Experimental "method". Requesting feedback.

[Edward]

"Ugg, not another one of these"

Anyway, I was just in my room thinking one day and I was like "Hmm... OMG IDEA". So I applied that idea. This was the crap that spewed out onto my cube.

1. 3 cross pieces
2. Rotation-less F2L (lack of one cross piece allows use of M slice and r moves)
3. COLL
4. EO of U layer while inserting Cross edge
5. Edge PLL (H perm, U perm, Z perm)

Note that step 4 and 5 can be combined.

I'd consider it a hybrid with tweaks. I'd like some feedback.

inb4wellrespectedmembersbringingideadown

 

[wing92]

By three cross pieces, i'm assuming you mean all of the pieces but DF right?

The problem I'm seeing with that is that some F2L cases I think would take longer using r and M moves instead of just rotating. I'll wait and see what some people who know what they're talking about say though. Also, would that inserting the last cross edge use algorithms from Roux?

 

[Edward]

By three cross pieces, i'm assuming you mean all of the pieces but DF right?

The problem I'm seeing with that is that some F2L cases I think would take longer using r and M moves instead of just rotating. I'll wait and see what some people who know what they're talking about say though. Also, would that inserting the last cross edge use algorithms from Roux?

It can be DF or DB.

Yes they are from Roux. Thats why I called it sort of a hybrid.

 

[Cyrus C.]

I think it's faster and better for lookahead just to rotate.

 

[jms_gears1]

Edward....
use
yo
brainz

rike srs.

this is just another attempt to combine fridrich and roux....
aside from the idea being stupid itself. If you feel like it can be fast then go for it.

 

[CubesOfTheWorld]

I'm glad that you thought of a hybrid method, but it just doesn't seem very fast.

 

[TheCubeMaster5000]

this is just another attempt to combine fridrich and roux....
aside from the idea being stupid itself. If you feel like it can be fast then go for it.

Just because it doesn't make sense that it would be faster doesn't automatically make it "stupid".

 

[jms_gears1]

this is just another attempt to combine fridrich and roux....
aside from the idea being stupid itself. If you feel like it can be fast then go for it.

Just because it doesn't make sense that it would be faster doesn't automatically make it "stupid".

No but it being another attempt to make a roux/fridrich hybrid (even if hes saying he didnt (which he isnt exactly) he did (i put parenthesis in your parenthesis so you can read while you read)) does.

 

[cmhardw]

I like the idea of the method. I've honestly never tried doing F2L with the DF (or DB) cross edge missing. I imagine it would afford you some block building shortcuts for certain cases.

I think you could consider another LL approach as well. Doug Li already wrote up a list of all cases for inserting the last cross edge and also orienting the LL edges without regard to anything else in the LL. You could then do:

1. 3 cross pieces
2. Rotation-less F2L (lack of one cross piece allows use of M slice and r moves)
3. Insert cross edge + orient all LL edges

And from this point you could continue with a corners only OLL then PLL. You could do COLL then EPLL, or even ZBLL.

Just a thought. This might be a good beginner stepping stone for the more advanced version of combining your steps 4 and 5 into single step (i.e. alg list to solve both steps in one look).

I honestly think the method is a neat idea and seems to have potential. I'll wait until someone with more math and programming knowledge than I do comes along to do a case count for combining steps 4 and 5, as well as someone who can analyze that minus 1 cross edge F2L. I think it is at least worth pursuing a bit further before deciding the merits of how fast or slow it's future potential is.

Chris

 

[trying-to-speedcube...]

3. COLL
4. EO of U layer while inserting Cross edge

No.

 

[StachuK1992]

This actually isn't *that* bad.
Shall try a bit.

 

[Kirjava]

Change steps 3-5 to CLL » L5E plz.

 

[Feinster]

http://www.speedsolving.com/wiki/index.php/3CFCE
lol

 

[Edward]

http://www.speedsolving.com/wiki/index.php/3CFCE
lol

It's so hard to be original these days >.>