Older cubers discussions

[pglewis]

SOOOOOO CLOSE!

On my journey to sub 1 minute 3x3 solves I've been hitting the occasional 0:54 and 0:52 (which was my personal best 2 months ago) but then things train wreck and I get a 1:22 or more. I did get a new PB of 51.46 though.

I'm still spending way too much time on the cross and F2L. I keep working on memorizing PLL cases and my last layer is getting faster and faster but I think I can pick up 20 seconds or more just by getting better at the cross and F2L.

The journey continues.

I'm excited for you, vicariously sharing your victories here.

You aren't too far behind me, much the same story here just with different numbers. The lion's share of my gains are still from improving F2L lookahead. I'm working to take my eyes off the pair I'm solving sooner so I can spot my next pair sooner, cutting down the pauses between. It's much easier to spot pieces when they're moving vs. static, so as unintuitive as it seems I do my best right now when I'm turning a little slower at the end of solving each pair... I can turn fast enough to out-turn my lookahead otherwise. It's the pauses that really kill.

 

[openseas]

I'm excited for you, vicariously sharing your victories here.

You aren't too far behind me, much the same story here just with different numbers. The lion's share of my gains are still from improving F2L lookahead. I'm working to take my eyes off the pair I'm solving sooner so I can spot my next pair sooner, cutting down the pauses between. It's much easier to spot pieces when they're moving vs. static, so as unintuitive as it seems I do my best right now when I'm turning a little slower at the end of solving each pair... I can turn fast enough to out-turn my lookahead otherwise. It's the pauses that really kill.

@NewbieCuber @Shaky Hands @pglewis
So happy to see many cubers with similar stage
I think my progress was, Intuitive F2L cases --> sub 50s, All PLL --> sub 40, a little bit improved F2L --> sub 30. Now, I'm hitting sub 20 like +/- 10% but rarely practicing these days due to BLD or some other stuff (magnetizing..)

 

[Mike Hughey]

For 3x3x3, many years ago I got where I was actually pretty good at lookahead. I resorted to learning full OLL just to get sub-30. And I improved lots of things, but still couldn't get sub-20.

Now it looks like I'm finally about to go sub-20. The key was that I finally realized how important it is that the way you insert your F2L pairs is critically important. You have to do them in a movement-efficient way. I never could be bothered with that, and I didn't fully realize until the past few weeks how much it was holding me back.

 

[mafergut]

For 3x3x3, many years ago I got where I was actually pretty good at lookahead. I resorted to learning full OLL just to get sub-30. And I improved lots of things, but still couldn't get sub-20.

Now it looks like I'm finally about to go sub-20. The key was that I finally realized how important it is that the way you insert your F2L pairs is critically important. You have to do them in a movement-efficient way. I never could be bothered with that, and I didn't fully realize until the past few weeks how much it was holding me back.

I'm starting to think that's one of the main reasons I find it so difficult to improve past the sub-19/ sub-18 barrier. When I find nice F2L cases I can easily get a 15 to 16 second solve, even with a normal, not specially lucky LL, but when F2L cases get nasty I am happy if I can sub-20 that solve at all. Remembering about corners stuck in back slots is part of it too.

 

[pglewis]

For 3x3x3, many years ago I got where I was actually pretty good at lookahead. I resorted to learning full OLL just to get sub-30. And I improved lots of things, but still couldn't get sub-20.

Now it looks like I'm finally about to go sub-20. The key was that I finally realized how important it is that the way you insert your F2L pairs is critically important. You have to do them in a movement-efficient way. I never could be bothered with that, and I didn't fully realize until the past few weeks how much it was holding me back.

@Selkie has recently put the magnifying glass to his solves in order to push forward too. I've yet to hit any stubborn plateaus but I know it'll be coming before long so I closely watch what you experienced folks do to break out. "Three more seconds improvement just from spamming solves" can't happen forever.

I'm still only at about half of OLL, at least when the timer is running. At my current pace I'll probably be near or below :20 before I have them all. PLL was a lot easier for me to absorb due to the frequency I get to see them, I had those mostly down when I was around 1 min and still absorbing F2L. They're so automatic at this point I actually have trouble if I start to think about the execution at all. I literally had to AUF and AUF back in order to do a Gb today, I let my brain engage and couldn't get the muscle memory to kick-in without a small reset lol.

It wouldn't be surprising to find out I still have 3-5 seconds tied-up in lousy cross solving. At my times I probably shouldn't be planning 3 edges and figuring out the 4th on the fly as often as I do. It's just pure laziness, it's probably rare for me to go past 8-10 seconds of inspection while spamming solves so I'm not up against the time. I guess on the plus side there is still low hanging fruit if I start to bottom out on easy F2L gains.

Remembering about corners stuck in back slots is part of it too.

Yes, so very much THIS. I've actually done some practice with sexy/sledgehammer in the back but I still have to know the corner is back there and how it's oriented first.

 

[Lid]

PB? Thought my 3x3 a12 was better than in the signature but couldn't find it, so here is my new "PB" 15.363

14.048, 11.634, 13.657, 15.890, 16.834, 16.623, (19.597), (11.313), 17.909, 15.716, 16.768, 14.546
Both 11s was skips ;-) rest full steps including 2 Nperms.

Spoiler: scambles

Average of 12: 15.363
1. 14.048 U2 L2 B2 D' L2 B2 D' U L2 U' B R2 U' L' F' D' L2 D2 L R F'
2. 11.634 L2 U L2 F2 L2 D B2 D' B2 D2 U' L' D2 F' D U2 L2 B2 D' L F'
3. 13.657 D2 U2 L B2 L' R2 U2 B2 U2 B2 L' U' L2 D2 F2 L D L2 B R' F'
4. 15.890 U' R2 B2 F2 U B2 U2 R2 B2 F2 D2 F D2 R B L' U R' U B2 D'
5. 16.834 L2 U' L2 U' R2 F2 U R2 B2 F2 R' U2 R' B L' R' D2 U' F' D'
6. 16.623 R2 F2 U R2 B2 D' U2 B2 F2 R2 U2 L' F' L2 R' D' L2 U2 B' D' L2
7. (19.597) L2 U B2 D F2 D' L2 D U' L2 F R U F U B2 L U2 B' L2
8. (11.313) L2 U2 B' R2 U2 B' F2 L2 F2 D2 F L F2 U2 F L' D R2 D'
9. 17.909 L2 B2 L2 D' L2 D' F2 D U' F2 U2 F' R2 U R D2 L' D F2 R2 D
10. 15.716 U2 F L2 D2 U2 F' U2 L2 F2 L2 D' F2 L' B' R2 B R F2 L' U R2
11. 16.768 R D' L2 B2 U' B U2 R' B D2 L D2 F2 D2 R U2 L' D2 F2 R
12. 14.546 L2 D F2 U L2 D' R2 B2 L2 R' D2 B' U L R2 D R D F2

Now I hope that I get my new Magnetic 3x3 & 4x4 before the next competition.

 

[Mike Hughey]

One thing I discovered a while ago, but that I hadn't realized pointed to what I'm seeing now, is that it's generally true that if I get a 25 second solve on a particular scramble, and then scramble it again and try it again, even knowing exactly how I solved it before, I'll STILL get a 25 second solve. And I could even repeat it 3 or 4 times and still get the same kind of result - maybe shaving off a second at the most. That should have clued me in that the WHOLE problem (with that particular solve at least, and therefore certainly with a lot of solves) was that I was being entirely too inefficient solving it. And my F2L insertion move counts are good. My whole problem was that the WAY I was executing them was terrible. Tons of cube rotations and tons of awkward moves to get them done.

One way I've been practicing to improve it is to just limit my cube rotations, even to the point of sometimes insisting on zero cube rotations while slow-solve practicing. It's forcing me to learn new ways to do F2L insertions that are much more finger-friendly.

 

[pipkiksass]

Calling all magnetisers - if you fancy magnetising a GANs Air 356 make sure you "tension" the edges and corners correctly.

For anyone who's been cubing for 3+ years, remember tensioning an old Dayan cube, where you stretched the core out and counted the threads to check each centre was tensioned equally? Well the GAN anti-pop system is built in to the pieces, and each has to be screwed to the same tightness to allow the cube to turn evenly. I've had the cube apart 3 times since magnetising it, and although it's now even, I need to tighten it a little (after initially massively overtightening it).

Anyway, it's the most irritating of the cubes I've magnetised. There's definitely more of an art to it than, for example, a Valk or Weilong GTS2, where you can't really go wrong. With the GAN, magnet placement needs to be more precise, tensions need to be even between all 12 edges and 8 corners. Just a little fiddly - took me back to the days of Florian modding Guhongs and, as I say, counting the threads on screws to even it out. The GAN isn't made to be dismantled, it's made for you to use the plug 'n play spring system. It's not HARD, it's just fiddly. I could probably magnetise a WuQue AND a Valk in the time it took me to do the GAN.

Anyway, a verdict - I hated the cube out of the box. It flexed in ways that I hate, similar to the FangShi Shuang Ren, which I thought felt cheap and horrible, although loads of people loved it "back in the day". I swapped the stock nuts for the yellow ones as a test, and it was a BIT better, but still meh. It didn't really keep its shape very well, and caught a lot.

A couple of hours later, it was magnetised but WAY too tight (I'd overtightened both corners and edges). I've since reassembled and disassembled it twice, and I'm about to give it a 3rd go to get it a little tighter.

It's pretty good now, VERY smooth and buttery, but a little loose. I'm hoping that, with a little tweaking, it could be main-worthy. It's definitely an interesting alternative to the Valk and GTS2.

A worthwhile experiment, IMHO.

 

[openseas]

Calling all magnetisers - if you fancy magnetising a GANs Air 356 make sure you "tension" the edges and corners correctly.

For anyone who's been cubing for 3+ years, remember tensioning an old Dayan cube, where you stretched the core out and counted the threads to check each centre was tensioned equally? Well the GAN anti-pop system is built in to the pieces, and each has to be screwed to the same tightness to allow the cube to turn evenly. I've had the cube apart 3 times since magnetising it, and although it's now even, I need to tighten it a little (after initially massively overtightening it).

Anyway, it's the most irritating of the cubes I've magnetised. There's definitely more of an art to it than, for example, a Valk or Weilong GTS2, where you can't really go wrong. With the GAN, magnet placement needs to be more precise, tensions need to be even between all 12 edges and 8 corners. Just a little fiddly - took me back to the days of Florian modding Guhongs and, as I say, counting the threads on screws to even it out. The GAN isn't made to be dismantled, it's made for you to use the plug 'n play spring system. It's not HARD, it's just fiddly. I could probably magnetise a WuQue AND a Valk in the time it took me to do the GAN.

Anyway, a verdict - I hated the cube out of the box. It flexed in ways that I hate, similar to the FangShi Shuang Ren, which I thought felt cheap and horrible, although loads of people loved it "back in the day". I swapped the stock nuts for the yellow ones as a test, and it was a BIT better, but still meh. It didn't really keep its shape very well, and caught a lot.

A couple of hours later, it was magnetised but WAY too tight (I'd overtightened both corners and edges). I've since reassembled and disassembled it twice, and I'm about to give it a 3rd go to get it a little tighter.

It's pretty good now, VERY smooth and buttery, but a little loose. I'm hoping that, with a little tweaking, it could be main-worthy. It's definitely an interesting alternative to the Valk and GTS2.

A worthwhile experiment, IMHO.

I did bunch except Gans. Main reason is that they don't have decent stickerless cube. I'm having trouble solving cubes other than stickerless - 7th color (either white or black background) bothers me so much. So, not much experience or comment to add... but..

When it comes to the core tension, I tried maglev for Valk3. Replaced springs with two donut-shaped magnets. The verdict from many top cubers were negative mostly because I couldn't find a decent option. (N42) However, if one can find reasonable strength & thickness combination, it does have benefits over springs - more tolerance in terms of space compared to springs (force ~ 1/distance^2 for magnets).

Last month (US Nats) - I had a chance to discuss with Chris Tran regarding this maglev core - he also agrees to the benefits but didn't have enough time to explore optimum conditions yet.

 

[newtonbase]

Last month (US Nats) - I had a chance to discuss with Chris Tran regarding this maglev core - he also agrees to the benefits but didn't have enough time to explore optimum conditions yet.

I saw a video he did on maglev cubes and the difficulties in finding magnets that give the right tensions and that actually fit in the cube. It looks fiddly. I'd like to try doing one but need someone else to get it working first.

 

[JohnnyReggae]

Calling all magnetisers - if you fancy magnetising a GANs Air 356 make sure you "tension" the edges and corners correctly.

I've done 3 GAN's so far, and one of them I've done twice changing the magnets. The last one I ended up taking apart about 5 times to try get the corners and edges correctly set. It is definitely a finicky cube to get right. Of the big 3 the GTSv2 is by far the easiest to magnetise.

I've also settled on the Yellow GES nuts, although if you go with stronger magnets I'd recommend using either Green or Blue.

 

[pipkiksass]

Interesting thoughts @JohnnyReggae @openseas @newtonbase...

Personally, I'm not overly keen on the GAN, even having tweaked the corner tensions. I'm totally sold on stickerless these days, too. I struggled a little going back to stickered, and have to admit I'm not overly keen.

That being said, that actual assembly of the pieces is much easier on stickered cubes - because there's only one plastic colour, the edges and corners are each only 2 pieces, so you don't have to worry about building one red/white/blue corner, one orange/yellow/green (in that order), etc., etc.

But that's a small price to pay for the fiddlyness, IMHO.

Maybe I need to tweak the cube even more to get it right, but I've used 4x2 N35s (same as my Valk), and I don't know if it's because the plastic is really thin, or just something I've not set 100% right on the threads of the edges/corners, but the cube just doesn't seem too great. Don't get me wrong, it's still a massive improvement on the cube out-of-the-box, but I've put lubicle 1 on it, done 30-odd solves, squirted a couple of drops of DNM37 in, and it's still really meh.

Re: maglev cores, I seem to remember one of the CubicleLabs videos, Chris talks about the added weight from the maglev core. I'd definitely be interested in trying to make one - I have a spare Valk. I wonder if this is more of a modding job, i.e. dremel may be required to get enough space for the right magnets?

 

[JohnnyReggae]

Interesting thoughts @JohnnyReggae @openseas @newtonbase...

Personally, I'm not overly keen on the GAN, even having tweaked the corner tensions. I'm totally sold on stickerless these days, too. I struggled a little going back to stickered, and have to admit I'm not overly keen.

That being said, that actual assembly of the pieces is much easier on stickered cubes - because there's only one plastic colour, the edges and corners are each only 2 pieces, so you don't have to worry about building one red/white/blue corner, one orange/yellow/green (in that order), etc., etc.

But that's a small price to pay for the fiddlyness, IMHO.

Maybe I need to tweak the cube even more to get it right, but I've used 4x2 N35s (same as my Valk), and I don't know if it's because the plastic is really thin, or just something I've not set 100% right on the threads of the edges/corners, but the cube just doesn't seem too great. Don't get me wrong, it's still a massive improvement on the cube out-of-the-box, but I've put lubicle 1 on it, done 30-odd solves, squirted a couple of drops of DNM37 in, and it's still really meh.

Re: maglev cores, I seem to remember one of the CubicleLabs videos, Chris talks about the added weight from the maglev core. I'd definitely be interested in trying to make one - I have a spare Valk. I wonder if this is more of a modding job, i.e. dremel may be required to get enough space for the right magnets?

I've had a love/hate relationship with the various GANs I've had from the 356sv2 to 3 Airs. Out the box they have all been meh ... Even with magnets they have been meh ... They only really seem to start shining when you've done a couple hundred solves. GAN definitely takes more patience to get the setup right.

Re: magnets. I found the 4x2mm N35's to be a little too strong on the GAN, good on the Valk, and weak on the GTSv2.

I'm also intrigued to try a Maglev conversion. I have a couple spare's that I could try it on ... hhmmmmm

 

[openseas]

Interesting thoughts @JohnnyReggae @openseas @newtonbase...

Personally, I'm not overly keen on the GAN, even having tweaked the corner tensions. I'm totally sold on stickerless these days, too. I struggled a little going back to stickered, and have to admit I'm not overly keen.

That being said, that actual assembly of the pieces is much easier on stickered cubes - because there's only one plastic colour, the edges and corners are each only 2 pieces, so you don't have to worry about building one red/white/blue corner, one orange/yellow/green (in that order), etc., etc.

But that's a small price to pay for the fiddlyness, IMHO.

Maybe I need to tweak the cube even more to get it right, but I've used 4x2 N35s (same as my Valk), and I don't know if it's because the plastic is really thin, or just something I've not set 100% right on the threads of the edges/corners, but the cube just doesn't seem too great. Don't get me wrong, it's still a massive improvement on the cube out-of-the-box, but I've put lubicle 1 on it, done 30-odd solves, squirted a couple of drops of DNM37 in, and it's still really meh.

Re: maglev cores, I seem to remember one of the CubicleLabs videos, Chris talks about the added weight from the maglev core. I'd definitely be interested in trying to make one - I have a spare Valk. I wonder if this is more of a modding job, i.e. dremel may be required to get enough space for the right magnets?

@pipkiksass
Valk3 is the best when it comes to maglev. The reason is that there are enough space in the center core not like GTS V2. I also tried GTS V2 & Maying but I had to grind the center piece (inside) to fit those magnet discs.

BTW, I ran many experiments (still on-going) with magnets varying N35, N38, N42, N50, sizing also varies 3x1.5, 4x1, 4x1.5,4x2, 5x1.5, 6x1. I have many MF3Rs to spare since between my son and my self, we have 40-ish MF3Rs for multi blind . Anyway, so far, I made dozens of them with various conditions, N42 4x2 corner with N38 4x1 edge showed the best performance. Will post details a little bit later.

 

[muchacho]

Sometimes I'm just faster, and I have no clue why o_O

Ao12 PB: 16.080 (old was 16.237 from April)

Spoiler

51956 01-sep-2017 15:22:39 00:16.143 F2 D2 L2 U R2 L2 D F2 L2 D' F' L D B D' B2 R L2 B2 U2
51955 01-sep-2017 15:22:04 00:15.501 R2 D2 B2 R2 D L2 U2 R2 B2 R2 U' R B' R U R F2 D L F' R U'
51954 01-sep-2017 15:21:25 00:15.719 U2 B2 F2 U R2 U2 F2 U R2 U' R' D' U' F' D2 B' D' B2 R
51953 01-sep-2017 15:20:36 00:29.245 D' F2 L2 U2 B2 D' B2 F2 R2 U' F2 R' F2 U' F2 D' B F' L2 U B' D'
51952 01-sep-2017 15:19:53 00:16.046 L2 D U R2 F2 L2 F2 D B2 L2 U2 R D' B2 R2 U F D' F D2 U
51951 01-sep-2017 15:19:16 00:14.983 U F2 U L2 D' F2 U2 L2 F2 R2 F2 R' B2 U F D2 B2 L B D' B U'
51950 01-sep-2017 15:07:41 00:14.343 R2 B2 D U2 B2 L2 B2 L2 U F2 D2 L' D' F D2 U' F' U2 R' D R
51949 01-sep-2017 15:06:56 00:16.023 B2 L2 D' R2 D2 L2 U F2 U' R2 B2 L B R F2 L F2 U' F' R2 F2 U2
51948 01-sep-2017 15:06:17 00:14.839 D2 F2 L2 U' B2 L2 U2 L2 F2 D2 U' B D R' B' R2 B' F' L' D2
51947 01-sep-2017 15:05:41 00:16.024 R2 L2 D' R2 F2 D B2 D' F2 D' L2 F D' U L D2 B U F D' R2
51946 01-sep-2017 15:05:03 00:14.607 B2 L2 F2 D' U' R2 D B2 D' B2 U' F' D' U B2 R D' U2 B R2 D2 R
51945 01-sep-2017 15:04:24 00:20.919 F2 R2 F2 R2 F2 U B2 L2 D2 L2 D2 B L' D2 F U2 R D L U' B' D2

 

[mark49152]

My first clock average from last weekend.

 

[newtonbase]

I gave a friend of mine a cube a week ago. He's spent his family holiday teaching himself. He was aiming to get a guide free solve within a month but with a little goading he's now on a 2.35 PB and is asking how to practice his cross.

 

[JohnnyReggae]

Got my first sub-20 OH single yesterday and new PB of 19.79 ... super stoked !! Beat my previous PB by almost 2 seconds which had a PLL skip in it. This one was full-step which made it even more special for me.

F2 L U2 F2 L B2 L R F2 U2 R B D R2 U F2 D2 U' L B

z2 x' // inspection
R' U' F2 D2 // cross 4
x y' R' U R // f2l1 7
U2' z U' R' U z' R U' R' // f2l2 14
U' y R' U2' R U' R' U R //f2l3 22
U2' z U R' U' z' // f2l4 26
U2' R U2' R2 U' R2 U' R2 U2' R //oll 36
y R2 U R U R' U' R' U' R' U R' // pll U-Perm 47

// = 19.79 @ 2.37tps

 

[Fyzzna]

After all this craze about magnetic cubes I finally gave in and decided to make one myself. Bought myself a cheap cube (MF3RS) to test the water and went to work. It was surprisingly easy and a little under two hours later I was finished and had the cube assembled again.
The result: meh.
Honestly not very impressed with this. The outer layers feel pretty good but M slices are absolutely horrible on this cube (it's also very locky, almost more so than before magnetizing it). It might be because of the cheap cube but right now I definitely don't want to take the chance of ruining my valk by putting magnets in it.

On a happier note: With the MF3RS I also ordered a thunderclap (v1) and I absolutely love this thing! There is a serious possibility this could replace the valk as my main so there's that

 

[Jason Green]

I finally edited my Nationals video. It's really just me talking to people, with a couple tiny snippets of cubing.