Potential Square 1 Hybrid Method/Variation?

[joshuachan1995]

I've been pondering over a hybrid method for the past few weeks or so (hopefully not wasting them) and would like to know if its feasible for speedsolving to any extent from the pros. My idea borrows and utilises elements from various LBL Methods (most prominently the Baum method). I attempted to couple and implement a wee bit of the Vanderbergh system in there as well The main advantage of the method IMO is that the amount of algs you need to learn is quite diminutive in comparison to the most widely used systems.

So I will give you a rough gist of the method and if its worth it , I will try my best to continue optimizing and further developing the method

1. Get back to Cube Shape

2. First Layer Corners or FLC as I like to call it + Orientate 2 Adjacent Edges (0 Algs)
-This is pretty much intutive and insertion of two adjacent edges can be influenced by simple (1,0) or (0,-1) adjustments in the respective U and D layers when solving the corners. On average, this step should take 3.5 twists and no more than 5 maximum.

3. Edge Orientation or EO (3 Algs)
- I'm sure most solvers are familiar with this. Just orientate the remaining two edge or one edges. There are only 3 different cases. One that swaps (UR - DR) .Second that has the effect (UF - DR) (UB - DB) and the last one being (UR -DR) (UB - DB). The algs to solve these cases are widely found on the internet and come in many forms with AUF here and there.

4. Simultaneous First Layer Edge Permuation (FLEP) & Corner Permuation Last Layer (CPLL)
-I guess you could say that this is the crown jewel in the method. There are 4 different classifications to this step . [P/S. I think that learning Opposite Edges on FL is redundant, simply do a M2 U2 M2 move to correct it)

a) CPLL Skip/First Layer EP (8 algs)
if you know how to do EP , you can proceed on otherwise learn 8 new algs. Fortunately these algs are not pure , hence the twists count are all below 7 except for the W Perm (eg. If you have adjacent edges on bottom. You can basically do adjacent-adjacent.)

b) FLEP Skip/ CPLL (2 algs)
use regular J and N perms.

c) J CPLL (8 algs)
an adjacent corner swap is needed on LL. I've generate algs for certain cases such as U Perm and H Perm on FL. My main concern here is the move count. I've been using dougbenham's square 1 solver and I'm not sure how to ignore LL Edges functunality. i formally request any experienced programmers to help me generate this algs ... please? So far, the algs that I generated for W perms O perms and adjacent swap have move counts of 8 and 9.

d) N CPLL (8 algs)
a diagonal corner swap is need on LL. Same move count concern as above. Not sure how to use the solver and ignore LL edges.

5.EPLL (9 algs)

Therefore, a grand total of 38 algs is required. I would lament again that my main concern is the move count.Once again, i formally request any experienced programmers to help me generate this algs ... please?

Please be kind on the comments... I'm not usually active on the forums so yea

 

[mazei]

If you check Jason Baum's website, its a similar method.

 

[joshuachan1995]

If you check Jason Baum's website, its a similar method.

Are you positive. Cos Baum method is FLC + Solve 2 adjacent edges using M2 + Finish off FL with 1 Algo and then 1 LLL

 

[mazei]

Ok, I didn't read through it thoroughly the first time. Time for some real comments.

First, what do you mean by this sentence?

[P/S. I think that learning Opposite Edges on FL is redundant, simply do a M2 U2 M2 move to correct it)

Why is there suddenly M on a square one? Also which piece is FL? As far as I know, the edges are UF, UL, UB, UR, DF, DL, DB and DR.

What advantage does it give, doing steps 3, 4 and 5 over doing it the Jason Baum's way which is step 3, complete first layer and then going to PLL? Just wondering because you are doing an extra step here.

The reduction in the number of algs certainly helps.

I'd be interested in the step 4 algs since with Jason Baum's method, you sometimes get the edges already oriented but just have 2 adjacent edges not permuted. This would certainly help a bit guess. It's kinda like COLL on square-1.

I think this method is nice as a transition into Jason's method or Lars' method.

I've been trying to think of something to try and merge step 2 and 3 together. So far nothing I think of is practical.

Certainly an interesting method, would love to see the move count and algs and such.

 

[blade740]

There is no way to ignore pieces on jaap's solver. You'll have to generate an alg for every possible EPLL case and choose the best instead. The solver has an option to generate algs from an input file.

 

[joshuachan1995]

Why is there suddenly M on a square one? Also which piece is FL? As far as I know, the edges are UF, UL, UB, UR, DF, DL, DB and DR.

The sequence (1 , 0)/(0 ,-1) or (0 ,-1)/(1 , 0) is designated as an M2 move
I guess using the abbreviation FL is inappropriate as it might be confused with individual edge pieces . Basically FL translates to the First Layer.

What advantage does it give, doing steps 3, 4 and 5 over doing it the Jason Baum's way which is step 3, complete first layer and then going to PLL? Just wondering because you are doing an extra step here.The reduction in the number of algs certainly helps.

Step 3 of Baum method is splitted up into two sub-steps. First being,solve 2 adajcent edges on First Layer using M2 sequence. Frequently, at least one of the edges are place correctly. This requires a fair bit of intuition

The second sub-step is to solve the last remaining edges on First Layer similar to Step 3 of my variation. The amount of algs however required is 7 and 3 respectively. Sometimes in the Baum method you might run into a scenario where all of your First Layer edges are already orientated, just not in the right places. http://jmbaum.110mb.com/square1pages/step3c.htm
What my variation does is to efficiently deal with this scenario by permuting the last layer corners whilst permuting the first layer edges. As a result , full PLL can be avoided.

Parity PLLs are daunting to learn and recognition is a tad harder as well. Some cases can only be recognisable by looking at 3 faces. As a result, my variation avoids this.

Overall, aside from the number of algs aspect, my variation also has better recognition as it requires less intuition and most cases can be dealt with algs if that makes sense at all

I'd be interested in the step 4 algs since with Jason Baum's method, you sometimes get the edges already oriented but just have 2 adjacent edges not permuted. This would certainly help a bit guess. It's kinda like COLL on square-1.

I think this method is nice as a transition into Jason's method or Lars' method.

I've been trying to think of something to try and merge step 2 and 3 together. So far nothing I think of is practical.

Certainly an interesting method, would love to see the move count and algs and such.

Thanks ...

I will try my best to generate the algs with lowest move counts possible but currently clueless in using the solver. With regards to merging steps 2 and 3, I would think its impratical as well. Merging CO and EO in Lars' method has already been discussed in the Advanced Square 1 Methods/Theory thread. As David Woner puts it CO is too simple a step to bother complicating.

Performing Full Orientation whilst completing the First Layer corners in this context is over-ambitious.

anyway good luck for Malaysia Open , 6 days left

 

[joshuachan1995]

There is no way to ignore pieces on jaap's solver. You'll have to generate an alg for every possible EPLL case and choose the best instead. The solver has an option to generate algs from an input file.

I apologise for double posting but I'm using jaaps sq1 solver modified by doug benham which supposedly is able to ignore pieces. Again, I'm no expert on programming so please forgive me for any misconceptions or staggered assumptions. Btw, would be kind to teach me how to work out the input file...and cases