An amazing expansion for the petrus method based off of ejf2l

[JLarsen]

So here's the process:

1.) Solve any 3 adjacent edges to form a 2x2x2 block while ignoring a missing a corner. Be aware of "bad corners" during this process.
-A bad corner is one that rests in the same layer as the corner that belongs within the 2x2x2 block. If there is white in your 2x2, then the corner sandwiched by the edges must not contain white. The same rule applys if you have yellow in the block.

2.) Solve f2l ignoring the corner

3.) Rotate so that the block with the missing corner is on FDR, and move the missing corner over the slot.

4.) Perform full CLS.

5.) PLL

Credit goes to Fanwuq. What do the people think? As far as I can see it would allow for an incredibly short 2x2 block. Perhaps 3 moves? That gives incredible ease with 2x2x3 expansion.

 

[byu]

This would require the knowledge of a little over 100 algorithms if I'm correct (CLS + PLL). It sounds interesting, I might try it if I ever learn full CLS. I assume EO is done just like normal Petrus F2L.

 

[ErikJ]

yeah I've thought of this before I told anyone about EJF2L. I'm just too lazy to learn full CLS and I get confused seeing 3 corners during step 4.

 

[JLarsen]

Do you think it could make a 3 move 2x2 block though? Those 108 algs more are indeed brutal byu so I don't know if I'll ever do it.

 

[byu]

My first tries at the Edges-Only 2x2 (move count):

4, 6, 5, 4, 5, 5, 6.

So, averaging around 5, I guess. But that's just me, I'm sure better Petrus users can do better.

 

[ErikJ]

three movers should be easy but I'm not really sure. 37.5% of the time you'll have a bad corner which would be more moves to deal with. I would use this for expanding. placing only 2 edges next to a 2x2x2 is super fast.

 

[MistArts]

Interesting... I'll try a few solves...

B' D' B D2 B2 U F' L' U B' L' D R D2 L' U' L' D L2 U' R B2 D' R B

3 Edges: U' R' F R' D (5)
2x2x3 -1C: U' F' U F' R' F' R2 (12)
EO: R' F2 R2 U' R' (18)
F2L -1C: U y' R' U R' U R U2 R2 U' R' U' R U R' (31)
CLS: y' R U' R' U R U' R' U2 R U2' R' (42)
PLL: F' U' F R2 Uw R' U R U' R Uw' R2 U' (55)

F' D' L D' R2 F' U2 L B2 D2 L' F' L' U L B D R' U' B' U2 L' F' D L2

3 Edges: D2 B2 L2 D (4)
2x2x3 -1C: U2 F2 U' F R' (9)
EO: R U' R' (12)
F2L -1C: y' R U2 R' U' R U2 R' U R2 U R' U' R U R' (27)
CLS: y' U2 R U' R' U R U' R' U2 R U2' R' (39)
PLL: U Rw U' Rw D2 L' U L D2 Rw2 (49)

Not bad. I suck using Petrus though.

 

[EmersonHerrmann]

An idea that I believe was thought of many times before, no offense. The 3 edges at the beginning could improve some move count issues.

 

[JLarsen]

Well fanwuq just explained to me what is new about his idea. Being completely color neutral get's rid of bad corners. If you look at what corner you have, and where it belongs, simply put that location in the U layer, and expand to directions common with D.

"If you build 2x2x2 at FDL, and enclosed corner belongs to RDF, it's not a bad corner if you just do a z' and use what is on your L face as your D face. try it."


Here's a link to the video so you can see the discussion in the comments.

 

[Johannes91]

Solving three edges around a corner takes 4.6169 moves on average.
Solving normal Petrus S1 takes 6.0338 moves on average.

Both without color neutrality. Color neutral S1 is about 4.870 moves on average, not sure about just the three edges.

This seems kinda hacky and to complicate things too much for the 1-2 moves it would save in S1.