New 4x4 Idea

[PatrickJameson]

So here is the basic layout of the method.

1. Solve two opposite centers
2. Pair 4 edges(with the two solved centers on the R and L) and put them in the DR DL BR and BL positions
3. Solve the third center (using multislice r and l, and r2, l2(not multislice), U, U2 F, F2, R, L) Then do M' and continue until all centers are done.
4. Finish the remaining edges.
5. 3x3

I have gotten 1:05 so far with it. I think this can be improved. [Edit: New pb with this is 59.84]

Another alternate method is to do 6 edges instead of 4. I'm not sure which is better yet though.

 

[Dene]

Ah. This sounds like a better way of doing what I do, where I get 8 edges, then finish the 4 centers. Much better idea, good job Mr. Jameson!

 

[PatrickJameson]

Why, thank you!

I also just got 59.84 with it.

I am finding that doing 4 edges is a little bit faster than 6 for me. I'm guessing it's personal preference for 4 or 6. The reason I like 4 better is because you have less restrictions when doing the last centers.

 

[alexc]

Sounds like a good idea, but what advantages would this method have over a normal reduction?

 

[PatrickJameson]

It's faster to pair those four edges than if you did it normally.

I don't know if it's going to be faster than reduction but its still an idea that may lead to other methods or something.

 

[fanwuq]

six edges is basically stadler if you mean solving them not pairing random edges. I'm actually solving six edge with D corners (block of 1x3x4). But i don't know CLL, so I end up pairing the rest of centers and edges and finish with Fridrich LL instead of the whole 3x3 solve or (last step of roux as intended).
Now I'm considering your way. Is it faster to pair random edges or solve specific ones?

 

[PatrickJameson]

six edges is basically stadler if you mean solving them not pairing random edges. I'm actually solving six edge with D corners (block of 1x3x4). But i don't know CLL, so I end up pairing the rest of centers and edges and finish with Fridrich LL instead of the whole 3x3 solve or (last step of roux as intended).
Now I'm considering your way. Is it faster to pair random edges or solve specific ones?

Solving specific edges takes longer to look for the pieces. When your just looking for random edges you can pair much quicker IMO

 

[fanwuq]

I was pairing random edges before and it wastes a lot of time to look for pieces. I'm going to see if my look ahead is better for solving specific pieces, it think it will be better. I might switch back once I get faster at this or adapt more algs for 4x4 solving.

 

[PatrickJameson]

I guess its personal preference for whether you look for specific edges or random ones.

 

[Dene]

I shouldn't think it would matter, because you still have to solve the rest of the edges afterwards... But this way definitely gets in a few really quick edges without worrying about the centers. I want to see what Erik can do with it.

 

[PatrickJameson]

I shouldn't think it would matter, because you still have to solve the rest of the edges afterwards... But this way definitely gets in a few really quick edges without worrying about the centers. I want to see what Erik can do with it.

Yea, that's true. I'm only averaging like 1:07 with it but my times are really inconsistent.

 

[Dene]

I'm going to revive this temporarily and suggest another idea that I just had.

How about getting two opposite centres, then pairing up the remaining centre pieces but not completing them. After this you can get 8 edges. Then you can finish the centres using only U2 moves. Then you can finish the edges.

Come to think of it this would be restrictive like normal reduction... but a bit quicker maybe.