Sortega!

Lucas Garron

Lucas Garron

Premium Member
Joined
Jul 6, 2007
Messages
3,757
Location
California
WCA ID
2006GARR01
YouTube
Visit Channel
In case you missed the party, there's a new 2x2x2 method on the block*, thanks to qq.

qqwref said:
2x2 method idea:
0) 2x1 bar of the same color on left half of D face.
1) Orient remaining 6 pieces without messing up the bar.
2) Finish sorting by placing the other two pieces of the D color.
3) PBL.

Step 1 can easily be done in inspection, and steps 1+2 should be reasonable. Step 1 can often be done efficiently 2gen, and step 2 always can. I kinda like the idea of pure 2gen up to PBL.
Click to expand...
(NOTE: I changed the word "block" to "bar" un qq's description to match terminology conventions.)


Here are some algs.

Talking about it with qq, we came across a lot of interesting properties new to this method. I'm definitely planning to learn it and play around with it.
There are also a lot of opportunities for improvement. My guess is it'll be incorporated into fast solvers' toolsets more clearly.

By the way, my favorite property is how every 2-gen move gets you closer to solved while still being a sortega case. I call this "cascading" and it doesn't show up very much in cubing: Square-1 shapes, and as qq pointed out, some versions of Roux L6E.

*It's not actually on the block. That's supposed to be a terrible pun.

EDIT (2010): This method is rather new and experimental. It should be pretty clear to you how it works if you're familiar with 2x2x2 speedsolving, but nevertheless, here are 5 random sample solves: 1 2 3 4 5

EDIT (2023): Embedded links for the sample solves from last edit.
 Twizzle link 
 Title Sortega Example Solve #1
 Setup 
F2 U' R U R2 F' U' R
 Moves 
Rv' // Bar
R2 U' R U2 R' U2 // Orientation
R' U R2 // Separation
U' R U' L U2 R' U L' // PBL
 Twizzle link 
 Title Sortega Example Solve #2
 Setup 
F U' R F' R' U2 R U2
 Moves 
z y' // Bar
R U' // Orientation
R U R2 U' R2 // Separation
U' // PBL

 Twizzle link 
 Title Sortega Example Solve #3
 Setup 
F2 U' F2 R F' U' F2 R2 F'
 Moves 
// Scramble already has a bar at DL
U R' U // Orientation
R U' R2 U' R2 // Separation
x L' U R' U' R L U2 R' U' R U2 //PBL

 Twizzle link 
 Title Sortega Example Solve #4
 Setup 
U' F2 R2 F U R2 F U R2 U'
 Moves 
z' // Bar
U' R // Orientation
U2 R2 U' R2 // Separation
U2 R2 U' R2 U2 y R2 U' R2 U // PBL

 Twizzle link 
 Title Sortega Example Solve #5
 Setup 
U R2 F' R2 U' F R2 F R
 Moves 
y' // Bar
R U' R2 U R' U2 // Orientation
R' U2 R2 // Separation
U L U' R U2 L' U L U2 R2 // PBL
 
Last edited:
So in fact it's just Guimond with some more cases you can start from (since guimond requires this pattern of 3). The feature of 2-gen ness if you also have the same colour of the first block is nice. Only 3 possible orientation cases = yummy. Might learn it for fun one day when I have spare time (will be like.. never :( )
 
As I said in the other thread, I was surprised at how nicely the orientation step worked out when restricted to 2gen only. I had a feeling some cases would end up kinda long, so Lucas's short algs are a pleasant surprise.

Erik said:
So in fact it's just Guimond with some more cases you can start from (since guimond requires this pattern of 3).
Click to expand...
Well, one benefit over Guimond is in the sorting step. Since there are only 2 pieces to trace, you can easily look ahead to sort during the solve. Thus you can be decently fast without actually seeing the sort step in inspection (in Guimond you would often have to look at D).
 
QQ that's what I meant with 3 possible orientation cases ;-) (probably named it wrong)

EDIT: I think with Sortega it might be a bit hard to predict this orientation/separation case since, mostly in a solve where I use the XLL algs I can see everything up till the XLL case in inspection (no matter if it's Ortega/Guimond/last piece + orient/method with no name)
 
Last edited:
I haven't got a 2X2 yet, though I will be getting one soon. Would you recommend going head first into this method? The order of the algs makes it look like it could be learned in phases (starting with having all the bottom layer oriented then just 3 cubies oriented, before going for just 2. It seems attractive as an advanced method that gets easier to set up the more algs you learn.
 
This method looks fun. I was actually about to just go learn EG1, but I might look into this more. I guess this would definitely be better if I started learning PBLs from various angles. Stachu, I'm really interested in seeing the chart you're going to make!
 
StachuK1992 said:
I'm going to make a cubeshape-like progression chart for this.
Click to expand...
Hmm, I already have picture of the cascading graph. But it might be annoying to make it usable, so go ahead. Do you want a list of optimal algs for each case, full graph data. or something?

fatboyxpc said:
I guess this would definitely be better if I started learning PBLs from various angles.
Click to expand...
Yeah, I've been playing with it all day, and it feels much cooler do do PBL if you always start holding the cube the same way (and much easier to do AUF/ADF variations consistently).


qqwref said:
You don't even have to do them as algs to start. In my test runs I did the orientation part using only intuition and Ortega orientation algs.
Click to expand...
True, but I think the cool part comes from the (previously) unintuitive tricks. Not even the regular OCLLs stay the same.
 
StachuK1992 said:
If you could send the pictures and optimal algs, that'd be appreciated.
email
Click to expand...

Was about to post something, but then my computer did its lovely crashing thing.

Nevertheless, here's a CSV of the 243 cases; you can get customized images from the second column through VisualCube, and you can decide how to lay out the graph.
http://cube.garron.us/sortega/sortega.csv
 
Lucas Garron said:
Yeah, I've been playing with it all day, and it feels much cooler do do PBL if you always start holding the cube the same way (and much easier to do AUF/ADF variations consistently).
Click to expand...

Yeah, I've already looked at doing the PBL that's double bars in back (J perm on U and D) in the front (basically doing the inverse mirror of lefty version, I guess?) and if I were to practice it I could get it really close to as fast as I do the on in the back. So I'd need to learn how to do that same PBL with bars on R and L, and then the J/Y perm on R, L, and B as well. It would probably also help to know a J and Y perm on D so I can avoid a rotation. Ideally I should learn the Guimond separation cases from all angles too, but maybe not if I can get this Sortega method down :P
 
Anyone know the probability of getting an unpermuted bar after a random scramble? (Step 0)

This looks really cool. Guess it's another thing I have to put on my memo list, but BLD COMES FIRST DAMMNIT.
 
You must log in or register to reply here.
Back
Top