Older cubers discussions

[mafergut]

For PLLs where you can cancel the final AUF, like my J-perms and R-perms I do AUF-PLL-AUF. For others that do not I prefer to rotate if it's just a y/y' and it will save me a final AUF, like with G-perms or T-perm. For algs that have x rotations the more so (like A-perms) to avoid a new rotation or an AUF on F/B. In some cases I have an alternative PLL to avoid y2, like lefty A-perms, alternative U-perms for bar in the front / back, or the T-perm + U2 variation. I also learned U-perm + U2 but never got proficient at them so I don't use them.

 

[pipkiksass]

I'm not so fussed about AUF for PLLs, but @mafergut's comment about x rotations reminds me that PLLs with x rotations probably cost me about 1/2 second. I just find the regrips really clunky. If I time them without the rotation and regrip, they're probably fast, but WITH rotation and regrip, they're awful.

May look for alternatives...

 

[pglewis]

I generally AUF-PLL-AUF unless I quickly see that it's already aligned and only a y or y' off. There's something very satisfying knowing I can just drop the cube after PLL.

New PBs : single: 23.01, ao5: 30.44, ao12: 31.51, ao50: 33.55

I thought I was mostly kidding about you overtaking me by this weekend

Great progress, I assume it's mostly from F2L lookahead?

 

[pipkiksass]

There's something very satisfying knowing I can just drop the cube after PLL.

For about 1/2 of my PLLs, I know what the AUF will be while executing, and build it into the alg. So for those, it's almost like the alg is 1 turn longer, so I'm just dropping the cube after PLL. Some of these I can predict AUF, based on relative sticker positions during PLL recognition; some I recognise mid-execution. I'm pretty sure I could apply this to all PLLs, if I could be bothered to spot patterns, and make PLL really fast.

I watched a video a couple of years ago, maybe a walkthrough of Feliks's, which talked about finger tricking PLLs differently depending on the AUF required after. Talk about getting advanced! But I can see that this would be really useful - if you know that a PLL ends with a U2, having your finger there ready to double-flick, then immediately drop the cube must be a good second faster than PLL - check AUF - regrip - U2.

One day...

 

[pglewis]

Some of these I can predict AUF, based on relative sticker positions during PLL recognition; some I recognise mid-execution. I'm pretty sure I could apply this to all PLLs, if I could be bothered to spot patterns, and make PLL really fast.

And then I guess the next step: recognize the PLL/AUF situation when you recognize OLL... and alternate OLL algs to influence the PLL case. More often than not I'm frantically trying to recognize PLL and fail to note the AUF situation; if I'm lucky I can catch it by the end of execution. I'm just now trying to work CP recog into the routine as a step towards 2 sided recog and I think that has helped me spot the AUF situation earlier and more often.

The rabbit hole runs deep.

 

[newtonbase]

The new A-perms I'm trialling leave the cube off by an x rotation so I'm usually left with a ABF. I'll need to remember to fix that before execution.

 

[JanW]

Then I know that for the inverse of that RUD alg, the FU edge sticker doesn't move, and I'm already prepared to AUF accordingly at the end with no pause or regrip.

Some very good disguised advise here! Look for the sticker that doesn't move. Previously, in the case of R perms, I've been looking at the headlights thinking they will move to the left or the right, which makes it more complicated to predict while executing. I now looked at all R and G perms for stickers that return to the same place after execution and by focusing on those it becomes a lot easier to predict the AUF case.

 

[pglewis]

The new A-perms I'm trialling leave the cube off by an x rotation so I'm usually left with a ABF. I'll need to remember to fix that before execution.

One of mine would leave me with an AFF. But it ends in an R2 which I perform with left thumb on the bottom L and right thumb on top R. It restores the rotation and does the R2 at the same time, leaving me with AUF.

 

[Jason Green]

I think the U-perm rotations might be a remain from my early cubing career when most solves ended in a U-perm and I probably had many bad habits involving unnecessary cube rotations. I shall try to stop doing those.

Caleb Miller recently had a video taking about rotations before PLL being a beginner mistake. I know some people shared his videos a while back. The only time I tend to still rotate (on purpose, if I know what's going on) is if it's a y or y' and my hands are in bad position anyway and I basically have to regrip to AUF anyway, then I just rotate and grip.

 

[Fyzzna]

I thought I was mostly kidding about you overtaking me by this weekend

Great progress, I assume it's mostly from F2L lookahead?

Thank you, although I haven't actually overtaken you yet

I would actually say most of the progress comes from improved lookahead, yeah (not F2L since I'm using Roux but you get the idea).

 

[mark49152]

Some very good disguised advise here! Look for the sticker that doesn't move. Previously, in the case of R perms, I've been looking at the headlights thinking they will move to the left or the right, which makes it more complicated to predict while executing. I now looked at all R and G perms for stickers that return to the same place after execution and by focusing on those it becomes a lot easier to predict the AUF case.

All my PLL algs give me a non-moving sticker. It's not necessarily one the of the stickers that appears not to move in a diagram of what the PLL swaps or cycles, because an alg might include an inherent AUF, as discussed before. I also find that thinking about the AUF during alg execution reduces the incidence of screwing up algs by thinking about the alg rather than relying on muscle memory. I used to do that quite a lot

 

[JohnnyReggae]

I continue to wrestle with the sluggish Valk syndrome. I did a full break-down and full cleaning of my stickerless last week, complete with q-tips, to get the gunk out. Reassembled and only put 2 drops of Silk in and worked it in. It was incredible. About six hours of solving later and it had slowed again. I think it's just the nature of the cube that it slows up with any gunking at all. My Thunderclaps are probably a wreck with gunk but they don't seem to care, with months between cleanings. A couple drops of Maru in the Valk helps; it might be a smidge too fast for me on the first solve or three then it's right where I want it for a session.

I found the same with my Valk. I left it unlubed for a while just because it seemed to perform better without anything in it. I've recently got hold of some DNM37 from TheCubicle and I must say that the Valk seems to love it. It is now buttery smooth and fast yet nicely controllable due to the magnets I put in, which I didn't like at first as they are quite weak but now after a few thousand solves I really like the fact that the magnets are there .. but not They provide some nice stability without having a magnetic force to push through when turning.

On the talk of the R perms ... for Ra I use the lefty version of the the Rb perm. I started doing it as it was on Badmephisto's list and I've stuck with it. I can do Ra lefty as fast as Rb righty.

I use a few lefty perms even though some suggest not doing it because of the regrips, which I can understand. But for now I'll stick with them.

 

[pglewis]

One of mine would leave me with an AFF. But it ends in an R2 which I perform with left thumb on the bottom L and right thumb on top R. It restores the rotation and does the R2 at the same time, leaving me with AUF.

I explained this very poorly but the left does the equivalent of an Lw' while the right does the R2 simultaneously. Not sure I actually get the left thumb fully on the bottom but that's the concept. Might be something similar you can do with an x rotated one (my current one starts x').

Thank you, although I haven't actually overtaken you yet

I would actually say most of the progress comes from improved lookahead, yeah (not F2L since I'm using Roux but you get the idea).

Me and my assumptions... I'm a CFOP guy but glad to see another Roux-er in here.

I'm tinkering with a few aspects of my solves again so I'm on the regression side of the see-saw, seeing a lot more high 30s and even sup 40s the past week and a half. There's still the chance of a lucky sub 20 single for me right now so I'm timing often just in case, but I refuse to let the timer pressure me into old habits for the sake of averages right now. If my past cycles are any indication I should see pay off in a couple more weeks.

I found the same with my Valk. I left it unlubed for a while just because it seemed to perform better without anything in it. I've recently got hold of some DNM37 from TheCubicle and I must say that the Valk seems to love it.

How much more life do you get out of DNM37 vs. Maru? I probably get an hour or a little more out of Maru, enough for a solid session and not much more. Not like the answer matters much, I know I'm going to pick some up to try regardless lol.

 

[JohnnyReggae]

How much more life do you get out of DNM37 vs. Maru? I probably get an hour or a little more out of Maru, enough for a solid session and not much more. Not like the answer matters much, I know I'm going to pick some up to try regardless lol.

Can't compare as I have never used Maru. Locally it's quite pricey and having a lube that wears out after an hour or so is not good value IMO.

DNM37 seems to last quite a while, perhaps not the lube itself but the feeling it gives the cube seems to last quite a while. For me I can easily go a week with a couple of drops, and with the Valk I average around 60 to 120 solves a day.

I've just put some in my new GAN Air that I put some magnets in last night, and it certainly seems to smooth the plastic out in that the feel is a lot softer without slowing it down like Traxxis for eg.

 

[pglewis]

I may have started an avalanche last night. It began by looking farther down the road to "what comes after full OLL for me", now that I'm about half way through, and the obvious answer seems to be COLL. Then I started thinking about edge control in order to land COLL cases consistently which lead me to thinking "or you could just use ZZ".

So I figured out how solving EO works just to get my feet wet, solving EO+CFOP just for the sake of demonstration. It turned into a giggle-fest for me with rotationless, super-easy F2Ls and OCLL every time, of course. EOLine + blockbuilding wouldn't be a difficult adaptation with a little practice to learn the ropes. I would already be able to 2-look the last layer worst case every time with OCLL + PLL and could start picking up COLL immediately. Honestly, if I felt I could work out EOLine or even EO+Line fairly quickly I'd switch this second.

 

[newtonbase]

I may have started an avalanche last night. It began by looking farther down the road to "what comes after full OLL for me", now that I'm about half way through, and the obvious answer seems to be COLL. Then I started thinking about edge control in order to land COLL cases consistently which lead me to thinking "or you could just use ZZ".

So I figured out how solving EO works just to get my feet wet, solving EO+CFOP just for the sake of demonstration. It turned into a giggle-fest for me with rotationless, super-easy F2Ls and OCLL every time, of course. EOLine + blockbuilding wouldn't be a difficult adaptation with a little practice to learn the ropes. I would already be able to 2-look the last layer worst case every time with OCLL + PLL and could start picking up COLL immediately. Honestly, if I felt I could work out EOLine or even EO+Line fairly quickly I'd switch this second.

So you've learned all those OLL algs and you want to switch to a method that only uses 7 of them?
I've been looking at it myself and EOLine is really difficult to plan in full never mind in just 15s. I like the F2L but I end up making a full cross first so really need to work on proper block building. It's a fun challenge overall.

 

[One Wheel]

I may have started an avalanche last night. It began by looking farther down the road to "what comes after full OLL for me", now that I'm about half way through, and the obvious answer seems to be COLL. Then I started thinking about edge control in order to land COLL cases consistently which lead me to thinking "or you could just use ZZ".

So I figured out how solving EO works just to get my feet wet, solving EO+CFOP just for the sake of demonstration. It turned into a giggle-fest for me with rotationless, super-easy F2Ls and OCLL every time, of course. EOLine + blockbuilding wouldn't be a difficult adaptation with a little practice to learn the ropes. I would already be able to 2-look the last layer worst case every time with PLL + OCLL and could start picking up COLL immediately. Honestly, if I felt I could work out EOLine or even EO+Line fairly quickly I'd switch this second.

ZZ has a lot of potential. I've mostly played around with untimed solves, just doing ZZf2l, OCLL, PLL, and it seems that I often end up with ridiculously inefficient last slot cases. I should probably look up some algs for them. When I've timed myself on ZZ I run roughly 1:00, versus 30-35 CFOP. I'm reasonably certain that if I figured out those LS cases and could actually plan EOLine in inspection I could average low 30s easily, and with a little practice break the 30 second barrier.

 

[newtonbase]

ZZ has a lot of potential. I've mostly played around with untimed solves, just doing ZZf2l, OCLL, PLL, and it seems that I often end up with ridiculously inefficient last slot cases. I should probably look up some algs for them. When I've timed myself on ZZ I run roughly 1:00, versus 30-35 CFOP. I'm reasonably certain that if I figured out those LS cases and could actually plan EOLine in inspection I could average low 30s easily, and with a little practice break the 30 second barrier.

Yes. I had to learn a new alg for when the corner is already placed which is a bit crazy but some are really fast and pretty easy despite the fixed orientation. I'm not expecting to do 3x3 in comp for at least 4 months so I have plenty of time to practice.

My latest batch of Thunderclaps hadn't turned up from Champion's Cube Store. It's been 4 weeks and it's rarely over 2 so I popped down to the post office on the off chance. They were in a sack about to be returned to China. If I'd been an hour later...

 

[pglewis]

So you've learned all those OLL algs and you want to switch to a method that only uses 7 of them?

IKR? And I'd have to give up my beloved hedge/sledge tricks I've worked into my F2L .

I've been looking at it myself and EOLine is really difficult to plan in full never mind in just 15s.

Yeah, I would have had a tough time working out EO from the outset. Even 8 months into CFOP I still had difficultly planning all 4 cross edges and it's only recently that I've developed a grasp of general EO and its implications. I think it's a foregone conclusion that I'm going to give it a go, though. I was literally giggling on every solve for over an hour. You know what has to happen when it's that much fun.

 

[One Wheel]

I think it's a foregone conclusion that I'm going to give it a go, though. I was literally giggling on every solve for over an hour. You know what has to happen when it's that much fun.

The ultimate goal of cubing is to have fun. Speed is just one way of doing it. Incidentally ZZ has to potential to be pretty fast, too.