F2L before Cross?

[Brilliantsam]

I was just playing around with random block-building on my 3x3x3 cube, thinking about "different" ways of solving it. Now I say different but what I discovered on my own was basically the first two steps of the fridrich method but instead the other way around... Before I got to this point, I spent some time just randomly placing 3x2x1 blocks in different places, using slice moves to match them up when I accidentally placed what was a bottom edge from the top of the cube to the bottom in the UD slot by hiding two f2l pairs and rotating the front face. I seriously hope I am not going crazy and that the way I'm doing this this is not already part of some already-invented, crazy block-building method (not Fridrich) that I haven't researched or that this was ticked off the list of tried and failed ways for speedcubing.

I feel that placing f2l pairs first and then the cross offers the following benefits (and some horrible defects):

Benefits:

1. More general freedom with f2l
2. The ability to look for f2l pairs during inspection rather than the cross to plan this first.
3. Cross pieces are left in more "discoverable" positions
4. Cross pieces can be placed in *moderately* few moves
5. Higher chance of a roux-like formation or even *possibly an x-cross*
6. Possible LL edge control

Defects :

1. F2l first can lead to higher difficulty locating f2l pieces
2. The change of order may lead to a larger move count or awkward moves
3. Cross pieces may be all placed in awkward postions leaving large move count

*= Ideas I am not totally sure about
If anyone has ever tried this, please also post any ideas or changes to this.

P.S: "Or you cannot just use petrus."

 

[Nuceria]

Hmm, it does lead to some awkward f2l pairings and kills most hope of look ahead, for me at least, but it is fun to mess with. I don't think that this is much of an improvement over normal fridrich f2l as it creates more, often hard to find cases for the f2l and increases cross move count at the same time. It does make for some cool cross insertion maneuvers though.

 

[miniGOINGS]

When I build my first block, I make 2 F2L pairs, and then insert the edge and center in a few moves, It usually take about 10-15 moves and 5 seconds.

 

[Brilliantsam]

Yes, I did think it would have more problems than benefits. I will look into cube explorer and experiment with other ways. At the moment I use full fridrich and I was just probing into methods. It is a fun way to go about the first two parts of the CFOP method but it does have huge downsides. If your lookahead was to adapt to this it might come close to a sub 25 at best. This is kind of a mixture between roux and fridrich, getting rid of the cross for a while (giving a sort of roux-slice scenario) and then inserting edges but now I think about it I'm sure this can't be advanced on...

 

[DavidWoner]

The move count is horrific. I am only getting sub-40 on lucky solves, and thats in STM using M-slice pairing. Considering it takes 3-7 moves STM (5-12 STM) to place just one cross piece, compared to about 7 moves for the whole cross using regular fridrich I don't feel like you gain much compared to what you lose.

 

[Brilliantsam]

The move count is horrific. I am only getting sub-40 on lucky solves, and thats in STM using M-slice pairing. Considering it takes 3-7 moves STM (5-12 STM) to place just one cross piece, compared to about 7 moves for the whole cross using regular fridrich I don't feel like you gain much compared to what you lose.

Well I guess it's down to: "Or you can just use Roux." lol

 

[miniGOINGS]

Well I guess it's down to: "Or you can just use Roux." lol

...never...

 

[Brilliantsam]

Well I guess it's down to: "Or you can just use Roux." lol

...never...

Roux is far, far, far better than this disgusting, high-move count, mutated Fridrich mod I've spent a couple more times solving like this and I realise now. XD

 

[miniGOINGS]

Roux is far, far, far better than this disgusting, high-move count, mutated Fridrich mod I've spent a couple more times solving like this and I realise now. XD

...then use pure Fridrich.

 

[PlutoCuber]

well actually that is a good idea but what if you put in f2l slots and edges at the same time using l r moves instead of R L.

 

[Brilliantsam]

...then use pure Fridrich.

I already do, I'm just building on the method in interesting ways to give benefits to certain parts of a fridrich solve. I would have expected my ideas to be rejected (and now I am seeing that rejection too) but I still love my horrendous ideas!

 

[miniGOINGS]

I already do, I'm just building on the method in interesting ways to give benefits to certain parts of a fridrich solve. I would have expected my ideas to be rejected (and now I am seeing that rejection too) but I still love my horrendous ideas!

How about:

1. F2L minus cross
2. LL Corners
3. Edge orientation
4. Edge permution

 

[Brilliantsam]

well actually that is a good idea but what if you put in f2l slots and edges at the same time using l r moves instead of R L.

That sounds interesting, it might take some of the move count off the cross. It's still pretty ineffecient though. It would create a lot of intuitive pressure as well and I'm sure there wouldn't be a reasonable number of cases where algorithms could be developed for.

 

[robertpauljr]

It is interesting that you bring this up today, because just last night when I was working out a solution to the Master Pyramorphinx, I discovered that I was doing the equivalent of the very thing you describe. F2L then cross. I haven't had time yet to experiment further to decide if it is a better way to do the Master Pyramorphinx than getting the cross first, but I had fun seeing that this is yet another approach to solving a normal cube that I hadn't thought of before. I'll have to add it to my list.

 

[Brilliantsam]

It is interesting that you bring this up today, because just last night when I was working out a solution to the Master Pyramorphinx, I discovered that I was doing the equivalent of the very thing you describe. F2L then cross. I haven't had time yet to experiment further to decide if it is a better way to do the Master Pyramorphinx than getting the cross first, but I had fun seeing that this is yet another approach to solving a normal cube that I hadn't thought of before. I'll have to add it to my list.

It's nice to see that this could be applied to other puzzles, I don't own a Master Pyramorphinx or a Pyramorphinx but I can understand how this could be used.

 

[spdcbr]

Ugh, I can't get any faster than sub 40 with "f2l before cross"