top-front edge.

FYI - I came up with it in my head, and no computer was used for this one. I cannot promise the same for some others I might come up with.

I am no expert at memorization or speed, but it appears to me that my alg is easier to memorize and faster to execute than your "DoubleTrouble". I hope you like it (if so, then, none of my research was in vain...I found this 19q double parity fixer in the process of looking for the algs that you thought were ANTI-Algorithms).

I hope you like it...let me know what you think.

Because it is very similar to DoubleTrouble™, it is IMO - relatively easy to memorize and relatively fast to execute. I do like it, but I personally don't think it is any EASIER to memorize, or any FASTER to execute. The beginning is nice, but the extra x2 rotation does not fully compensate for that, and the ending is also not as symmetrical as the original. Placing the flipped dedge at UB would also require (at least for me) some extra cube rotations or face turns.

Failing to restore LL for no other reason than to gain 2q, shows that you are very resourceful, and want to be proven right at all times, even when YOU think that you are wrong. :rolleyes:

Your research in definitely not in vain, and the DoubleTrouble™ parity alg, proves it.

 

[Kenneth]

What don't you understand about not bringing tension? This is not healthy for the community. I never said that my alg for that case is good for speedsolving.

This is the theory section of this site. Obviously you do not like me because you consistently write in language that is very sarcastic (and I don’t see any reason for you to do so, especially since I made it clear to you before that I don’t want to fight anymore, anywhere). Whether an alg is fast or not, it is still an alg. Just because I didn’t find a fast alg to solely break OLL parity doesn’t mean anything…I found a 21q that is interesting.

I'm not trying to be sarcastic at all, I'm just trying to tell you that what you do is not useful to anybody, your work is for nothing and obviously you cannot see it yourself. You don't even prove anything because you cannot go throu all combinations to see if there is a shorter or better alg. The only way to do that is to brute force using computers but that is not possible either because of too many combinations.
....

An alg that fixes parity that scrambles that much is only useful if you look for parity before you start the 3x3x3 part of the solve and that takes eons, much longer than it takes to execute a whole bunch of algs.

Maybe for FMC but that is always counted in half turns...

So, if you find the 14 turn alg, then gratz because that would be a gain. 14 is the optimal alg for the same case of 3x3x3 corners so shorter than that is not possible, cube explorer proves that the brute force way. I will even give you "all" optimal algs so you got something to start from but I don't think you can get any of them (it's the same, just mirrors and inverses) to do the job without messing centres:

Searching depth 14

R2 F2 R F2 U L' U L U' R U R' U' R (14f*) //order 2
R' U R U' R' U L' U' L U' F2 R' F2 R2 (14f*) //order 2
L U' L' U L U' R U R' U F2 L F2 L2 (14f*) //order 2
L2 F2 L' F2 U' R U' R' U L' U' L U L' (14f*) //order 2

Searching depth 15