I though a thread with a list of known/named methods would be nice.
I use Fridrich F2L, and I've read about ZB and Lars's method.


Fridrich :
http://www.ws.binghamton.edu/fridrich/cube.html

Zborowski-Bruchem (ZB) :
http://www.zborowski.republika.pl/exper ... ethod.html

Lars Petrus :
http://lar5.com/cube/

This was mentioned in the LL Methods thread.

A whole bunch of algorithms from Bernard Helmstetter.

http://www.ai.univ-paris8.fr/~bh/cube/

(more)

Waterman's method(s): http://rubikscube.info
Gaétan Guimond corner technique: http://www.rubikscuberecord.com
My solving approach: http://grrroux.free.fr
And many other optimizations on different websites.

Gilles


PS: Hey it's my very first post on twistypuzzles.com!

Cool, a couple corners methods.

There is a nice movie! Shotaro "Macky" Makisumi has a pop, puts the piece back in and finishe under 20 sec...

http://www.rubikscuberecord.com/tv.html

Lots of stuff, including speedcubing tips.

http://cubefreak.hp.infoseek.co.jp/speedcubing.html

I do a "semi-Fridrich" with COLL/EOLL substituted for OLL.

There's Minh Thai's corners-first solution which I imagine isn't on the web anywhere, but has quite a few variants.

Also a Philip-something has an edges first solution, that supposedly has only a couple of algorithms.

Then of course there are the not-so-fast ones - Rubik's 7-step, Simple Solution by Nourse, Ideal Solution, etc.

When I am in the mood to be REALLY pokey, I do one of the following, just for some relaxing fun:
- Make all sides two-colored first (e.g., the red side has to be all red/orange, but not solidly red), and only then one-colored.
- Solve one side (not necessarily with the adjoining colors correct), then another adjacent side, then the 8 corners, then the cube. This is, in fact, what the original 1980's insert for Rubik's Cube suggests. Takes a mere 150 moves or so.....

for the longest time, i've been telling people that i'm using an "adapted" form of the Fridrich method. i just found out that i've been actually doing the VH system ("Vandenbergh-Harris"). it's very fun.

I've been using a mixed method.

I'm consistently under a minute now (normale 50 sec, lucky 40), but I'm not dedicated enough to go really fast.


I do the F2L:
Cross on the bottom.
Fridrich (missing/lazy on a couple) for 3 edge-pairs.
As I do the last edge-pair, I try to get two of the edges in the LL flipped correctly. Two or four is ok.

With 4 edges flipped: Position the corners with Niklas and Sune. *1

With 2 edges flipped: Position corners and flip the last two edges *2

Orient the corners.
Position the edges.






1* Niklas,Sune and others: http://lar5.com/oldcube/yF.html


2* I'm not sure how to explain them all without diagrams, but these are the sequences I use to flip 2 edges and postition the corners. A couple are from the

Fridrich LL and used for something less precise, a couple are suggestions. I also used a solver, but I don't remember if I use anything from it.
Mostly, it means if I lose the only piece of paper I've written it on, I can find it again.


The corners are either wrong beside each other (b) or on opposite corners (o).
If they are next to each other, the bad corners go at FLU and FRU. If they are opposite, only the edges matter.

The edges are either wrong on near (FR,BR,BL,LF) or far edges (FB,RL).


Some of them have reflections on the left

b RL : RUR' F'U'F RU'R'
b FB : FR'F' U' FRF'R'UR
b FR : [U] FRU R2 U'F'U FRF'
B BR : [U2] RUR' F'UF RU'R'

o FB : R'F'U' FUR
o LB : R'U'F UFR

Lets not forget keyhole methods and L2L4 and other methods by Duncan Dicks.

There is also mine "Fish'n'Chips" method, still in the developing, but I'm getting there.

One of many fine things about my method is that it is possible to solve the last two stages as one using an ZB-alg. So if I know the case (about 15 at the moment) I change to ZB, else it is the normal way.

Same thing goes for the "keyhole" stage, I do like VH, pair up and then solve pair and OLL-edges. But again, if I know the ZB-F2L case (around 25 at the moment), then I go that way

I'm putting up a web-page about my method at the moment and I just uploaded the only page there is that is in English, (the rest is in Swedish but will get an translation soon).

http://members.chello.se/solgrop/kub/Special%20cases.htm

bump

Good brief descriptions:
Speed Cubing Methods
viewtopic.php?t=5063

I am doing Petrus method + flipping all bad edges while assembling 2x2x3 block.It is definately faster then ZB, but it will take so long to get used to .I tried ZB but there are too many patterns left open after extended cross or 2x2x3 normal. ehhhhh.

Something I've been developing, similar to ryan heises method except tainted by dirty memorised algorithms.

I call it the L5C method because it uses a set of algorithms to solve the last 5 corners instead of the last layer. It goes:

Step1. 2x2x3 + 2x2 (or cross + 3 c/e pairs)
Step2. Solve the remaining edges
Step3. Orient the remaining Corners
Step4. Permute the remaining Corners

I've made a page to learn the method and algos from at:
http://l5cmethod.about.vg/

I'm still optimising it and working on some branches off it like:
Step1. 2x2x3 + 2x2 (or cross + 3 c/e pairs)
Step2. Solve the remaining edges
Step3. Orient the remaining Corners and permute the dfr corner
Step4. Permute the remaining Corners

Here are some others that I know about.

Guus Razoux Schultz:

http://rokumentai.akimoto3.com/methode.htm

Adam Cheyer / Jeff Varasano:

http://www.ai.sri.com/~cheyer/rubiks/rubiks.html

And mine:

http://www.ryanheise.com/cube/method/

There are also some interesting unpublished methods including Anthony Snyder's, Doug Li's, and a corners-first method developed by David Allen and Gene Means.

And then there's Georges Helm's impressive list of 638 solutions which he has collected:

http://webplaza.pt.lu/public/geohelm/myweb/cubbib.htm

You might check out my method:

http://www.speedcubing.com/DavidJSalvia.html

David J

I think what would also be interesting is a list of dates/discoveries of solving techniques.

The 2x2x2 block idea was discovered independently by Lars Petrus, Mirek Goljan and Anthony Snyder in the early 80s.

The cross by Singmaster?

The 4 corner/edge pairs of F2L by who knows (but adopted by Fridrich and many others).

The 2x2x2 block + cross idea by Anthony Snyder in the early 80s, and then rediscovered by Chris Hardwick in 2003 (although Chris does them simultaneously).

The "Fish and Chips" idea by Anthony Snyder in the early 80s, and rediscovered by Kenneth Gustavsson in 2006.

The OLL PLL idea independently by Fridrich and also Hans Dockhorn and Anneke Treep (in 1981).

The ZBLL by Lars Petrus.

The ZBF2L independently by Zbigniew Zborowski and Ron van Bruchem in 2002.

I know I'm leaving out some more recent discoveries (and some early corners first discoveries - e.g. Erno Rubik himself gets credit for some of that, I think) but this is a start.

I've begun constructing a webpage covering the evolution of Rubik's Cube solving techniques over the years:

http://www.ryanheise.com/cube/methods/

Additions/corrections welcome.

Just being my persnickety self as usual, but I always thought that petrus conflated COLL and EPLL and that ZBLL was simulataneous COLL and PLL, that is to say lars does his corner swap first then solves.

Although Lars describes how to solve the last layer in 3 separate steps on his page, that is just the basic beginner approach. This is mentioned on his step 5 page:

http://lar5.com/cube/fas5.html

which says "For the advanced cuber, the steps 5-7 meld together to one single step of dozens or even hundreds of special cases."

In the 80s, Lars developed a set of algorithms by hand which could solve the LL in many special cases using either one algorithm, or by melding two algorithms together (with some overlap) to form fairly efficient solutions (sometimes optimal, or close to optimal). Then in 2000, Bernard Helmstetter developed a computer program which found all of the optimal solutions. His work can be found here:

http://www.ai.univ-paris8.fr/~bh/cube/

under the section "Lars' method", "Step 5, 6, 7 simultaneously".

Then in 2003, Zbigniew announced his new method, which started with a Fridrich style F2L, but orienting the LL edges while solving the last corner/edge pair. To generate his set of solutions, Zbigniew wrote a computer program. The result of his work can be found here:

http://www.zborowski.republika.pl/exper ... step2.html

Then, to solve the LL (what is now known as ZBLL), he used the already existing optimal solutions to Lars' method as found by Bernard Helmstetter. This is mentioned on Zbigniew's page here:

http://www.zborowski.republika.pl/exper ... step3.html

Ron independently developed the same strategy as Zbigniew, and attempted to search for all of the solutions to the last F2L pair (with LL edge orientation) manually using his CubeSolver applet program. In 2002, Ron announced his goal to the speedsolvingrubikscube mailing list asking for assistance to help search for all of the optimal solutions. Ron used to have a page where he published his work, although I can no longer find it.

Thanks. I knew he had step 5+6 in one step and step 6+7 in one step, but I didn't know he had developed ZBLL algos

oh my...

I've been going to his site for a year...

Does he link to this somewhere??
is part of the advanced step being able to find the page? Maybe I'm just super retarded ...
wow ... that's impressive.

why is it 270 instead of 177 does that count include mirrors?

Not yet, because he wants to finish it first. But it looks really nice IMO.



There are really 493 algorithms. Lars doesn't count mirrors so it's 270.

More info about methods, nice summaries.

viewtopic.php?t=5063

Shiii ... I cannot even scroll that page normally. It's too bloated with applets. It would be a good idea to replace the applets with imagecube pictures, and clicking them pops up a small window with an interactive applet. Just like Joel van Noort has done for his site also

-Per

I'm not having any problems with it.


Personally, I think that's a very bad idea. Having to open and close many popups is IMO very frustrating when compared to having all the applets on the same page and just pushing the start and/or stop button(s).

After all, I don't think there are many people learning the "270" method, so there's no real need for Lars to consider how the page will render on slow/bad computers/browsers. He possibly made that page mostly for himself.

Hmm. i tried to open the page in Firefox and it works fine. I guess there's an issue with IE7 when using many applets

-Per

One of the things I find really interesting in this old thread is the L5C method. I've never seen it or heard of it, but it seems like it could be quite fast with practice. I'll have to check it out in a bit more detail.