# The New Method / Substep / Concept Idea Thread

#### Noahaha

##### blindmod
New method

You can also put in one corner while you are doing the edges. That would eliminate a large part of identification of the corners.
Or you can put in 4 corners while solving the first two layer of edges and then finish with OLL and PLL.

#### rj

##### Member
I say that anyone who doesn't want to learn fridrich, but also isn't intuitive should try this.

#### Noahaha

##### blindmod
New method

I say that anyone who doesn't want to learn fridrich, but also isn't intuitive should try this.
This seems very intuitive, much more so than beginner's method.

#### rj

##### Member
Niklas? I am a fridrich guy. What is the alg?

#### JasonK

Niklas? I am a fridrich guy. What is the alg?
Niklas is the alg you posted, URU'L'UR'U'L, and its variations.

#### rj

##### Member
WOW! I know two of the three petrus algs. And what, may I ask, is Allen?

#### JasonK

WOW! I know two of the three petrus algs. And what, may I ask, is Allen?
The Allan alg on Petrus's site is L2 U' B F' L2 B' F U' L2. Just a U-perm.

#### rj

##### Member
An idea for a new 5x5x5 method--I think it's unique. It goes like this: Solve 2x2x2 cubes on all the corners, then solve the edge "inner squares" as Heise would call them, then solve as a 3x3x3. Any more ideas or algorithm suggestions?

#### Godmil

An idea for a new 5x5x5 method--I think it's unique. It goes like this: Solve 2x2x2 cubes on all the corners, then solve the edge "inner squares" as Heise would call them, then solve as a 3x3x3. Any more ideas or algorithm suggestions?
On you go, solve the 2x2x2 corners of a 5x5 and tell us how you get on.

#### elrog

##### Member
I don't think solving as a 3x3 with double moves is such a bad idea. I think it would be easier to turn quickly. It is kind of hard to start out with corners though.

#### Noahaha

##### blindmod
The New Method/Substep/Concept Idea Thread

I don't think solving as a 3x3 with double moves is such a bad idea. I think it would be easier to turn quickly. It is kind of hard to start out with corners though.
This made me lol. It is an absurd thought to change to the most impractical reduction method just so that the 3x3 stage uses a different number of layers. And aren't you "Mr. Movecount" or something? What happened to your argument that what turns you make don't matter as long as it's fewer moves?

#### elrog

##### Member
Trying to solve with a method that is hard to recognize doesn't necessarily mean higher movecount.

I was thinking about a easier way to finish F2L (minus 1 slot) and edge orientation at the same time. It can be done intuitively, but algorithms help to increase reaction time. You can do things intuitively until they basically become algorithms, but this is not always the case if you have a large amount of possible positions.

I talked about doing ZBLS before. You could use it on the 2nd and third slot using 86 of the ZBLS algorithms to orient two top layer edges each time. To reduce the number of cases greatly, you could use ELS on the 2nd to last slot and keyhole in the slots corner. Then you would solve the last slot with whatever other substep you wish.

It would be possible to solve the corner of the 2nd to last slot, then keyhole your ELS alg to make the slot and orient eges, but you may not be able to bring the edge above the corner of the slot into the top layer if it is a bad edge. Solving the corner while preserving edge orientation may take a few more moves for some cases, but I think it is better than trying to deal with the corner being in the way.

I think it would be cool if they had competitions where you won if you had the best average of moves and speed.

This could be calculated something like this: Time/(100 - number of moves) to some power.

You would have to disqualify movecounts over 100 or something. You would need a high power to make using less moves actually worth it, otherwise, it would be like regular speedsolving. Doing FMC could also provide you with ok ratings with a high enough power. You would want to get the power somewhere in a happy medium so that it doesn't turn into either complete speedsolving or FMC, though there may be some people who use both approaches. You could also change the 100 if needed.

Last edited by a moderator:

#### Noahaha

##### blindmod
The New Method/Substep/Concept Idea Thread

Trying to solve with a method that is hard to recognize doesn't necessarily mean higher movecount.
I was referring to how you justified thisethod because of it's "more ergonomic" 3x3 stage.

#### elrog

##### Member
I've been messing around with L5P cases with edges oriented. I don't include L3C in this because thats just solved with commutators. It only would take me 15 algs without mirrors, but it would be extremely useful for the Heise method. I am thinking about learning them because it is only 15. I also realized that for cases of L5P that need 2 edges swapped and flipped, you can use a single setup move to convert them to a case with edges oriented and needing to be swapped. Heres a few examples:

convert to j perm case - Bw' (R U2 R' U' R U2 L' U R' U' L) Bw

convert to v-perm case - Lw' (R' U R' Dw' R' F' R2 U' R' U R' F R F) F z'

convert to random case I know - Rw' (U) (R U2 R D Rw' U2 Rw D' R2) Rw

You may need to do the third one backwards to see what it case it solves.

#### mark49152

##### Super Moderator
Staff member
I think it would be cool if they had competitions where you won if you had the best average of moves and speed.
Much as I enjoy reading about theory and ideas, I'm still confused about what you're trying to achieve. Speed, or fewest moves? A number of times you've expressed a preference for saving moves over using faster algs/techniques, but you're not doing full-on FMC either. So yes I can imagine that kind of competition event would suit you! Perhaps there should be a forum called "elegantsolving.com"

Rather than use a formula, which kind of makes it hard to care about the score, you could use a metronome. For example, solver chooses a bpm and has to solve cube without missing a beat, or maybe with N misses allowed before DNF. Speed of thinking becomes the challenge rather than tps, move count is the deciding factor for a given bpm, and to my mind there's no solve more elegant than one that's done at continuous turn rate with no pauses.

##### Member
My idea for 2x2. Very similar to ortega. Solve a face, but always solve it diagonally, OLL, PBL. PBL will consist of 2 algs.

Basically 2-look EG-2

#### jayefbe

##### Member
The New Method/Substep/Concept Idea Thread

A competition taking both time and number of moves will never work (unless we are talking about FMC which already IS about move count and time). The reason it will never work as you suggest is that what constitutes a "good" solve is totally dependent on how you score it. From one extreme, reducing moves can be worth everything, to the other extreme, reducing time is worth everything. In which case, those are two separate events that already exist. Trying to force an event that is in between will only lead to arguments over which aspect should be rated higher over the other, and thus who is actually best will always be up for debate. There is no objective way to rank solves using two wholly different aspects of solving a cube.

#### elrog

##### Member
@ mark: Hmm... I hadn't thought of that. I think it would work well though. Cool idea.

@ Username: Thats not a bad idea to save on alg count, but why always solve the bottom face where it needs a diagnal swap? I can see that it would get rid of the need for AUFing the bottom, but if you solves the face the correct way, you would have a pretty good chance to skip PBL with using the same number of algorithms.

@ jayefbe: Yes, I see your point, but it still would be fun for a non-official type of thing. I really don't expect it to be added to WCA or anything (if I did, I'd make it its own thread).

#### The Supreme One

##### Member
the "big block corners" was how i originally tried to solve my v-cube 5. the idea is great, but there are almost no algs in existence for it, and there is almost zero intuitive portions, it would be all alg. the only advantages are having an epic looking solve, and a very fast 3x3 phase because double layer turns are faster on a v-cube

elrog, as per your question, the idea i was sharing was not to solve the F2L with offset, but rather to move the bottom layer to correspond with pairs that are offset and already formed in the cube, then to finish the F2L using two keyholes (or one more offset), and the the cube usually then gives you a free fourth pair

#### redbeat0222

##### Member
So I was thinking for a CFOP who likes m slices. I think that you could do F2B, COLL/CMLL, fix cross, ELL. It is not a new method and I'm sure A LOT of you have found this out or experimented to find this out but I think it was cool for me to find it out on my own. I believe I've gotten sub 25 without practice on the first try.