Older cubers discussions

[pglewis]

Great I will look although I’ve been scared stiff to take it apart haha.

For spring noise I determine which face(s), take the center cap off, remove the screw completely without disassembling if possible (magnetic cubes are good here), track down the spring since it probably leaped when I wasn't prepared, and put one drop of silicone lube on spring's contact points (the screw-head and the washer in the center).

I take a good look at roughly how far the screw is in beforehand so I can return it to roughly the same tension afterwards.

 

[SpartanSailor]

No I did mean back face moves. For example, take HJ speffz.

Plain M2: (L' B L B' M2 B L' B' L) (U R U' M2 U R' U')
Advanced M2: (B' M2 (U R U' M2 U R' U') B)

Oh! Sorry. I just tried that... totally works! (Of course). But do you have to learn that for every letter pair?

The process I was describing is I believe essentially the same, but I don’t know it well. It does the same thing by solving two pieces at once... but I’d have to see if it provides the same solution...

Edit to add: Ahhhhhh.... the thing I was thinking about has to use an M-slice target as part of the pair. You’re doing something more advanced. But I like it!

 

[Old Tom]

For spring noise I determine which face(s), take the center cap off, remove the screw completely without disassembling if possible (magnetic cubes are good here), track down the spring since it probably leaped when I wasn't prepared, and put one drop of silicone lube on spring's contact points (the screw-head and the washer in the center).

I take a good look at roughly how far the screw is in beforehand so I can return it to roughly the same tension afterwards.

Now I'm scared worse, but will wait til I get to the comp and have someone show me. My GAN Magnetic is not as smooth as it was.

 

[mark49152]

But do you have to learn that for every letter pair?

No, it's just a useful tactic for saving moves. Basically it is a set up move to a commutator (3-cycle) involving the A sticker. Most stickers could be set up to A but some work better than others.

 

[openseas]

Oh! Sorry. I just tried that... totally works! (Of course). But do you have to learn that for every letter pair?

The process I was describing is I believe essentially the same, but I don’t know it well. It does the same thing by solving two pieces at once... but I’d have to see if it provides the same solution...

Edit to add: Ahhhhhh.... the thing I was thinking about has to use an M-slice target as part of the pair. You’re doing something more advanced. But I like it!

@SpartanSailor / If you still have the document I gave to you a while ago, it shows all those combinations.
The advantage of advanced M2 is pretty "intuitive" - not meant for much of memorization of algs.

If you have UB-FR pair, UB alone is M2, FR alone is U R U' M2 U R' U'. If you execute two consecutively (not adding or canceling anything inbetween), M2 U R U' M2 U R' U' = [M2, U R U'] commutator. It is basically interchange with M2, inserting with U R U'. So, when you solve "a pair", you can setup any piece to UB then do [M2, setup for the 2nd pair] or [ setup for the 2nd pair, M2].

RB or LB is a great example since you can setup to UB by B or B'. Same goes for UR or UD, simply U' or U sets up to UF location. DB can be setup to UB by B2 as well.

 

[openseas]

BTW, for those of you who want to learn 3-style, this is a very nice tutorial made by Tim Goh (one of Texas BLD gansters ).
His channel has very good tutorials of 3 style, 4 / 5 BLD, and MBLD room concept.

 

[pglewis]

No I did mean back face moves. For example, take HJ speffz.

Plain M2: (L' B L B' M2 B L' B' L) (U R U' M2 U R' U')
Advanced M2: (B' M2 (U R U' M2 U R' U') B)

I'm wondering why stop there... this seems to work for a lot of pairs and is blowing my mind. I haven't sat down and thought deeply enough about what the restrictions are.

 

[mark49152]

I'm wondering why stop there... this seems to work for a lot of pairs and is blowing my mind. I haven't sat down and thought deeply enough about what the restrictions are.

No, don't stop there! The end goal is 3style, and this is just one example of how M2 is a good method to use as a stepping stone to 3style.

Technically what is happening here is a setup to a commutator with M2 interchange. There are lots of other examples. Taking it one step further, you can set up to an M'/M interchange with the FU sticker too. The trick discussed in this thread for efficiently solving the FU and BD targets is an example of this.

Once you get used to setting up to a variety of commutators that interchange on the M slice, it's relatively easy to learn to interchange on the E and S slices instead, and even on the outer layers.

 

[SpartanSailor]

@SpartanSailor / If you still have the document I gave to you a while ago, it shows all those combinations.
The advantage of advanced M2 is pretty "intuitive" - not meant for much of memorization of algs.

If you have UB-FR pair, UB alone is M2, FR alone is U R U' M2 U R' U'. If you execute two consecutively (not adding or canceling anything inbetween), M2 U R U' M2 U R' U' = [M2, U R U'] commutator. It is basically interchange with M2, inserting with U R U'. So, when you solve "a pair", you can setup any piece to UB then do [M2, setup for the 2nd pair] or [ setup for the 2nd pair, M2].

RB or LB is a great example since you can setup to UB by B or B'. Same goes for UR or UD, simply U' or U sets up to UF location. DB can be setup to UB by B2 as well.

I do still have it. Relooking at it makes more sense. As I begin to better understand what I’m doing, I can better process new ways to do it.

I’m excited to learn some of these setups, but with my next comp on Saturday I’ll stick with what I’m doing with M2

 

[Logiqx]

Reached my goal of sub 2 minutes just now and I am very happy with that (it’s a single solve time, my mean average is a full minute more than that).

Do you think I am justified in answering the question “how long does it take you to solve” with “about a couple of minutes”

“Under a minute” is my next goal. I feel chopping off between 60 and 90 seconds will take intuitive F2L (as suggested already by a few of you - thanks) and some training on my finger tricks. And time.

... and so it begins. I started with a goal of 3 minutes, then 2 minutes, then 1:30, then 1:00, then 0:45, then 0:30 and the cycle never seems to end!

You're best placed to answer how long it takes you to solve. After a fair a while, you'll know what is normal, good and bad for you.

You'll always see variation in your times and my rule of thumb is that "normal" solves fall within your median +/- 15% regardless of your speed. e.g. 20 seconds average means that 17 to 23 feels "normal" and just due to how the solve flows or cases encountered. If you're averaging about 2:00 then 1:40 to 2:20 will probably feel normal.

When I'm asked how long it takes me, I usually answer with my mean time (or higher) but if asked "what's the fastest you've ever done it?", I'll tell them my best single and say that it usual takes my mean time.

 

[auienrst]

After a few weeks of inactivity (I've been doing a few solves casually watching the TV to train CN), I did a quick ao5 before a meeting, and was pleasantly surprised! (44.11-40.05-36.16-31.76-47.33)

I used "unlimited" inspection, to find the first pair in x2/y (I'm using Roux). And I nearly broke my old ao5 PB (40.11 vs 39.17) and single (31.76 vs 29.92) even though my "real" pb are with 15s inspection and on a fixed position.
I didn't find a good pair for the last solve, but all resolutions were without skip. I think it's time to practice again

 

[Logiqx]

I'll try to get (this year) a sub-20 Ao5 in OH on film, it won't be easy. This is a 23.595 average, I think I shared a video with a 26.483 Ao5, but that was 11 months ago.

Let's go... OH race to sub-20 Ao5 on film. Here's my first video!

Times: (21.22) 21.61 21.63 21.81 (22.48) = 21.68 Ao5

Cube: Valk Power M

 

[muchacho]

That looked easy, I will need 6 months to get a sub-20 Ao5, and I fell you can get it any time now.

 

[Logiqx]

That looked easy, I will need 6 months to get a sub-20 Ao5, and I fell you can get it any time now.

I'm still a fair way from feeling I can get a sub-20 Ao5. I think I need to be over a second faster globally. Catching one on camera will be harder though!

Don't forget that you're officially using the best method for OH as it requires less TPS. That's gotta be a good thing for an oldie!

Do you know full CMLL for OH?

 

[muchacho]

It may be the right method but it's the wrong person, my move count is embarrasing for a Roux solver.

I know 15 CMLL algs (and I'm still missing 5 for 2H), some are not the best for OH. It's what I'm working on right now.

 

[mark49152]

Just ordered a Giiker Smart Cube. Looks like that could be a lot of fun .

 

[40yearstosolve]

So intuitive F2L. I found this guy and I really like his style of dumbing it down to basics. This is how I got to where I am today, boiling things down to their lowest common denominator- maybe we all do this?

I can’t fit any more algos in my aging skull so perhaps the ‘intuitive’ part of intuitive F2L is supposed to be learn this way.

But this guy is the best I have found.

Two questions...

1. It feels so awkward and slow beyond words - how the heck do I get fast at this?? Is it purely repetition?

2. Is there a better video anyone can recommend. I know pglewis kindly shared a video but for some reason the guy (I forget his name, but he seems quite well known in YourTube cubing circles).

I am determined to keep at it ready for my summer holiday where instead of reading a book this year I shall become known as ‘that dorky bloke sat over there playing with his stupid puzzle day in day out’ - I hope I am known as that, I’d get a kick out of people not getting it haha.

 

[JanW]

@40yearstosolve did you watch the video I posted? Pglewis suggested JPerm's F2L video. You can always watch more videos to get new views on the subject. Maybe some explanations work better for you than others.

You will get very fast with it if you keep practicing. In the beginning it seems like a lot to digest, and you will keep making mistakes. Just go slowly at first and try to see how the pieces move on the cube. Once it clicks, you will notice how those F2L pairs starts falling into place with ease.

 

[mark49152]

@40yearstosolve : There's endless debate about algorithmic versus intuitive F2L, but in reality it's not black and white and the end goal is the same. There's nothing wrong with figuring out an intuitive solution then forgetting how it works and executing it from muscle memory for the rest of your life. And there's nothing wrong with finding a cool alg then studying how it works and applying its elements intuitively to your solutions. Ultimately though, you need to be knowledgeable and flexible so that you can choose the best pairs and solutions in each case and solve to different slots without wasting moves or rotations.

For me, getting good at F2L was a mixture of approaches, depending on what I was enjoying most. It just takes time. Keep at it!

 

[Logiqx]

@40yearstosolve

My take on the word intuitive is that once you've grasped the basics you can figure out decent solutions for yourself. I remember some cases being hard to figure out but with a bit of trial and error you eventually stumble upon a good solution and through repetition it eventually becomes automatic.

I suggest you put the timer away when learning F2L as it means nothing until you are comfortable with the new approach. It will more than likely double your times to begin with (or worse) but it really is the way forward. It reduces the number of mini-steps in your solve and is far more enjoyable than the robotic beginner method.

I'd thoroughly recommend Feliks' website for beginners through to advanced cubers. Check out the F2L section:

https://www.cubeskills.com/tutorials/f2l