The "Square-1 Help / Alg Sharing" thread

[Neo63]

Hey everyone!

Did anyone use this site to get fast?

Yay Canadian
sorry, back on topic. Yeah I started off with Lars' site, partially because there weren't that many resources when I started in 2008, and partially because I wasn't very good at finding websites (I don't think I found speedsolving then...)
Lars is the one that came up with the square-1 speedsolving method that almost everyone uses nowadays, but just like Jessica Fridrich, his algs are outdated. It's still a good site to get the concept down, but I suggest looking at threads that people posted above, particularly MTGjumper and pkFeng's EP algs which I found to be very useful.

HELP!
How I get sub 20? (consistently)
Urgh.

I like the style of the movie :P
I was going to comment on how you take too long to recognize CP, then I realized you use parity CP (I think...) anyway I still think you can make that step a bit faster. I subconsciously recognize parity when I do CP but I use normal vandenbergh method. A good way I think would be to count how many "blocks" of corner+edge are solved before doing CP, and that way you can easily figure out whether you have parity on each layer or not.

[Cheese11]

Yay Canadian
sorry, back on topic. Yeah I started off with Lars' site, partially because there weren't that many resources when I started in 2008, and partially because I wasn't very good at finding websites (I don't think I found speedsolving then...)
Lars is the one that came up with the square-1 speedsolving method that almost everyone uses nowadays, but just like Jessica Fridrich, his algs are outdated. It's still a good site to get the concept down, but I suggest looking at threads that people posted above, particularly MTGjumper and pkFeng's EP algs which I found to be very useful.

Heh heh, Canadians are awesome.

I remember back in 2007 the resources I could find on cubing were Dan Browns tut and Rob(lot's-o-numbers) tut. I didn't even know what speedsolving was until a friend introduced me to Roux. Back on topic; I just learnt square-1 and I just learnt the necessary algs off of the site I posted and I'm starting to learn some more of them. Now I'm averaging about 1:30 and it's pretty much all I'm practicing.

[Czery]

I like the style of the movie :P
I was going to comment on how you take too long to recognize CP, then I realized you use parity CP (I think...) anyway I still think you can make that step a bit faster. I subconsciously recognize parity when I do CP but I use normal vandenbergh method. A good way I think would be to count how many "blocks" of corner+edge are solved before doing CP, and that way you can easily figure out whether you have parity on each layer or not.

Hmm... Now that you mentioned it, I actually do pause a lot during CP. I actually have no method of identifying parity on separate layers.
Using blocks seems like a good idea if there are blocks at all. The Q, E, X cases aren't blocked at all and recognition is nasty on those. I think Andrew wrote something about it somewhere...

[vcuber13]

look at an edge corner edge, if there are 2 or 4colours its parity, if three its good

[rubixwiz031]

Heres the parity I use:
/, (-3, -3,) /, (3, 0), /, (-3, -3), /, (2, 0), /, (-4, 2), / (4, 2), /, (1, 0), / (-3, -3), /.
That will do 2 things. It will solve parity on top, but make a H perm on the bottom.
To solve the H perm on the bottom ONLY, here's my alg: /, (-3, 3), /, (-3, 3), / (0, 1,) /, (3, -3), /, (3, -3), / (1, 0)

[blade740]

The Q, E, X cases aren't blocked at all and recognition is nasty on those. I think Andrew wrote something about it somewhere...

3-Color rule. Look at one corner (any corner), plus the edges on either side of it. If there are 3 colors total, it's an E, so no parity. If there are 2 colors or 4 colors, it's X or Q, respectively.

[sa11297]

It really depends on how much you practice. Using the method posted there, and no more, it's certainly feasible to average sub20 without learning anything else.

As for what to improve: EO is an easy choice, since there are only a few algs to learn in order to do it 1-look. Cubeshape is the sort of thing that you learn gradually with lots of solves, sorta like cross on 3x3, or maybe intuitive F2L. And EP... well, just learn a few EPs at a time until you know them all.

how many EP cases do you know?

[blade740]

how many EP cases do you know?

All of the nonparity ones, plus O-perms and adjacent swap.

[Neo63]

Hmm... Now that you mentioned it, I actually do pause a lot during CP. I actually have no method of identifying parity on separate layers.
Using blocks seems like a good idea if there are blocks at all. The Q, E, X cases aren't blocked at all and recognition is nasty on those. I think Andrew wrote something about it somewhere...

I think the block method gets most of the cases down, and after a while of practicing you'll do it subconsciously. The rest Andrew covered pretty well.

[mrpotatoman14]

Out of curiosity, does anyone know how fast people have gotten with the screw method?