The New Method / Substep / Concept Idea Thread

[VenomCubing]

Thanks!
Very helpful.

 

[Sue Doenim]

After rethinking my previous post on a MGLS style WVCP, I have thought that it would be better like this:
1) EOF2L-1
2) Place last F2L edge and solve CP
3) Place last F2L corner and solve CO (2-gen CLS)
There are certainly more algorithms, but I think ergonomics would be better.

 

[Neuro]

After rethinking my previous post on a MGLS style WVCP, I have thought that it would be better like this:
1) EOF2L-1
2) Place last F2L edge and solve CP
3) Place last F2L corner and solve CO (2-gen CLS)
There are certainly more algorithms, but I think ergonomics would be better.

You might as well be doing WVCP. It has nearly the same # of algs and recog is still pretty bad

 

[Shiv3r]

That's fish and chips or snyder LL. It's not too bad anD better in terms of movecount though you can get really horrible cases like 3 twisted corners

3 twisted corners alg is actually pretty fast. mine is sune+Uperm, and I can sub-2 it. even though I don't practice it...

 

[shadowslice e]

3 twisted corners alg is actually pretty fast. mine is sune+Uperm, and I can sub-1 it.

video please.

 

[Shiv3r]

video please.

I probably exaggerated, I could probably sub-1.5 it. I'm saying that for L3C cases, that one isn't that bad.
Kinda like T-perm imo. It's pretty inefficient but really fast.

 

[Anonymous696969]

What if someone could make a new megaminx method? Maybe it could help with look ahead (meaning not having to look over the megaminx so much to find pieces so look ahead is easier, that be awesome).

 

[Shiv3r]

What if someone could make a new megaminx method? Maybe it could help with look ahead (meaning not having to look over the megaminx so much to find pieces so look ahead is easier, that be awesome).

look at Hyu da Hyun S2L order. it's a variant of westlund which imo is better lookahead-wise. there isn't a straight tutorial on it, but this is what I do:
-F2L on white
-Rotate and do another "F2L" with Yellow
-continue with some strange form of westlund and finish with a LL color adjacent to your F2L color.

 

[Anonymous696969]

look at Hyu da Hyun S2L order. it's a variant of westlund which imo is better lookahead-wise. there isn't a straight tutorial on it, but this is what I do:
-F2L on white
-Rotate and do another "F2L" with Yellow
-continue with some strange form of westlund and finish with a LL color adjacent to your F2L color.

Thanks man, i'm going to look into that concept more, pretty different and smart idea. Once again thanks. This will probably help a lot with my megaminx times as I am really into it right now.

 

[Sue Doenim]

You might as well be doing WVCP. It has nearly the same # of algs and recog is still pretty bad

I guess. My method, I believe, would have a slightly lower movecount and slightly better ergonomics. I guess I'll go back to the first version. Here are some other orders in which to do it, though.
E=F2L edge, C=F2L corner, F=F2L pair, O=corner orientation, P=corner permutation (as in reduction to a 2-gen solvable state)
F, O+P: WVCP
E+P, C+O: my 2nd version
E+O, C+P: some weird TSLE/TTLL type thing
C+O, E+P: Seems bad
C+P, E+O: 1st version
O+P, E+C: Looks interesting, kind of like reverse WVCP, but would have 126 algs for first step.

 

[Neuro]

Interesting proposal to finish MCELL in 3 looks

I'll just go through all the steps here (based on my preference):

1: FB on L
2: 2x2x2 in DBR
3: TSLE variant (orients corners, doesn't pay attention to EO) 105 algs
4: Roux VH (insets edge corresponding to corner in FDR while orienting the edges) 15 algs?
5: Last CE (uses TTLL-like recog but very few algs due to 2x1x1 block in CE on bottom) 54 new algs (?) and 21 PLL's

I'll put a solve in later but it looks pretty promising IMO, should be efficient with good recognition while being under 200 algs!

EDIT: Solve. 43 STM but might've been more efficient with different starting block or CE+LL alg

 

[Shiv3r]

Interesting proposal to finish MCELL in 3 looks

I'll just go through all the steps here (based on my preference):

1: FB on L
2: 2x2x2 in DBR
3: TSLE variant (orients corners, doesn't pay attention to EO) 105 algs
4: Roux VH (insets edge corresponding to corner in FDR while orienting the edges) 15 algs?
5: Last CE (uses TTLL-like recog but very few algs due to 2x1x1 block in CE on bottom) 54 new algs (?) and 21 PLL's

I'll put a solve in later but it looks pretty promising IMO, should be efficient with good recognition while being under 200 algs!

what else is under 200 algs? ZZ-b and ZZ-CT.

 

[Neuro]

what else is under 200 algs? ZZ-b and ZZ-CT.

Only CT works well with MCELL? Also it has overall less algs than CT because Last CE has a super low case count.

 

[shadowslice e]

what else is under 200 algs? ZZ-b and ZZ-CT.

They control more though and take more moves to get to the alg excecution point.

 

[Shiv3r]

They control more though and take more moves to get to the alg excecution point.

however more people main them then main M-CELL at this point

 

[shadowslice e]

however more people main them then main M-CELL at this point

yes but at one point more people main cf than cfop.

 

[Shiv3r]

yes but at one point more people main cf than cfop.

true true. I'm not against M-CELL. one thing I didn't like about the original method proposal was the lack of algs proposed
One of the things I have against the method is the structured "solve these 3 pieces , then solve this last pair here", because it has the possibility to give nastier cases than other methods Like CFOP or ZZ or even Roux which give you easier cases more often.
But if someone genner all the TCLL and then the L5E algs, or the algs posted here, I think maybe It will be ok.

@Teoidus How's 2Gr going?

 

[Neuro]

But if someone genner all the TCLL and then the L5E algs, or the algs posted here, I think maybe It will be ok.

@Teoidus How's 2Gr going?

Misoriented L5E would have terrible recog. You'd be MUCH better off orienting and permuting. As for TCLL, I have some algs that I can modify that I made for TCMLL. Oh and my ZZ substep doc is coming along pretty nicely, it just takes awhile. Hopefully I can release my findings by the end of the month.

 

[Teoidus]

@Teoidus How's 2Gr going?

I'm practicing it when I have free time.

As for the MCELL variant: FB+2x2x2 (18), TSLE (10), RouxVH (8), last CE (~13) => ~49 moves

So I could imagine it being alright, but the fancy LSLL trickery may be worse off than a simpler method that achieves similar movecounts.

Personally, I favorite MCELL variant is one I stumbled on (which I considered old version of 2GR, since I didn't know about MCELL at the time): FB+2x2x2 (18), CP+EO (8), sq to complete F2L (10), 2GLL (15) => 51 moves.

While not anything amazing, the CP+EO step was surprisingly efficient and you can get pretty lucky with the psuedo2gen and 2gen steps following it.

 

[Allahjabark]

A sub 10 rouxer on the subreddit posted his own lse method that he used before switching to EOLR or something. I just like the braindead recog on it and I average about 4 secs with lse now (as opposed to 4-a-b-c) coming from 6 seconds. I'll stop rambling and drop the link now so you can try it for yourself.

https://docs.google.com/document/d/19fYocBPuClXnpZiggBm3vJKicztwuj0iXIuZtm0aRak/edit?usp=sharing