Thoughts about the BH method

[CT_Warrior]

I started learning BH recently (very late, I know, haha), so I was wondering how one could optimize BH for speed rather than for move count.

I started working on the first 30 cases, and so far it seems pretty fast.
Which cases would need to optimized and how would they be optimized?

 

[cmhardw]

I started learning BH recently (very late, I know, haha), so I was wondering how one could optimize BH for speed rather than for move count.

I started working on the first 30 cases, and so far it seems pretty fast.
Which cases would need to optimized and how would they be optimized?

I would say to use the U PLL alg in a lot of the edge cases. The reason we optimize for move count is to be able to use the entire method on bigger cubes. Technically the page that talks about 3x3x3 corners and 3x3x3 edges that Daniel and I made up should be titled "corners on the n x n x n cube" and "central edges on any odd n x n x n cube"

As for optimizing for speed, basically if an alg sucks then just use a PLL that is faster. I also recommend to use all of the "Slice-Plane" type edge commutators, as they are crazy fast on 3x3x3 compared to the supercube safe versions.

Sorry that this post will not really be all that much help. Daniel and my focus is really to be able to have a fast and efficient method for big cubes. It just so happens that using BH is possible on the 3x3x3, but we did not intend the method to maximize the speed for 3x3x3.

I would say that the BH corner algs would be hard to beat in terms of speed, unless you are just really crazy fast at your A perms. I think Haiyan might disagree with me on this point but I think that BH corners would already be fairly well optimized for speed already. BH edges would have a long way to go to being optimized for speed, because there are a lot more capabilities on 3x3x3 since the U perm is just so incredibly fast. Seriously though, the "Slice-Plane" commutator type which works on 3x3x3 just can't be beaten in terms of speed: M' U2 M U2. I'm sure a huge number of algs could be speed optimized using Slice-Plane algs compared to even a U perm.

Hope this at least helps a tiny bit,
Chris

 

[CT_Warrior]

I see, that was pretty helpful, thanks.

I'll definitely learn BH corners since I pretty much agree that it's fast. I've already started to hate R2 after playing with BH for a bit.

I think I'll learn DIADEM for edges since it seems exactly like the speed-optimal method I was looking for.

 

[cmhardw]

I think I'll learn DIADEM for edges since it seems exactly like the speed-optimal method I was looking for.

At the risk of receiving a :fp I do have to ask, what is the DIADEM method for edges? I've searched the forum, google, and the Wiki and can't find anything relating to it.

Chris

 

[CT_Warrior]

It's Deadalnix's Edge Method.
Not sure how he derived the acronym, but oh well.

He has a beta page of his method out for those who wanted to learn or check it out earlier that you can ask for, but it's not quite public yet.

 

[deadalnix]

Hi chris,

effectively, this is the name I gave to my method for edges. This some kind of bastard child between BH and M2. Anyway, BH where not so well known when I invent the first version so I didn't inspire myself of it. It's more something like a parallel progress. But M2 was obviously a big inspiration.

I have discussed this in the M2 tutorial, and a tutorial is under progress. But due to personal issues, I'm very late.

Anyway, if you are interested, a draft english version is done, and it's very close to the final version. I have to correct some details (mainly langage details because english isn't my langage).

Finaly, the method require more moves than BH, but is more finger friendly and require less thinking (or less learning if you want to do it braindead). It's big cube safe, but I present also some shortcut that aren't center safe, mostly based on M'U2MU2 which is really fast.

The method can be adapted to wing I think, but theyre are some special case I have solution for yet. Most of patterns can be used « as this » on wing, but theyre are extra patterns like when you have 2 adjacent wings.

 

[Sakarie]

How to train BH corners?

One question about training BH; I can find every cycle in about ten seconds, but i want it two be 4-5. How do I most effective train? I just started to time doing all the 18 cycles starting with URF, and then RFU and so on, but is that the best way to go?

Only using BH for corners.

 

[deadalnix]

It's what I do when I'm training, and also solving some cube using BH, but looking at the cube.

 

[mazei]

I would scramble, start the timer, and try to solve corners with BH as fast as possible.

 

[MatsBergsten]

One question about training BH; I can find every cycle in about ten seconds, but i want it two be 4-5. How do I most effective train? I just started to time doing all the 18 cycles starting with URF, and then RFU and so on, but is that the best way to go?

Only using BH for corners.

One thing I did was to solve the 2x2 bld a lot, using BH.

 

[deadalnix]

2x2x2 can be easily speedblinded

 

[MatsBergsten]

2x2x2 can be easily speedblinded

Not by me it cannot . But I explicitly said 2x2 BLD using BH.

The 2x2 BLD in itself is not the funniest event, maybe if you can speedblind it is.
In my opinion 4x4 BLD is the best event of all. Not too long, not too short. A
splendid mix of speed and memory and method

 

[mazei]

In my opinion 4x4 BLD is the best event of all. Not too long, not too short. A
splendid mix of speed and memory and method

Not for me. But I do enjoy it in a certain way as well.

 

[Sakarie]

Practised Corner commutators, by doing all of the threecycles (21x18=378).
I did them in sets of 18, by doing URB>UFR>[all of them], and so on. The average was about 2:35, with the highest at 3:03 (BLU), and the lowest at 1:36 (UBL).

How's that compared to you? How low should I be able to come?

 

[Mike Hughey]

Practised Corner commutators, by doing all of the threecycles (21x18=378).
I did them in sets of 18, by doing URB>UFR>[all of them], and so on. The average was about 2:35, with the highest at 3:03 (BLU), and the lowest at 1:36 (UBL).

How's that compared to you? How low should I be able to come?

I can do most of them between 1:30 and 2:00, so it sounds like you're doing pretty well compared to me. I'm terribly slow turning, though.

 

[mazei]

Well I hope I can get sub-2 with BH corners. Still using visual(tapping) for memo. I tried letters but I felt that it was too troublesome. Any ideas for the corner memo system?

 

[cmhardw]

Well I hope I can get sub-2 with BH corners. Still using visual(tapping) for memo. I tried letters but I felt that it was too troublesome. Any ideas for the corner memo system?

If you don't like letters this may be even more troublesome, but single syllable words has been, for me, the easiest to use, and quickest, memory system I've ever tried for corners.

Good luck with your BH corner practice! Corners are one of my favorite parts of the solve, because it goes so quickly with commutators!

Chris

 

[Swordsman Kirby]

single syllable words

I may adapt this method up to 5x5BLD. I don't think it scales that well for multi, etc. Also, I noticed an error in your words, namely "Zool (XU)." It's actually "Zuul."

 

[Mike Hughey]

single syllable words

I may adapt this method up to 5x5BLD. I don't think it scales that well for multi, etc. Also, I noticed an error in your words, namely "Zool (XU)." It's actually "Zuul."

I haven't bothered with learning a new set of single-syllable words for corners, but I use a sort of halfway implementation of this. I use the same image words that I normally use, but I don't use them as images. Instead, I treat them as verbal utterances. The problem is that some of them aren't single-syllable, so it takes longer than Chris, but it's probably not that big of a loss. And some of my words easily shorten to single syllables, and if it's obvious I'll just use those. It's still quite fast and required a lot less work than inventing a whole second set of words. And for multi, I just do the same thing, but I actually use the images as pictures, instead of just words.

 

[Sakarie]

Practised Corner commutators, by doing all of the threecycles (21x18=378).
I did them in sets of 18, by doing URB>UFR>[all of them], and so on. The average was about 2:35, with the highest at 3:03 (BLU), and the lowest at 1:36 (UBL).

How's that compared to you? How low should I be able to come?

I can do most of them between 1:30 and 2:00, so it sounds like you're doing pretty well compared to me. I'm terribly slow turning, though.

You are, according to yourself, turning very slow, and I'm obviusly much slower than you. Still, you say that I'm doing pretty well! That's kind of you!