Making A New Speedcubing Method

[Michael Wimberly]

So I'm working on a new speedcubing method. It's based on a combination of CFOP and corners-first methods. Here's a rundown of the method:

1. Put all edge pieces with white/yellow tiles around the white center. This will act as a temporary cross, so that F2L pairs are easier to form.
2. Solve the F2L pairs, keeping the "cross" solved.
3. Do COLL or OLL/PLL for just the corners of the last layer.
4. Solve the cross intuitively, using only M and U moves (similarly to roux).
5. Orient the edges of the last layer, using regular OLL.
6. This will leave you with only a U, Z or H perm.
You can also use whatever 1LLL method you want for steps 5 and 6.

This is all I have as of right now. The alg count is relatively low, ergonomics are decent, and I don’t have a move count in yet. I will reply to this thread for updates and answering your questions.

 

[efattah]

If you have a pseudo-cross on the bottom with edges that are mixed & disoriented, how you can fix that with M-U moves? The DR and DL edges would have to be solve and that would take S moves....

 

[Michael Wimberly]

If you have a pseudo-cross on the bottom with edges that are mixed & disoriented, how you can fix that with M-U moves? The DR and DL edges would have to be solve and that would take S moves....

It fixes itself when you make the cross, I’ll prob make a video explaining this whenever I can.

 

[THE_V01DL0RD]

Im kinda stuck

I think you need D moves

also cant you do EO while doing the cross?

 

[efattah]

I'm still confused how you can fix DR and DL using only M-U moves.

 

[Michael Wimberly]

Im kinda stuck
View attachment 43559
I think you need D moves

also cant you do EO while doing the cross?

I have been using cube rotations, but you can do D moves also. Also I guess you can do EO, I didn’t think of that.

 

[Michael Wimberly]

I’ll make a video with a walkthrough solve hopefully this week, just to clear up some confusion.

 

[THE_V01DL0RD]

got a 21 with charlie holmes cross(M2 U2 M2)

 

[Michael Wimberly]

I just took my first ao5 with this method, and got 45.09. To put that into perspective, I usually average around 28-30 with CFOP. I still need to work on this method a little though, I hope it will get better.

 

[OreKehStrah]

That’s a terrible method for speedsolving

 

[Michael Wimberly]

UPDATE: I’m calling this method PFCL. This stands for pseudo-cross, F2L, COLL, and last edges.

 

[pTr]

I am spending some time to think about this.

 

[L4NC3_D4_AS14N]

So I'm working on a new speedcubing method. It's based on a combination of CFOP and corners-first methods. Here's a rundown of the method:

1. Put all edge pieces with white/yellow tiles around the white center. This will act as a temporary cross, so that F2L pairs are easier to form.
2. Solve the F2L pairs, keeping the "cross" solved.
3. Do COLL or OLL/PLL for just the corners of the last layer.
4. Solve the cross intuitively, using only M and U moves (similarly to roux).
5. Orient the edges of the last layer, using regular OLL.
6. This will leave you with only a U, Z or H perm.
You can also use whatever 1LLL method you want for steps 5 and 6.

This is all I have as of right now. The alg count is relatively low, ergonomics are decent, and I don’t have a move count in yet. I will reply to this thread for updates and answering your questions.

not optimal or efficient, but very fun!

 

[pTr]

Ya, I agree. It's rather fun and kinda a brain teaser.

 

[Hazel]

Yeah, the issue with this method is that it is usually faster to solve a subset of the cube correctly on the first pass than it is to solve it partially and correct it later. You sacrifice only a little bit by making the cross correctly to start with, and you gain back tons of time by skipping the D-layer fixing step later.

 

[Michael Wimberly]

not optimal or efficient, but very fun!

That’s really what I was doing it for, just as something fun to do when I’m bored. I really like solving puzzles intuitively, so making my own method was really just for fun more than anything else. I do intend to continue to try to improve it.

 

[Michael Wimberly]

Here’s the minimum algorithm count.
COLL: 7 OLL, 3 PLL
Final OLL: 3
Final PLL: 4
This comes out to a total of 15 algs. This could be good for beginners who don’t want to memorize a lot of algorithms.
I’m not counting F2L because that can be solved intuitively.
I guess you can count the bottom cross formation (the one after f2l and coll) as it has only 2 cases for each cross piece that are each 3 moves long, but it’s so simple that I’m counting it as intuitive.

 

[Michael Wimberly]

Here’s the minimum algorithm count.
COLL: 7 OLL, 3 PLL
Final OLL: 3
Final PLL: 4
This comes out to a total of 15 algs. This could be good for beginners who don’t want to memorize a lot of algorithms.
I’m not counting F2L because that can be solved intuitively.
I guess you can count the bottom cross formation (the one after f2l and coll) as it has only 2 cases for each cross piece that are each 3 moves long, but it’s so simple that I’m counting it as intuitive.

Also forgot to mention you can learn full COLL if you want, but that’s a lot of algs.

 

[Deleted member 36445]

or you could just use petrus

 

[Nitrophenol]

I think it would be more efficient to use EO right after the temporary cross, then you can get a cross case on the last layer and solve it using Pseudo-OLL, as I like to call it, you just assume that the white and yellow pieces are the same colour. Now you can just use M and S moves to orient and solve the last 8 edges in their correct place, and we can invent algs for that