Advanced Human Thistlethwaite Method

[Pseudoprogrammer]

Hello all. I used to lurk here a bit, and recently got back into cubing. Before I left I was working on the Advanced Human Thistlethwaite Method. An advanced version of that detailed by Ryan Heise on his website.

The basic solution steps are as follows:
1: Orient all edges (Get the cube into a <L R U D F2 B2> group)
2: Put all equator edges in the equator (this in turn solves a yellow/white cross on the top and bottom automatically, which is what we are really doing, it's just easier to think of it in terms of putting the equator pieces in the equator)
3: Solve the equator (in solving, really just combine steps 2 and 3, just be sure to stick to your F2 B2 restriction)
4: Orient all corners with 1 alg at once
5: Permute all corners at once
6. Get all edges on their own color or the color opposite
7. Permute all edges

Here is an example video: http://www.youtube.com/watch?v=bCYYx7yOyWk
Skip to 2:30 if you read all of that above, unless you like my voice or watching my hair bounce ;D

The solve on the video is a typical 40 move solve. The algs look awkward because I use optimal algs instead of the 1 or 2 move longer fingertrick algs.
Here is the same solution with the steps broken into pieces:
http://alg.garron.us/?alg=D-_F_R2_U..._B2_R-_B2_F2_D2_L2_F_L2_D_R-_B2_F_D2_U-_R2_F-

 

[somerandomkidmike]

I would like the algorithms... looks like fun!!!!!!!!!!

 

[rachmaninovian]

yes it is interesting...so you average 40 moves with it? how many algs do you know?

 

[Pseudoprogrammer]

I personally only know about 30 so far, none of which are in muscle memory (you may have seen my lips moving a little as I recited algs in my head during the video). The hardest set of algs is corner orientation, where one orients all corners at once with a single alg (1 of 60 algs). Since this method is basically all algorithms after step 3, the turn count is very consistent. Generally about 39-43 moves every solve, but there are of course absolute worst case scenarios that would go up to about 50 moves (although I have never encountered one, they are something like 1/800). If you listened to the video, I had to do nearly a dozen shoots to get a scramble in which I knew all the algorithms. Anyone interested in learning this method needs to first learn the basic human thistlethwaite method found on ryanheise.com/cube or in my other youtube tutorials.

 

[ostracod]

I am really interested in this variation of HTA. Please post the algorithms, or a link to somewhere we can get them!

 

[MistArts]

I think orienting corners before solving the E slice is better.

 

[Pseudoprogrammer]

I'm converting all of my algorithm tables which are in incomprehensible handwriting in my notebook or a table that would make no sense to the average person in excell, into a user friendly tutorial. Hopefully it'll be done by tomorrow night (or perhaps tonight if i'm lucky), unfortunately I have a few hundred words to add to my history research paper and will be at a nearby college working on that until night. Several people have offered to host the tutorial so it should be online soon.

 

[ostracod]

Thanks so much! If I like AHTA after trying it out for a while, you may have your first convert.

 

[Pseudoprogrammer]

I think orienting corners before solving the E slice is better.

The reason I do it before is the first 3 steps (really only two steps as I described before) can be all planned out in the preinspection, whereas if you did the corners first, you would have to stop and think for a moment to solve the equator after doing an algorithm. That would get rid of the main advantage of this method (no thinking, all automation)

 

[cuBerBruce]

The main idea behind the Thistlethwaite method seems to be to solve the cube in 4 phases, based upon some nested subgroups of the cube group. The subgroups are <U,D,L,R,F2,B2>, <U,D,L2,R2,F2,B2>, and <U2,D2,L2,R2,F2,B2>. In this new method, the four phases of the Thistlethwaite method are not being strictly followed. More precisely, elements of Thistlethwaite's phase IV are being done earlier in the solve.

The basic solution steps of this proposed "Advanced Human Thistlethwaite" method and the corresponding Thistlethwaite phases (in Roman numerals) are as follows:
1: I
2: II
3: IV
4: II
5: III,IV
6. III
7. IV

So this is clearly less "Thistlethwaite-like" than the Heise version.

I also note that this method solves the corners in only two steps. This seems to require a lot of cases if indeed you are using only 1 alg execution on each of these steps, which seems to be what you are claiming. Perhaps you are partially solving these two steps with some intuitive moves and then applying an alg to finish each step. Or possibly you relying on a variety of setup moves.

 

[mcciff2112]

i left a comment on youtube already but i would really like to look more into this.
It looks like a very interesting and efficient way to solve it.

 

[Pseudoprogrammer]

Cuber Bruce: One orients all corners in one algorithm (one of 60), and then "intuitively" (although this is really trivial, someone who doesn't know how to solve a cube at all could do it) puts all the white corners on the white side and yellow corners on the yellow side. Then using one of five algorithms to permute all corners at once.

And the only difference in my method and heise's method is A) the equator solving and B) combined steps

Heise's method:
Orient all edges, make opposite colored crosses on top/bottom

Orient all corners, two at a time

Seperate corners into their respective layers (intuitive)

Permute all corners (usually takes about 3 steps, because you just use the same alg over and over)

Permute edges onto opposite colored sides (usually takes two steps)

Permute entire cube using only double turns (intuitive)

Whereas my method is:
Orient all edges

Solve equator (which in turn forms crosses for you due to step one)

Orient all corners at once

Seperate all corners into their respective layers (intuitive)

Permute all corners in 1 of 5 algs

Make all edges on opposite colors (or their own color) in 1 alg

Since I use an alg for the step above I don't mess up my corners, which allows me to solve this step in 1 alg also: Permute remaining 8 edges in 1 alg

So if you look at them closely they really aren't that different. The major one which causes the differences is the equator solve.

 

[cuBerBruce]

Heise's Human Thistlethwaite method is based upon the Thistlethwaite method. The Thistlethwaite method has four distinct phases. The Heise Human Thistlethwaite method has the same four distinct phases as the Thistlethwaite method. So while Heise's version keeps the four Thistlethwaite phases distinct, this "Advanced Human Thistlethwaite" does some re-ordering of the steps rather than doing each of the four Thistlethwaite phases separately. I agree it isn't a big departure, as phases I, II, and III are still done in that order, but various aspects of phase IV are done earlier rather than after the completion of phase III. It seems to me you are relating your method only to the Heise version rather than the original Thistlethwaite method.

Cuber Bruce: One orients all corners in one algorithm (one of 60), and then "intuitively" (although this is really trivial, someone who doesn't know how to solve a cube at all could do it) puts all the white corners on the white side and yellow corners on the yellow side. Then using one of five algorithms to permute all corners at once.

Using AUF, ADF, and flipping the cube over to reduce the cases, it seems to me there would be 99 cases. Maybe counting mirrors as the same alg reduces it to 60. Either way it's a significant amount of algs, but not a ridiculous amount.

And the only difference in my method and heise's method is A) the equator solving and B) combined steps

Heise's method:
Orient all edges, make opposite colored crosses on top/bottom

These are two distinct steps.
1. Orient all edges. (Thistlethwaite method phase I)
2. Make opposite colored crosses on top/bottom. (Part of Thistlethwaite phase II)

Orient all corners, two at a time

The other part of Thistlethwaite phase II.

Seperate corners into their respective layers (intuitive)

I don't think separating corners into their respective layers is actually required by Heise's method here. I think it is only a suggestion for helping to ensure that the corners end up in a configuration that is solveable using only half-turns.

Permute all corners (usually takes about 3 steps, because you just use the same alg over and over)

The above sentence seems to imply solving the corners. The Heise method merely gets the corners into a configuration that is solveable using only half-turns at this stage. As recognizing when this condition is satisfied is somewhat tricky, this may be a bit of a drawback for speedsolving purposes, but I believe Heise does this to be faithful to the Thistlethwaite phases. This is part of Thistlethwaite phase III.

Permute edges onto opposite colored sides (usually takes two steps)

The other part of Thistlethwaite phase III.

Permute entire cube using only double turns (intuitive)

"Permute entire cube" seems a bit too general to me. The corners are permuted within two orbits, and the edges are permuted within three orbits. This is, of course, Thistlethwaite phase IV. (And Heise "cheats" and doesn't strictly limit himself to half-turns in this phase.)

 

[Pseudoprogrammer]

Yes I count mirrors as the same alg. And yes "Advanced Human Thistlethwaite" refers to an advanced version of Heise's Human Thistlethwaite, not an advanced human version of the Thistlethwaite algorithm.

 

[qqwref]

Yeah, Bruce, this method is a more advanced (in the sense of more speedsolvable) version of Heise's Human Thistlethwaite, so as I see it the goal is more to be speed-friendly than to be like the original Thistlethwaite algorithm.

I've tried a bunch of solves with this method (without the specialized algorithms) and seem to average around 40 seconds, so I wouldn't be surprised if this method can go sub-20 or maybe sub-15. The movecount seems pretty good too.

 

[Am1n-]

I'm very interested in this, are the algs available somewhere?

mvg

 

[miotatsu]

I know this is a pretty big necropost but I think it is time someone puts more information about this method out there.

First step: EO
like EOLine except without the line. good tutorial here http://cube.crider.co.uk/?p=eoline

Second step: equator
here is how I do this step: I find/place a pair of edges that need to be fixed onto the L/R slice (I use F and B for the EO so I am free to turn L and R in the next step). Next I find a U/D color and place it above the pair, and place the opposite color below. now when you place the two U/D edges it also partially solves your equator. usually doing this you either end up with a nearly solved E slice or solved.

Third step: corner orientation
for a very long time now I have been wondering about the algs for this. I was convinced up until yesterday that I would at some point need to generate them myself. However, I stumbled upon this by chance:
http://www.speedsolving.com/forum/s...Long-Comeau-Belt-Method)-Guide-and-Algorithms
The parity+EPOCLL step of this method is exactly what is needed for AHTA.

Fourth step: corner permutation
There are only 5 cases, I generated my own algs for these but it would be very easy to generate others if you do not care for these:
if D is solved do Y or J perms on U, else you will have one of these 3:

Double J: (place the correct corners on L)
F2 U' R2 U D R2 D' F2

Double Y: (no setup needed)
F U' R' B' L2 B R U F'

YJ: (place the correct pair on top left)
F' L U L F D2 F' L' U' L' F

Fifth step: edges
I do this intuitively, you can use things along the lines of M2 U2 M2 U2
(note that it should be easy to generate algs for all cases though, there are not that many)

Final step: edges
again, should be easy to generate algs for it but I use intuitive edge commutators and an occasional H-perm

I hope this information helps anyone else out there interested in AHTA

 

[Cubenovice]

Hi Miotatsu,

Would you mind reposting this in the active Thistlethwaite thread?
http://www.speedsolving.com/forum/showthread.php?24355-Human-Thistlethwaite-Help-Discussion

I think it is good to keep all (A)HTA info in one thread so it is easier to find for new users.

We already discussed LCBM a little bit there.

I would also like to comment on some other stuff in your post but prefer to do it in the other thread.