• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Last edges (Roux amoung other methods)

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
I have been working on an article in the wiki about a method that solves the last 8 / 6 / 5 edges. It is useful for Roux, Columns first, some CF methods and more...

Have a look:

http://www.speedsolving.com/wiki/index.php/E15_/_E35

I'm not compleatly done but the important things are there, it only lacks a few special orientation algos and a little more in depth description on the first part... and possibly a number of examples.

If you plan to learn Roux and know EPLL, then consider, recognition is faster using this (aspecially for the orientation part) but maybe not execution (my algs are short but maybe not the easiest ones, you need to practice them, but once you found the grips the shortness makes them fast).
 
Last edited:

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
So are these the steps for your Roux LSE:

1. Have DB solved
2. Orient the last 5 edges and possibly solve DF at the same time
3. Permute remaining 5 (or 4 if DF is solved) edges

If so, I used to use Step 3 when I was using a different method. I have all of the cases written down somewhere with short solutions, I'll try to find them. I think the average number of moves was 7 and they were all based on M and U, though yours are short also..
 

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
Exactly =)

I know MU algs for most cases but I think it is easier to learn if the the algs are diffrent from each other in several ways = not just diffrent turns but also the use of diffrent sides. That also make them easier to remember because you do not confuse an R d MU for being the same as a pure MU. That's why I did not put only pure MU's in the tutorial.

But I think I will put up a page that lists algs in the same manner as the CxLL or OLL and PLL pages.

Thanks to Jason MU algs for all cases without leading AUF can be found in this thread:
http://www.speedsolving.com/forum/showthread.php?p=52705

As you can see I invaded the tread after a while :D

If you find your algs, then you can post them here because I will not create that page now. First I will write a tutorial for my way to solve F2B (I used to call it "FS-F2L" = solve half a "cross" at RD / LD but ignore D centre, then build and place four pairs to compleate F2B, then CMLL and E15 = my big cube method but for 3x3x3 (you are going to love the 2x2x2 version = FL + CLL =) )

Advanced you go: place BD and centres before CMLL and then, if possible orient edges while doing the corners (insert M turns into the CLL). If the pice in FD is oriented then go for all oriented and if FD is unoriented then go for three unoriented in LL (if all are correct use a COLL). Then it will be only EP5 or O + EP5 for edges = 7-12 turns something to solve the last step. EDIT: I just found you can go for 1+3 unoriented if you initially got 0+4. Here is the Antisune: r U2 r' U' R U' R' ... Man! no extra turns =)
 
Last edited:

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
Note that learning this method is a good exuse for not learning ELL, I used to do ELL (and still does for some cases) but after learning this method I started to do most ELL's 2-step (orient and loose the solved FD edge, then permute it back, and the rest of course) That becuse the recognition time is wery short for this and it's not for the ELL's. It is not that turn consuming either, the 2-step E5 uses just a few more turns than 1 step ELL.

OK, this makes almost 20 algs to learn and it's 29 in ELL totaly but that includes 4 EPLL's so it does not differ many, but the ELL's are like 11 turns on average, these are less than 7!
 
Last edited:

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
Still haven't found them but I have another place I'll check later.
 

cubeRemi

Member
Joined
Jul 7, 2008
Messages
73
Location
Holland
I only see how to solve the last 5 edges, will the >5 edges part come ? I'm even more interested in that part!

remi
 

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
I found the sheet, a few of them were the same as yours, or similar at least.

F(H-PLL): MUM2UMU2M' (7)
F(Z-PLL): M2UMU2MU'MU2M' (9)
F(Z-PLL'): M2U'MU2MUMU2M' (9)
H-PLL: M2U2M2UM2U2M2 (8)
Z-PLL: M2U'E2ME2MUM2 (8)
f(d' P): MU2MU'M'U2M' (7)
f(d P): MU2MUM'U2M' (7)
U-PLL: M2U'MU2M'U'M2 (7)
U-PLL': M2UMU2M'UM2 (7)
P U P: M'U2MUM'U2M (7)
P U' P: M'U2MU'M'U2M (7)
P Z-PLL: M2UM2UMU2M' (7)
P Z-PLL': MU2M'U'M2U'M2 (7)
P U-PLL: MU2M'U'M2U'MU2M (9)
P U-PLL': MU2M'UM2UMU2M (9)

EDIT:

I had them drawn out differently from how you named them on the wiki. I used an arrow system.

Here is a picture of the sheet to show what I mean:

http://i140.photobucket.com/albums/r25/Athefre/L5EPartial.jpg

How many moves do you think this way of solving takes?

17 maybe?

1. Insert DB (3)
2. Orient edges (5 at least)
3. Permute edges (.75 for U adjustment + 7.5 for solution, +.75 for final U adjustment = 9)

If your final pair is BR+DBR you can insert the pair and BD at the same time (and make sure when solving the corners that it doesn't mess it up). I didn't see that mentioned in the article.

For the usual way of doing LSE, I average 13 moves.
 
Last edited:

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
Yes, but the move count is not everything, I like the look ahead times, (exept for the times I get 5-cycles permuts, needs more practice) and the oppurtunity to do things (solve BD or if it is solved do the orientation or make it a better case) while solving the corners (I don't know if there is any pice you can solve during CMLL for Roux, it is the orientation but more pices to work with). And I don't do 3 turns for BD, often I solve it during the end of F2B in the same turn as putting a pair, or in the CMLL, (I do FD if it's easier and then y2) add the skips and the average is less than 15 turns.

I was thinking of an arrow system but it made the images so messy :)

Well, thank you for the algs =)
 
Last edited:

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
Yes, but the move count is not everything, I like the look ahead times, (exept for the times I get 5-cycles permuts, needs more practice) and the oppurtunity to do things (solve BD or if it is solved do the orientation or make it a better case) while solving the corners (I don't know if there is any pice you can solve during CMLL for Roux, it is the orientation but more pices to work with). And I don't do 3 turns for BD, often I solve it during the end of F2B in the same turn as putting a pair, or in the CMLL, (I do FD if it's easier and then y2) add the skips and the average is less than 15 turns.

I know move count isn't everything, I was just pointing out that if the recognition isn't faster than in Roux....then you are using more turns to accomplish the same goal.

Kenneth said:
I was thinking of an arrow system but it made the images so messy :)

The way I did it doesn't look messy at all. All the image has to have is the arrow pattern for the upper layer. For example look at the top of the image I posted. See the Square that has the arrow like this:

<----->

Then it has the up and down arrow below it. All you have to show is the <-----> arrow because it's the only case like it. You don't have to show anything else, nothing about the edge that is at FD, no other special patterns. The same goes for every other case.

Another example: Look at the third pattern on that sheet image. See how I have a line coming down from the big square to a little square with an arrow pointing up? That isn't needed. It's the only case that has an arrow pointing north-east on the big square if placed in the lower right corner (the other case like it has the arrow pointing south-west if placed in the lower right corner.

I hope that completely makes sense because I do believe it would be a good way of listing (along with the name) the case because it shows how to recognize it.
 

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
Yes, but you may need to use B&W or greyscale images then, otherwise it will be to much information in the picture and that makes it messy.

I'm not done with recognition but I'm getting closer to have colour pattern recognition, some are easy like P U P because it looks like the H-PLL from two of sides. I will probably improve the page a little later when I have gotten more used to the method. First I will, as I said work on a page for my style to build F2B.

BTW: for me recognition in Roux always starts with orienting the centeres (D to D or D to U), else I have no chance to see orientation =) Then I usally also orient the BD to avoid the 6 edge orientation and I always end it in EPLL because I do not like the RU/LU edges, it is slow for me (lack of practice) and that makes about the same as E15 so... :p
 

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
Yes, but you may need to use B&W or greyscale images then, otherwise it will be to much information in the picture and that makes it messy.

There is no need to show the colors on each edge, all you need is a 3x3 grid with arrows showing the direction each edge on the U layer needs to go.

I drew an example:

http://i140.photobucket.com/albums/r25/Athefre/L5Egridexample-1.jpg

It's just like this page, no colors are shown because they are unnecessary.

Kenneth said:
BTW: for me recognition in Roux always starts with orienting the centeres (D to D or D to U), else I have no chance to see orientation =) Then I usally also orient the BD to avoid the 6 edge orientation and I always end it in EPLL because I do not like the RU/LU edges, it is slow for me (lack of practice) and that makes about the same as E15 so... :p

Inserting the last pair in Roux almost always ends with a R or R' so I do r or r' to orient the centers, or do it while solving the corners, you don't even have to memorize where to insert an r or r' move for each case your mind will do it for you.

Orienting 6 edges is as easy and as fast as any other case actually, I made a post here showing how I do each 6 misoriented edge case.
 
Last edited:

Kenneth

Not Alot
Joined
Aug 10, 2007
Messages
1,693
Location
Kauparve, Hejde, Gotland, Sweden
WCA
2005GUST01
YouTube
Visit Channel
Yes, I'm sure you are right, but I need to practice if I'm switching to LSE and I won't. Because of many reasons, I like my 3x3x3 method to be as much as my big cube method as possible so I can use the same algs, look ahead style and even grips. And as as I wrote from start, you can use E35 for more methods than Roux: F2L without the last edge. Petrus, 2x2x3, Orient using O/O' solve last two pairs but ignore FD edge, COLL EP5 and so on, it is simply more useful than LSE :)
 

cubeRemi

Member
Joined
Jul 7, 2008
Messages
73
Location
Holland
for the >5 edges:

what about this:

-petrus 2x2x3+eo / zz left block
-corner orientation
-corner permutation ( 2 intuitive moves to reduce number of cases )
-edges to their right layer
-permute edges

comments?

Remi
 

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
Yes, I'm sure you are right, but I need to practice if I'm switching to LSE and I won't. Because of many reasons, I like my 3x3x3 method to be as much as my big cube method as possible so I can use the same algs, look ahead style and even grips. And as as I wrote from start, you can use E35 for more methods than Roux: F2L without the last edge. Petrus, 2x2x3, Orient using O/O' solve last two pairs but ignore FD edge, COLL EP5 and so on, it is simply more useful than LSE :)

I'm not telling you to switch, actually I want you to do something of your own :). I was just trying to show you that orienting centers early and orienting 6 edges in LSE isn't complicated.

The part you should have focused on in my last post is the image I drew, that's all I really wanted you to see because I think it is a good way of showing each case instead of having to show a cube detailing specific colors.

cubeRemi said:
for the >5 edges:

what about this:

-petrus 2x2x3+eo / zz left block
-corner orientation
-corner permutation ( 2 intuitive moves to reduce number of cases )
-edges to their right layer
-permute edges

comments?

Maybe. But it's not something I would want to use because if you don't want to use more moves in step 4 and 5 by doing it intuitively you would have to memorize the solution to each case. I'm sure recognition wouldn't be hard in step 5 because the UR edge would always be solved after doing step 4.
 
Last edited:

cubeRemi

Member
Joined
Jul 7, 2008
Messages
73
Location
Holland
cubeRemi said:
for the >5 edges:

what about this:

-petrus 2x2x3+eo / zz left block
-corner orientation
-corner permutation ( 2 intuitive moves to reduce number of cases )
-edges to their right layer
-permute edges

comments?

Maybe. But it's not something I would want to use because if you don't want to use more moves in step 4 and 5 by doing it intuitively you would have to memorize the solution to each case. I'm sure recognition wouldn't be hard in step 5 because the UR edge would always be solved after doing step 4.

for step 4 I think there are 1 + 3x3 + 3x3 = 19 cases but only if you are "right - up neutral". if you are not I think there are 3x3 x4 + 1 = 37 cases thats way less then oll!

for step 5 here are 2x2 + 3x3 = 13 cases but if you do just one setup move there are 2x2 + 2x2 = 8 cases:p

also:
step 2 is 2-gen
step 3 can be 2-gen 50% of the cases( ? )
step 4 is 2-gen
step 5 is 2-gen

this means that this is useful for OH-cubing. but M moves are not at all.

Remi

EDIT:

just like some fridrich users have learned COLL ore ELL you could learn some extra algs like this:

R'U'R'U'RURURU', just the verry short ones
But now I'm thinking ahead :p
 
Last edited:

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
I only said it's not something I would want to use because I would have to memorize more stuff. But for someone else, maybe it would be fast.
 

Athefre

Member
Joined
Jul 25, 2006
Messages
1,248
Kenneth, I made a page showing every case, the way I was having a hard time describing, I used your naming system because I think that's important:

L5EFinal.png


And applet examples here. You sometimes have to reload the page for every applet to show correctly. It's on a page with the method I told you about. This topic helped me quit being lazy and finally do the work to put the idea online :)
 
Last edited:
Top