does anyone know a faster square-1 meathod?

[trying-to-speedcube...]

Come on, Jason's method isn't even the second-fastest method used in competition.

Because noone uses it!

Only one person uses the 2nd-fastest method, what's your point?

because every good square-1 solver uses Vandenbergh's method

rofl

There is no fastest method, use the method that suits you best.

It's like people who play SSBM and say that "there's no best character" even though it has been demonstrated that Mewtwo sucks. Would you say that old pochmann can be executed in sub-15?

You don't really think that I was serious in saying that litteraly every good square-1 solver uses Vandenbergh's method? I meant that it is very widely used, and the world records are set with that method.

WTF is SSBM?

Old Pochmann doesn't have the same move count as freestyle.
Baum method has about the same move count as Vandenbergh's.
Would you say that a 50-move Fridrich solution could be executed faster than a 50-move Petrus solution?

 

[Swordsman Kirby]

You don't really think that I was serious in saying that litteraly every good square-1 solver uses Vandenbergh's method? I meant that it is very widely used, and the world records are set with that method.

You might be serious, but you're also wrong.

WTF is SSBM?

It's a fighting video-game.

Baum method has about the same move count as Vandenbergh's.

I want some raw numbers.

Would you say that a 50-move Fridrich solution could be executed faster than a 50-move Petrus solution?

While I believe Petrus can be as fast as Fridrich, I still think a 50-move Fridrich solution would be faster because: 1) Petrus uses less moves (at least, someone who's good at it), 2) It is harder to turn as fast when using Petrus. Since you eliminated 1), yes.

 

[blade740]

I agree with the request for raw numbers. I can't find any such comparison of movecounts, besides jason's own statement on his page.

 

[trying-to-speedcube...]

Manual calculation gives this:

CO: 2.2
EO: 4.3
CP: 5.1
EP: 8.7
S5: 2.7

TT: 23 twists for Vandenbergh after cubeshape.

FLC: 4
F2E: 4
L2E: 5.4
PLL: 9.3
ML: 1.5

TT: 24.2 twists for Baum after cubeshape.

 

[blade740]

Well, first of all, Vandenbergh's numbers are a bit different in reality from what they should be. With intuitive CO, it's much more likely to skip EO (or get a better case) than just pure alg solving. Same with V'bergh's CP influencing EP.

Second of all, S5 should never take more than 1 twist when movecancelled into the last algorithm.

 

[trying-to-speedcube...]

And you don't mention shortcuts for Baum's method.

About 50 % of the time there is already an edge solved, and if not, I can easily force an edge during CFL.

 

[Swordsman Kirby]

Manual calculation gives this:

CO: 2.2
EO: 4.3
CP: 5.1
EP: 8.7
S5: 2.7

TT: 23 twists for Vandenbergh after cubeshape.

FLC: 4
F2E: 4
L2E: 5.4
PLL: 9.3
ML: 1.5

TT: 24.2 twists for Baum after cubeshape.

With obvious cancellations, ML and S5 are both 0.5.

S5 is really more like 0.25 and ML is really 0.5. If we ignore method-specific shortcuts/mini-skips, our new totals are 20.8 and 23.2.

I have no exact numbers for the officially second-fastest method, but it looks to be ~21.2 twists after cubeshape.

 

[trying-to-speedcube...]

How did you come to 0.25 for step 5?

That's fixing middle layer and/or switching layers.

 

[DavidWoner]

Switching layers adds zero twists, since it always cancels with the last twist.

There are also algs for all EPs that are 1 more twist than optimal that fix middle slice, so fixing middle slice is reduced to .5

I don't know why SK wrote it like that, but 20.8 is the correct number.

 

[joecubes]

hahahahahah there so many people arguing on which method is the best lol.

 

[joey]

hahahahahah there so many people arguing on which method is the best lol.

Awesome post there man.

 

[Dene]

hahahahahah there so many people arguing on which method is the best lol.

Awesome post there man.

Don't be a jerk. I found his post helpful and insightful >_>

 

[Swordsman Kirby]

How did you come to 0.25 for step 5?

Yeah, uh, oops. Vault312 is on top of it. Oh well, I guess the same could be done for PLL but

Spoiler

I only know one of each. (don't misinterpret this and think that I use Jason's method)