Human Algorithm Methods

[TheNextFeliks]

There are two methods known as Human Thistlewaite and Human Kociemba. Would these ever be possible to use for cubing? I know about Ryan Heise's HTA tutorial but do you know any more tutorials for both of these?

 

[5BLD]

No they would not, they are called human for a reason

 

[Noahaha]

No they would not, they are called human for a reason

I thought they were called human because they were the "human" versions of computer algorithms.

 

[5BLD]

Sure, hence they are inefficient human representations of how a computer algorithm works. Show me a way for it to be nearly or as efficient as the computer alg.

 

[nqwe]

But they could be used. In fact, it isn't that hard to get sub1 with HTA.

 

[5BLD]

But they could be used. In fact, it isn't that hard to get sub1 with HTA.

That's not 'reasonably fast'. I just look at the movecount and think, it's not worth it after weighing up the disadvantages of the non-ergonomicity of using such a method

 

[Noahaha]

Sure, hence they are inefficient human representations of how a computer algorithm works. Show me a way for it to be nearly or as efficient as the computer alg.

I fully understand that those are terrible for speedsolving. I just thought you made it sound like people couldn't use them.

 

[TheNextFeliks]

I have just been looking into new methods for 3x3. I was just wondering about these. I guess they are pointless.

 

[mDiPalma]

I guess they are pointless.

but they are fun

try ZZ/HTA:
1) eoline
2) solve BL and FL edges into BL and FL spots (can be swapped)
3) solve BR and FR edges into BR and FR spots (can be swapped)
4) orient the corners (U and D colors facing up/down)
5) solve the remaining F2L blocks using only R2, L2, U while permuting corners (if you can)
6) epll (or just pll)

 

[Cubenovice]

For thistlethwaite info and example solves go here:
http://www.speedsolving.com/forum/showthread.php?24355-Human-Thistlethwaite-Help-Discussion

5BLD:
It is not that hard to be (nearly) as efficient as the computer alg.

 

[vagabond]

but they are fun

try ZZ/HTA:
1) eoline
2) solve BL and FL edges into BL and FL spots (can be swapped)
3) solve BR and FR edges into BR and FR spots (can be swapped)
4) orient the corners (U and D colors facing up/down)
5) solve the remaining F2L blocks using only R2, L2, U while permuting corners (if you can)
6) epll (or just pll)

Hi Matt! This is a very nice analysis and this method really has some potentials in FMC! What's the average move count that you get with this method? What's been your best and least move count with this method?

 

[42isgoatmethod]

That's not 'reasonably fast'. I just look at the movecount and think, it's not worth it after weighing up the disadvantages of the non-ergonomicity of using such a method

The move count for human thistlewaite is 45.7 moves. This is very good for a method. and i feel if the ergonomics were improved it could be good