[Help Thread] The "Square-1 Help / Alg Sharing" thread

[Neilggghhh123]

That's literally less than 1 tps. Get faster at cubeshape.

OK fine ill try

 

[TMarshall]

Does anyone have any alternate algs for the 4 w + u perm ep cases? I can't execute 4's or -4's on the bottom fast. Or, if there aren't any other good algs, could someone make a video showing how to execute those algs fast? Thanks in advance.

 

[Sam N]

Does anyone have any alternate algs for the 4 w + u perm ep cases? I can't execute 4's or -4's on the bottom fast. Or, if there aren't any other good algs, could someone make a video showing how to execute those algs fast? Thanks in advance.

what algs do you use?

 

[TMarshall]

what algs do you use?

Currently I'm not using the algs, I just do an M2 and then deal with the U-adj. I've tried the algs that I've found on Simon crawford's ep list and I can't get them fast at all.

 

[Sam N]

Here is a link to Sarah Strong's website for the EP cases.

http://sarah.cubing.net/square-1/ep

I don't feel like it's mandatory to know all the EP cases since you can turn all of the bad cases into reasonably nice ones. If these happen to be the same algorithms you saw on Simon's website, I apologize.

 

[Lid]

Does anyone have any alternate algs for the 4 w + u perm ep cases? I can't execute 4's or -4's on the bottom fast. Or, if there aren't any other good algs, could someone make a video showing how to execute those algs fast? Thanks in advance.

You can also do H/op + adj/adj for these.

 

[TMarshall]

Here is a link to Sarah Strong's website for the EP cases.

http://sarah.cubing.net/square-1/ep

I don't feel like it's mandatory to know all the EP cases since you can turn all of the bad cases into reasonably nice ones. If these happen to be the same algorithms you saw on Simon's website, I apologize.

Ya i know I don't need to know full ep, but that case always was super annoying to me, so I kinda just wanted to learn algs for it. Either way, Opp-Opp then u-adj isn't too bad, so I think I'll stick to that for now.

 

[Kudz]

In which order shall I learn EPs? Pls don't sort them to two groups.
Should I learn parity CP algs before?
Why no one uses Roux for squan? Is it significantly worse than Vandenberhg method?
How fast should be CS for 15/20/25 sec solver?
Thanks in advance!

 

[Sam N]

In which order shall I learn EPs? Pls don't sort them to two groups.
Should I learn parity CP algs before?
Why no one uses Roux for squan? Is it significantly worse than Vandenberhg method?
How fast should be CS for 15/20/25 sec solver?
Thanks in advance!

For EP, I would suggest learning how to do all the cases that don't involve parity first. This includes thing like aj-aj, op-op, aj-op, cw U, ccw U, H, Z, etc. Most of these ep cases should be intuitive, and you'll figure out what works best for you over time. They should not take too long to figure out. I would also suggest learning parity EP cases. The ones that I feel are the most beneficial are O, W, aj, op, and aj-U cases. Ive learned other ep parity cases along the way, but those will help give you a good foundation.

 

[blade740]

In which order shall I learn EPs? Pls don't sort them to two groups.
Should I learn parity CP algs before?
Why no one uses Roux for squan? Is it significantly worse than Vandenberhg method?
How fast should be CS for 15/20/25 sec solver?
Thanks in advance!

Adj/adj, opp/opp, H, Z, Opp/O, opp/adj, U, U/U, W/adj... Then the rest in whatever order you feel like.

Parity CP is good to learn early. Recognition is the hardest part, so the sooner you start the better off you'll be. Plus it works wonders for your EP lookahead.

Roux is a decent method for intermediate solvers. It requires only a few algs to be fairly quick (say, 20-25). But it lacks the inherent advantages of v'bergh, the forced good cases and lookahead tricks that advanced solvers use. Basically, you'll hit a wall at some point where the only thing you can do to improve your method is to learn increasingly large and difficult alg sets. Whereas with Vandenbergh there are so many cool tricks and intuitive cases to learn that improve your average movecount without having to resort to a table of hundreds of algs.

As for cubeshape... I dunno, sub5 for the worst cases? That seems pretty reasonable but I'm just pulling numbers out of my ass at this point.

 

[Kudz]

Adj/adj, opp/opp, H, Z, Opp/O, opp/adj, U, U/U, W/adj... Then the rest in whatever order you feel like.

Basicly I wanted to know should I do it like. non-parity U perms then non-parity zperms...

But thanks

 

[blade740]

Basicly I wanted to know should I do it like. non-parity U perms then non-parity zperms...

But thanks

Well, I don't really know many parity EP cases. I simply don't do them because of parity CP. But after learning the ones I mentioned, I would just go set by set: you'll already know all the 1-1 swap and 0-2 swap cases, so learn the rest of the 2-2 swap cases (H/H, Z/Z, H/Z, U/Z, U/H). Then the rest of the 3-1 swap cases (O/adj, W/opp). Then the 3-3 swap cases (O/O, W/W, O/W). Then if you really want to learn parity cases, start on those. Learn all of the single swaps, then the 3/0 cases (O, W), 1/2 cases (adj/U, opp/H, etc), then the 3/2 (W/U, etc).

This order is pretty easy to remember and takes you roughly from easiest to hardest. The ones I listed in the previous post are mostly to allow you to use 2-look efficiently for the ones you haven't learned yet.

 

[Kudz]

Well, I don't really know many parity EP cases. I simply don't do them because of parity CP. But after learning the ones I mentioned, I would just go set by set: you'll already know all the 1-1 swap and 0-2 swap cases, so learn the rest of the 2-2 swap cases (H/H, Z/Z, H/Z, U/Z, U/H). Then the rest of the 3-1 swap cases (O/adj, W/opp). Then the 3-3 swap cases (O/O, W/W, O/W). Then if you really want to learn parity cases, start on those. Learn all of the single swaps, then the 3/0 cases (O, W), 1/2 cases (adj/U, opp/H, etc), then the 3/2 (W/U, etc).

This order is pretty easy to remember and takes you roughly from easiest to hardest. The ones I listed in the previous post are mostly to allow you to use 2-look efficiently for the ones you haven't learned yet.

Okey now it makes sens for me.
Thanks!

 

[TMarshall]

Did anyone figure out a fix to the Qiyi Square1 core stripping problem? I tried epoxy but it didn't work too well.

 

[blade740]

I tried super glue, which didn't work for long on one cube but seems to be holding on another.

 

[not_kevin]

I switched out the old cap with a lock nut I bought from Home Depot. Looks like the #6-32 is the right fit for the screw and the passage. I've done maybe ~200 solves since, and it hasn't had an issue yet, but that's not much of a promise given that flipping the core seemed to last for liek 600 last time.

 

[bobthegiraffemonkey]

I lubed the core and tensioned it before doing any solves, and may have flipped the piece the screw goes into in the process. No issues so far, done a couple hundred solves maybe.

 

[Hssandwich]

No problems after 1000+ solves.

 

[blade740]

Reposting some numbers from the facebook group for the sake of posterity:

Weighted for case probability:

For cases without parity:
CP = 4.722 twists
EP = 7.507 twists
Total: 12.229

For cases with parity:
Parity CP = 9.333 twists
Nonparity EP = 7.507 twists
Total: 16.840

Nonparity CP = 4.722 twists
Parity EP = 11.028 twists
Total: 15.750

So full EP saves on average 1.09 twists over parity CP (for parity cases alone - it's only about half a twist savings on average across all solves - 14.535 vs 13.990)

That said, for nonparity cases, I average 10.181 twists rather than 12.229 because of various tricks, bringing my real average down to 13.511 twists even with parity CP. And that actually doesn't count a few forces that I do through parity CP, I just used straight probabilities there.

Once I finish running the numbers for the parity CP cases I'll post my whole spreadsheet but for now I'm going to sleep.

TL;DR: Parity CP costs only 0.5 twists on average, not accounting for special tricks.

 

[EMI]

^ Not taking into account that you save blocks during (non parity) CP almost every time - which probably lowers the average EP twist count - and not taking into account the look-ahead advantage of non-parity CP, especially if you have blocks (Maybe I'm wrong here, but that's how I would intuitively argue)