Older cubers discussions

[pglewis]

I picked up an MGC 5x5 and have enjoyed the solves so much more vs. my old QiYi. I'm still glacially slow (over 5 mins) because I refuse to learn any more algs and I also end up 2-looking OLL and PLL fairly often because a respectable number of my 3x3 LL algs end in failure. That hasn't stopped me from binge-solving since I got it though.

I also got a Square-1 that I'm hoping to figure out on my own, being in no particular hurry. With enough fiddling I can get it back to cubic shape and will continue toying around to see if I can come with some algs from there.

Not much news on 3x3 in a while. I had a good spell a few weeks ago where I threatened my Ao5 (17.57) and just missed by like a few tenths. My top 10 singles are all under :14 now, after a sudden burst of 13s including a few full-step ones. I've had less practice time recently and the new 5x5 has stolen most of that, plus I haven't been timing as much again either. I'm still maintaining a global in the low-low 20s though, mostly the sup-24 mistake solves keeping me over :20.

 

[OtterCuber]

I got my first square-1 these days, making my collection almost WCA complete. 6x6 + 7x7 on the way. Seems my oldish (in Cuber years) brain still can learn

Wow, you've learned a lot of WCA events! I can only manage two or three events at a time, I think.

 

[GodCubing]

If ur old, you should try a low movecount method like roux or petrus

 

[theos]

I also got a Square-1 that I'm hoping to figure out on my own, being in no particular hurry. With enough fiddling I can get it back to cubic shape and will continue toying around to see if I can come with some algs from there.

Good luck with the square-1! After learning 3x3, I was able to eventually figure out my own solutions for all WCA puzzles except square-1. Something about it just broke my brain. I eventually had to look up some algorithms because I wanted to compete in square-1 in a competition, but if you have the time and patience for it, I think you'll find it a nice challenge.

 

[OtterCuber]

Good luck with the square-1! After learning 3x3, I was able to eventually figure out my own solutions for all WCA puzzles except square-1. Something about it just broke my brain. I eventually had to look up some algorithms because I wanted to compete in square-1 in a competition, but if you have the time and patience for it, I think you'll find it a nice challenge.

I see that you are a FMC expert. How did you learn it, sir?

 

[theos]

I see that you are a FMC expert. How did you learn it, sir?

When I started cubing, the first method I learned was the Heise method. It's already a fairly move-efficient method and it includes the concepts of block-building and commutators so I was already well set up to tackle FMC. Then when I wanted to take FMC more seriously I used Sebastiano Tronto's FMC tutorial, often considered the bible of FMC. Now with the advent of domino reduction for FMC, I've been studying this Domino Reduction guide but so far with mixed results.

 

[Bayamo Penguin]

wow, as a 50yo i'm glad i found this thread today

 

[pglewis]

Good luck with the square-1! After learning 3x3, I was able to eventually figure out my own solutions for all WCA puzzles except square-1. Something about it just broke my brain. I eventually had to look up some algorithms because I wanted to compete in square-1 in a competition, but if you have the time and patience for it, I think you'll find it a nice challenge.

Thanks! I definitely empathize with the brain breakage but I'll keep at it unless/until it stops being fun. As a sanity check while avoiding spoilers: my approach so far has been 1) get it back to cubic, 2) orient the top and bottom, 3) permute to solve. Returning to cubic shape still stumps me for a bit on occasion and I've yet to orient from there, even by accident. I can get to a single corner off on both sides, or half and half on both sides (the orientation always ends up such that it denies being fixed up as easily as it seems like it should be). I get so far on step 2 and then seem to settle into an endless loop, so a new approach is needed there. Anyway, let me know if I'm vaguely headed in the right direction, or at least, if there's a serious flaw with the approach I'm taking.

 

[OtterCuber]

When I started cubing, the first method I learned was the Heise method. It's already a fairly move-efficient method and it includes the concepts of block-building and commutators so I was already well set up to tackle FMC. Then when I wanted to take FMC more seriously I used Sebastiano Tronto's FMC tutorial, often considered the bible of FMC. Now with the advent of domino reduction for FMC, I've been studying this Domino Reduction guide but so far with mixed results.

Thank you for sharing your pathway to success, theos.

 

[pglewis]

RE: my probs trying to orient on Square 1: I finally realized I'm wrestling against parity. I need to swap opposite edges on one side and no amount of playing around was getting me there. A quick Googling verified this and, while trying not to spoil too much for myself, I did see that the parity fix-up is a somewhat long alg. That's not something I'm going to work out for myself right away so the next plan of attack is repeatedly scramble out of cubic form and just hope I luck into fixing it up one of these times.

[Edit: bingo, lucked into fixing parity on the first random try. Just permutation left]

 

[qwr]

RE: my probs trying to orient on Square 1: I finally realized I'm wrestling against parity. I need to swap opposite edges on one side and no amount of playing around was getting me there. A quick Googling verified this and, while trying not to spoil too much for myself, I did see that the parity fix-up is a somewhat long alg. That's not something I'm going to work out for myself right away so the next plan of attack is repeatedly scramble out of cubic form and just hope I luck into fixing it up one of these times.

[Edit: bingo, lucked into fixing parity on the first random try. Just permutation left]

I don't remember how to solve square-1 but I don't think there is an easy way out of parity that preserves cubeshape intuitively.
If I study the puzzle more, I will try to come up with an invariant at the start that determines if parity occurs at all, like squanners do with CSP and some 4x4 methods can do.

 

[pglewis]

I don't remember how to solve square-1 but I don't think there is an easy way out of parity that preserves cubeshape intuitively.
If I study the puzzle more, I will try to come up with an invariant at the start that determines if parity occurs at all, like squanners do with CSP and some 4x4 methods can do.

I seem to have managed to mess up parity without leaving cube shape. I had been experimenting with ideas for permuting for a while when I accidentally messed up orientation. "No big deal" I thought, but before I knew it I was back to the case where U and D are half and half but not lined up such that I could orient. I'm pretty sure I didn't leave cube shape, though I'll be damned if I know what I did.

I have at least one home brew alg that I've worked out how it changes the permutation. As it sits on my desk right now I have just two corners swapped and basically a U Perm on one side, though getting it this far hardly qualifies as "speedsolving" by any definition.

 

[Mike Hughey]

When I figured out how to solve Square-1 "on my own", I actually came up with an 80 move algorithm (I think I remember it was 80 moves) that messed up all my progress, but was guaranteed to invert parity. I then just solved it a second time, and it was guaranteed to work. So I did indeed figure out a reliable method for solving Square-1; it was just a really terrible speedsolving method.

How did I find that algorithm? Just silly trial and error. It was mostly repetitive with a weird move or two thrown in. I don't remember at all what it was; as soon as I knew it was reliable, I decided I was done with that task (figuring it out on my own) and immediately learned a more practical parity algorithm.

 

[theos]

I seem to have managed to mess up parity without leaving cube shape. I had been experimenting with ideas for permuting for a while when I accidentally messed up orientation. "No big deal" I thought, but before I knew it I was back to the case where U and D are half and half but not lined up such that I could orient. I'm pretty sure I didn't leave cube shape, though I'll be damned if I know what I did.

I have at least one home brew alg that I've worked out how it changes the permutation. As it sits on my desk right now I have just two corners swapped and basically a U Perm on one side, though getting it this far hardly qualifies as "speedsolving" by any definition.

Just to be clear, when we talk about parity for square-1 we mean permutation parity and your current situation is exactly a parity case. That is when you have a solvable PLL on one layer and an unsolvable PLL on the other (solvable/unsolvable as determined by solving PLL on a regular 3x3). This would be impossible to solve on a 3x3 so that means you must have to leave cube-shape to solve it.

 

[pglewis]

When I figured out how to solve Square-1 "on my own", I actually came up with an 80 move algorithm (I think I remember it was 80 moves) that messed up all my progress, but was guaranteed to invert parity. I then just solved it a second time, and it was guaranteed to work. So I did indeed figure out a reliable method for solving Square-1; it was just a really terrible speedsolving method.

How did I find that algorithm? Just silly trial and error. It was mostly repetitive with a weird move or two thrown in. I don't remember at all what it was; as soon as I knew it was reliable, I decided I was done with that task (figuring it out on my own) and immediately learned a more practical parity algorithm.

That's precisely the approach I intend to take. Relying on random rescrambling and luck to clear up parity or working out some long but reliable fix-up is unlikely to remain inside the "fun zone" for long lol. A proper parity alg will be the first thing I end up learning.

Just to be clear, when we talk about parity for square-1 we mean permutation parity and your current situation is exactly a parity case. That is when you have a solvable PLL on one layer and an unsolvable PLL on the other (solvable/unsolvable as determined by solving PLL on a regular 3x3). This would be impossible to solve on a 3x3 so that means you must have to leave cube-shape to solve it.

Is there a form of OLL parity as well? I've gotten stumped at the orient stage numerous times and assumed it was a parity issue (I end up needing to swap opposite edges on just one side). If not then I still have some things to figure out in the orient stage. So far random rescrambling out of cube shape and back has eventually gotten me to a point when I can orient but it's sounding like I may be mistaking my ignorance about orienting certain cases for parity.

 

[One Wheel]

The "OLL parity" case I presume you're taking about, with a single edge on each face in the wrong layer, is not a parity case.

 

[pglewis]

Latest in the Square-1 saga: I managed to solve all corners, though edges aren't oriented yet. I think this means I've finally lucked out of parity for real, though. I have a home-brew alg that solved CP but I don't have anything for EP. More to work out once I fix EO (which I still struggle with sometimes).

[Edit: nope. Ended up with a UPerm on one side and not a PLL on the other, with two edges solved]

 

[theos]

Latest in the Square-1 saga: I managed to solve all corners, though edges aren't oriented yet. I think this means I've finally lucked out of parity for real, though. I have a home-brew alg that solved CP but I don't have anything for EP. More to work out once I fix EO (which I still struggle with sometimes).

Yeah, I think with EO, you can swap two pairs of edge between layers intuitively, but you'll have to work out an algorithm to swap a single pair of edges between layers.

 

[theos]

I really got into 4BLD last year and got 3 comps lined up with 4BLD in them. I have a real shot at getting sub-10 and maybe a NR. Unfortunately I triple-DNFed in the first of these comps, so hopefully things go better in the next two in March.

Good news... my first solve of today's competition was a 4BLD success, a 7:14 and a NR! Ended up just 2 swapped centers away from my first official 4BLD mean.

 

[pglewis]

I did finally luck my way out of parity, ending up with a GPerm/UPerm pair. I have a couple algs to play with CP and now I'm sitting on HPerm/UPerm... and nothing in my toolbelt for working with EP yet. Feels like a huge victory nonetheless!