ECE - New 3x3 Solving Method

[Shiv3r]

Don't know if anyone will care, but I got my first sub-9 with ECE! I think this method actually has great potential! The solve's lookahead was easy, especially since the first step was three moves. My global average with ECE, specifically EZD is around mid-14 seconds. Hopefully someone else will see this solve as proof of this method's potential.

This solve was done with the EZD variant of this method. I know all WV, CO, CP and EPBL algs (56 algs) for anyone wondering. You can find CO and CP algs in any Square-1 tutorial.

Scramble: L' U B' D2 F' D' F2 R L2 D' F2 R2 D2 F2 L2 U' L2 D2 B2

Solution:

y' U2 D R' // 3 Edges + 3 oriented corners
y' D r U r' d' U2 // Setup for CO
R U' L' U R' U2 L // CO
U' R2 U' R2 U D R2 D' R2 // CP
U D' M' U2 M U D' M' U M // EO + Edge seperation
x M U' D L U2 D2 L' U' D M' x' U' D2 // Edge permutation of both layers

I'm counting this as 42 moves since U D or U D' gets executed as one move, since I execute both moves at the same time.
42 moves / 8.97 = 4.68 turns per second

I'm very pleased with this method's low movecount!

thats actually really awesome!!!
yeah I'm making a youtube video overviewing the method.
this is how I teach it:
1.e-slice edges+corner orientation
2.insert pair+OCLL(7 cases, easy to remember)
3.seperate into layers w the 5 square-1 algs and then permute with T-perm and Y-perm on both sides(or use square-1 algs, I said u could use both)
4.solve d-layer and then use EOLL algs that preserve corners(found on speedsolving.com wiki)
5.EPLL


thats the beginner way to that method, after they master this way they can learn the corner perms then finally full winter variation.
The algorithm count on this method is insane! Even after you learn the full-fledged method, it takes something like 40 algorithms, wheras a full-fledged Fridrich takes about 78 algorithms(and that's only the OLL and PLL)
I believe in this method!
GO CRAFTO22!!!

 

[crafto22]

thats actually really awesome!!!
yeah I'm making a youtube video overviewing the method.
this is how I teach it:
1.e-slice edges+corner orientation
2.insert pair+OCLL(7 cases, easy to remember)
3.seperate into layers w the 5 square-1 algs and then permute with T-perm and Y-perm on both sides(or use square-1 algs, I said u could use both)
4.solve d-layer and then use EOLL algs that preserve corners(found on speedsolving.com wiki)
5.EPLL


thats the beginner way to that method, after they master this way they can learn the corner perms then finally full winter variation.
The algorithm count on this method is insane! Even after you learn the full-fledged method, it takes something like 40 algorithms, wheras a full-fledged Fridrich takes about 78 algorithms(and that's only the OLL and PLL)
I believe in this method!
GO CRAFTO22!!!

Thanks man! I'll keep practising with this method so long as it interests me. Right now I average about 14 seconds, so I definitely think this method has serious potential!

 

[Shiv3r]

I just have them all memorized, so I would have to re-generate all the algs. There are only 16, so I'm sure it wouldn't be too much trouble to find some good algs with Cube Explored.

what are the cases? at the start, it seems like it's just the EPLL cases on both sides, but inserting edges gets you really wierd cases. Can you at least tell me what the cases are? Ill gen the algs myself, but finding all the cases are beyond me.

 

[shadowslice e]

what are the cases? at the start, it seems like it's just the EPLL cases on both sides, but inserting edges gets you really wierd cases. Can you at least tell me what the cases are? Ill gen the algs myself, but finding all the cases are beyond me.

You have all the standard double EPLLs (ie U/U U/Z U/H Z/H) then all the cases where two sides are swapped by something like an M2 U2 M2

 

[crafto22]

what are the cases? at the start, it seems like it's just the EPLL cases on both sides, but inserting edges gets you really wierd cases. Can you at least tell me what the cases are? Ill gen the algs myself, but finding all the cases are beyond me.

The edges are separated into their own respective layers. There are only 16 cases which are just combinations of standard EPLLs.

 

[Shiv3r]

i know, i tried that. but you get wierd cases still, even seperated into their own layers

 

[crafto22]

i know, i tried that. but you get wierd cases still, even seperated into their own layers

Yeah, there are parity cases, M2 U2 M2, then do the alg

 

[Shiv3r]

Yeah, there are parity cases, M2 U2 M2, then do the alg

theres two parity cases: ones M2 U2 M2, the other one is when edges are swapped like in a J or Y-perm(without the corner swap of course)
I think the best course of action for those are maybe just bring them to the U layer and but them back in and do an alg???

 

[shadowslice e]

theres two parity cases: ones M2 U2 M2, the other one is when edges are swapped like in a J or Y-perm(without the corner swap of course)
I think the best course of action for those are maybe just bring them to the U layer and but them back in and do an alg???

No there is only one parity case: M2 U2 M2. It works for all just like 4x4 PLL parity works.

Also, did you not read my previous post on the alg cases?

 

[Shiv3r]

No there is only one parity case: M2 U2 M2. It works for all just like 4x4 PLL parity works.

Also, did you not read my previous post on the alg cases?

yes, I did. I guess I can see how it works thank you for clarifying.
BTW I taught my sister the classic 3x3 LBL method, and as her speedsolving method I'm teaching her ECE. She loves this method, and told me to say thank you Crafto22

 

[crafto22]

yes, I did. I guess I can see how it works thank you for clarifying.
BTW I taught my sister the classic 3x3 LBL method, and as her speedsolving method I'm teaching her ECE. She loves this method, and told me to say thank you Crafto22

Cool! I would love to hear about her progress with the method. I'm confident she can be sub-15 with this method as I am at the moment. This method definetely has great potential and I think with better lookahead and EZD execution somebody could be sub-10!

 

[Shiv3r]

Hey Crafto, I Genned an algorithm set here for EPBL
Turns out if you don't care about which face , U or D, is on top, you only get 10 cases, plus U2 M2 U2 for parity.
Its a really crappy alg set, with a lot of b moves and hard to do movesets. this is only meant as a base for other algsets for this method.

 

[efattah]

During the last couple of months of searching for new methods, I too (independently) came across this method (which I suppose is called ECE, and SSC...) This is now three people that have come up with it, furthering evidence that it has huge potential. However, I did not generate the algs for it, and I am still moving forward with my LMCF method, but I am very eager to monitor the progress of the growing ECE crowd.

I really look forward to the day when the final round at Worlds is comprised of four or more totally different methods instead of just 2...

Eric Fattah
Vancouver, BC

 

[Shiv3r]

During the last couple of months of searching for new methods, I too (independently) came across this method (which I suppose is called ECE, and SSC...) This is now three people that have come up with it, furthering evidence that it has huge potential. However, I did not generate the algs for it, and I am still moving forward with my LMCF method, but I am very eager to monitor the progress of the growing ECE crowd.

I really look forward to the day when the final round at Worlds is comprised of four or more totally different methods instead of just 2...

Eric Fattah
Vancouver, BC

Yes, there is no variation in the cubing world anymore. That's what made me switch to first ZZ for a few days, then now roux from CFOP.
I believe that this method may be the method that wins competitions.
Crafto, can you show us a solve? maybe a video of you speedcubing with it?
I can't get past about 30 secs with any method, so I want to see how someone fast does this method.

Plus if you use MGLS instead of WV, you only need to learn the I cases(7 algs with mirrors), waaay easier to learn than 27 cases. This is because if you insert the last e-slice edge, the corner under it has only 3 possible orientations: oriented, I cases or Im cases. for all 3, there is only about 7 cases(OCLL, CLS I-cases, and CLS Im cases)

 

[shadowslice e]

Plus if you use MGLS instead of WV, you only need to learn the I cases(7 algs with mirrors), waaay easier to learn than 27 cases. This is because if you insert the last e-slice edge, the corner under it has only 3 possible orientations: oriented, I cases or Im cases. for all 3, there is only about 7 cases(OCLL, CLS I-cases, and CLS Im cases)

I would say that you can learn the MGLS if you like but in general the algs for that aren't as nice. Especially not if you use SLS which averages ~5 moves IIRC and the WV is also nicer algs than MGLS though it could be ok in some situations such as on the off-chance that it happens to be solved.

 

[Shiv3r]

I would say that you can learn the MGLS if you like but in general the algs for that aren't as nice. Especially not if you use SLS which averages ~5 moves IIRC and the WV is also nicer algs than MGLS though it could be ok in some situations such as on the off-chance that it happens to be solved.

the thing is, for roux solvers like me who abhor large algsets, MGLS is a lot easier. Its EJF2l, you don't need to worry about the corner, just insert the edge and do the alg. and theres only 14 algs. I think that deserves some credit.

 

[shadowslice e]

the thing is, for roux solvers like me who abhor large algsets, MGLS is a lot easier. Its EJF2l, you don't need to worry about the corner, just insert the edge and do the alg. and theres only 14 algs. I think that deserves some credit.

I'm a roux user. And I agree that smaller alg sets are nicer. However, I also believe Ergonomics+Efficiency>smaller alg sets. Plus SLS is only ~20 algs anyway and a few of them are just insert and do standard OCLL while others are just cancelled OCLL inserts and stuff.

 

[Shiv3r]

I'm a roux user. And I agree that smaller alg sets are nicer. However, I also believe Ergonomics+Efficiency>smaller alg sets. Plus SLS is only ~20 algs anyway and a few of them are just insert and do standard OCLL while others are just cancelled OCLL inserts and stuff.

yes, but if I want to learn an algorithm set I want it to be useful in more than one case.

 

[efattah]

yes, but if I want to learn an algorithm set I want it to be useful in more than one case.

I'm a roux user. And I agree that smaller alg sets are nicer. However, I also believe Ergonomics+Efficiency>smaller alg sets. Plus SLS is only ~20 algs anyway and a few of them are just insert and do standard OCLL while others are just cancelled OCLL inserts and stuff.

If we are talking about the future of speedcubing (as in the fastest/newest methods) I think there is no way around large algorithm sets. During algorithm execution you can move the cube way faster than any system where you need to track pieces and make decisions while the pieces are moving. Roux is probably as fast as we will ever get for a 'low' alg set (which isn't even that low). The method I am currently working on has 200+ algs and I believe most new (good) methods will be at least 50 algs...

Eric Fattah
Vancouver, BC

 

[shadowslice e]

If we are talking about the future of speedcubing (as in the fastest/newest methods) I think there is no way around large algorithm sets. During algorithm execution you can move the cube way faster than any system where you need to track pieces and make decisions while the pieces are moving. Roux is probably as fast as we will ever get for a 'low' alg set (which isn't even that low). The method I am currently working on has 200+ algs and I believe most new (good) methods will be at least 50 algs...

Eric Fattah
Vancouver, BC

I'm not sure I entirely agree with this... You can reach almost as high tps in intuitve without the pause needed to recognise an alg case. There may not be anything wrong with learning large alg sets per se though I'm not sure it offers a significant advantage over less alg heavy methods. It just depends how good the lookahead for that part of the solve is. For example, people can sometimes reach 8+ tps in F2L for CFOP and that is almost as fast as the alg tps.

To be more succinct, larger alg sets may be nice as they solve more of the cube though at the price of recognition being worse. The only time I would say it is good to have a large alg set is at the very end of the solve unless we can easily predict what will happen next (such as CMLL-LSE in Roux).

My thoughts are that methods will move towards a lot of small steps with excellent lookahead where only a few easily identifiable will need to be tracked (I think LSE is likely the best example of this) and the steps will almost blend into each other ( I tried to do this with SSC and will explain how that works if I ever get around to the video but I think F2L and LSE are good examples of this)