Older cubers discussions

[h2f]

@pipkiksass, @Selkie, @earth2dan, @mafergut: Thanks for the kind words of encouragement

To answer Pip's point, 4BLD is exactly the same principle as 3BLD, just bigger. About 2.5 times the information, plus an extra piece type, but otherwise the same principles, process and skills. I did improve a lot in the last few weeks, and I think a big reason for that is that doing more 5BLD and bigger MBLDs has improved my memo skills so now I can easily do several 4BLDs in a practice session, whereas before my max was 1-2 before my memory got tired. So I just got more practice done. Maybe 80-100 solves in the last month. It's a really fun event and my favourite right now.

It's not 2.5 but less - Maskow had made some math about it and it showed that it's 2.25 3bld cubes. And you showed me the way to practice - I'm gonna to make 2-3 5bld attempts per day. Polish Champs in 25 days...

 

[earth2dan]

In other news, while waiting for my Chinese food last night there was a kid of about 11 with a Gans 3x3 who I reckon was about a 60s solver. He did 4 different OLLs one solve, and quite slowly. I resisted the urge to go over and show off, but almost wish I had!!
Just nice to see another cuber IRL, it's a first for me!

Only twice have I met another person in real life that solves. I took my kids to the park once and there was a girl with a Rubik's brand 3x3 there. The other time was when Dan Fast (CrazyBadCuber) was here in Winnipeg.

Maybe the twisty puzzle business is starting to grow out of the niche market after all?

I was visiting my sister in Alberta this summer and there was a boardgame store in the mall that had a bunch of Meffert's puzzles. First time I've seen non-Rubik's brand puzzles in a brick and mortar store. Hopefully as it becomes more popular we'll see even more retail options, and more events.

Looking for some cube recommendations for my next puzzle order:

• Megaminx - I know we have some good Mega solvers here. I have an old ShengShou and a Dayan Stickerless (ridged), but I can't say either of them are really good. The Dayan is better, but centers tend to pop on looser tensions.
• 5x5 - anyone have an opinion on the MoYu WeiChuang GTS vs QiYi WuShuang? I have a stickerless Yuxin, and both black/stickerless BoChuangs and they're all pretty great, but there's room for improvement.
• 3x3 - I am definitely getting the Gans 356 Air, but I'm torn on whether or not I want the Valk 3. Anyone here swear by the Valk 3?
• Others - Anyone fallen in love with a new puzzle in recent months that they wish to recommend? I'm all ears.

 

[mark49152]

It's not 2.5 but less - Maskow had made some math about it and it showed that it's 2.25 3bld cubes.

I'm not sure how the maths would account for cycle breaks, flips and twists, but just comparing target counts and ignoring cycle breaks, it would be 18 versus 53, and typically an orientation can be found that leaves 7 centres solved. So (53-7)/18 = 2.55. Not arguing with Maskow though - my maths is probably oversimplified.

 

[pipkiksass]

Anyone here swear by the Valk 3?

Yeah, it's awesome, and not MASSIVELY expensive (cheaper than the Gans and Weilong GTS). Very smooth cube which cuts anything you would reasonably want to cut. I find M turns lovely on it, my Z, H, and U perms are much faster on it.

But then I haven't tried either of the other cubes above. To be frank, I have no intention to buy either, I love the Valk. On the right tension it never pops, is incredibly smooth, and still really controllable.

Edit: I've fallen in love with 4x4, but that doesn't count, does it?! I think you want something a LITTLE more obscure than what is probably the second most popular puzzle in the world?

Sent from my iPhone using Tapatalk

 

[h2f]

I'm not sure how the maths would account for cycle breaks, flips and twists, but just comparing target counts and ignoring cycle breaks, it would be 18 versus 53, and typically an orientation can be found that leaves 7 centres solved. So (53-7)/18 = 2.55. Not arguing with Maskow though - my maths is probably oversimplified.

His argumentation looks the same as yours. But he argued that 46 targets in 4bld leads to no more than 26 words needed to memorize the cube but in 3bld 18 targets leads to 11,5 words in average needed to memorize the cube because of cycle breaks twists and flips. And I'm sure he checked it. 26/11,5 = 2,26. In fact I think it's something between 2,26 and 2,5 cubes.

See here his comment:

 

[Logiqx]

Oooh. I've just found a really sick F2L trick whilst playing around...

r U2 r' F2 - execute everything right handed

I've often looked at this case and thought there must be a clever solution.

My previous approach was y' R U R2' U' R

Pretty cool to find a 4 move re-gripless solution!

Edit: Solves FR pair when the edge starts in FL and corner is on top

 

[pglewis]

The weather has still been surprisingly nice during the day here so I've been taking my cube with me to go get my daily afternoon coffee and practicing in the outdoor plaza. For better or worse I've found that playing with a cube in public prompts a fair bit of unsolicited conversation. Usually it's either "have you solved it yet?" or "I just peel the stickers off" but I did have a UC student ask what my times are one afternoon, he's roughly in the same same ballpark. Over the weekend two kids wandered over, one could solve one layer fairly quickly... not permuted properly but it was obviously not his first time. Wish I'd had my Mars or something with me, they'd have gone home with one of my spares.

5x5 - anyone have an opinion on the MoYu WeiChuang GTS vs QiYi WuShuang? I have a stickerless Yuxin, and both black/stickerless BoChuangs and they're all pretty great, but there's room for improvement.

I don't have any other 5x5s to compare but I really like the WuShuang a lot. Can't imagine it popping without me deserving it and it turns far better than either my G4 or Yuxin 4x4s did out of the box. The only question mark in my mind would be how well it scales up for faster solvers. EDIT: I also really like the smaller size.

3x3 - I am definitely getting the Gans 356 Air, but I'm torn on whether or not I want the Valk 3. Anyone here swear by the Valk 3?

I've been a Thunderclap man since I got mine over the summer. My Cubicle premium Valk broke in with just a few solves and became rather fast. I resisted the temptation to adjust it and now my TC feels quite sluggish already. Kind of amazing how quickly I can adapt to a new feel but I'm gonna call it at this point: new main.

 

[pglewis]

@Logiqx: keep 'em coming, at this rate I may never get consistent again but you know how I love new F2L tricks right now!

 

[pglewis]

I was intending to do some solves, break it in a bit, time some 3x3 and 4x4 solves, then consider whether I should do a spring swap... when I had a small internal pop while scrambling. Thought I'd soldier on, when I had an internal explosion on OLL parity.

Similar experience here with the 62mm version, I just don't trust it much at the moment.

 

[mark49152]

His argumentation looks the same as yours. But he argued that 46 targets in 4bld leads to no more than 26 words needed to memorize the cube but in 3bld 18 targets leads to 11,5 words in average needed to memorize the cube because of cycle breaks twists and flips. And I'm sure he checked it. 26/11,5 = 2,26. In fact I think it's something between 2,26 and 2,5 cubes.

See here his comment:

Is that link correct?

Yes I can imagine flipped edges would add more to 3BLD but it would depend how you memorise them anyway. Would they be counted as one target or two? What about parity? The only objective way to eliminate the effect of memo system is to count only targets, not words. Accounting for the average number of cycle breaks and flips would make sense, but I don't know the math. I would guess 1-2 of each for 3BLD and 2-4 wing cycle breaks for 4BLD. Centre cycle breaks can almost always be avoided.

 

[Logiqx]

Oooh. I've just found a really sick F2L trick whilst playing around...

r U2 r' F2 - execute everything right handed
...
Edit: Solves FR pair when the edge starts in FL and corner is on top

Another discovery is the inverse... F2 r U2 r' which solves F2L case 31 from FR to FL.

Previously, I've been doing R L' U2 L R' or y' R' U R2 U' R' but I think tonight's find is better!

 

[pglewis]

I've mentioned it before but I'm sold on F l' U' l and the lefty mirror for hedgeslammer. Getting a fair bit of use out of it on 4x4 and 5x5 too.

 

[pipkiksass]

Similar experience here with the 62mm version, I just don't trust it much at the moment.

My review of the Yuxin Blue

So it took me nearly TWO HOURS to reassemble the cube. It's a long time since I've reassembled a 4x4, and getting the first 3 combined internal edge pieces to stay in was a nightmare. There's even less room inside the cube then there was in my Weisu, and I set the tensions pretty tight, which gave me very little room for flexing the internal centres to fit the internal edges in. Also I should've finished a corner before moving on, I tried to insert a 4th internal edge, should've done 2, then built the corner, then a 3rd.

In retrospect, looser tensions, less ambition. More haste, less speed(cyoobin)?

BUT, once lubed and correctly assembled, and with the Fangshi springs, it's AWESOME. Does it lock up? Yes, a little, but probably because I can't turn a 4x4 properly. Most of my lockups are at the 3x3 on 4x4 stage, so I did some 3x3 on 4x4 averages (per my previous post, I'm a noob at 4x4, so this is the stage where I can honestly tell if the cube is holding me back).

I did an Ao12 on my Weisu, which was 34.12.
I did an Ao12 on my Yuxin, which was 26.77 (best 22.53, used my phone timer, but lower variance probably than my 3x3 solves - will put this to the test with CSTimer tomorrow)

I was averaging about 2:40-2:50 with my Weisu. With no additional practice (just picked it up and solved it a few times, my rolling Ao12 is down to 2:36, with my last solve on the Weisu being 3:09.141, followed by a 2:20.003 and a 2:11.756 on the Yuxin. The 2:20 had double parity; the 2:11 had PLL parity.

I was partaking in the race to sub-30 in 3x3 on 4x4 to get faster turning on 4x4, and I'm pretty sure I'll graduate from that immediately. I was also taking part in race to 2:30 on 4x4, so I'm pretty sure I'll be getting past that landmark fairly soon as well. I'm confident will be sub-2 using this cube in no time whatsoever. Possibly later this week!

So it's 60mm, which may account for a bit of the speed increase, and also stickerless, which the WCA now say doesn't help... so... but it's pretty! It's also significantly quieter than the Weisu, even having only been solved about 5 times. Turning feels very smooth indeed. I have the tensions a LITTLE looser than my Weisu, but it feels infinitely MORE controllable (go figure). I'd imagine it's because everything internally fits together so tightly.

Sadly it exploded internally before I could judge whether the spring swap was worthwhile. I won't be taking it apart to swap the springs back in a hurry. With the new springs, lube, tensioned to avoid more pops, it's STREETS ahead of the Weisu. But who uses a Weisu anymore? Would I recommend it over a 2-year-old 4x4? Unsurprisingly, yes! Is this, therefore, the most pointless review ever? Yes!

Sorry for wasting your time. I'll be sticking with this cube until I'm sub-xx, or it starts to pop lots and makes me cry. Initial outlook, however, is good.

 

[One Wheel]

• 5x5 - anyone have an opinion on the MoYu WeiChuang GTS vs QiYi WuShuang? I have a stickerless Yuxin, and both black/stickerless BoChuangs and they're all pretty great, but there's room for improvement.

I really can only compare my Shengshou and my Weichuang GTS, and the Weichuang is obviously an improvement. It is very fast, potentially too much so. I end up getting a lot of lockups when I'm scrambling, but it doesn't seem to be nearly as much of an issue during the solve. I've had a very few pops of the corner center pieces. I lubed the core of mine with Cubicle Wt. 3 and the pieces with Silk, and haven't messed with the tensions too much. It still feels a little blocky and catchy to me, but my guess is that with another 200-300 solves (it's probably got about 100-125 right now) it could be absolutely fantastic, and it will probably really come into its own around 1000 solves. I know that the cube isn't supposed to make much difference until you're fairly fast, but going from the Shengshou to the Weichuang immediately improved my times from typically 4:00-4:30 to typically 3:30-4:00. As of today my best solves are 2:53.60 and 3:00.31. My pb with the Shengshou was 3:44.99 roughly a month ago. It will be a long time before I can take advantage of all this cube has to offer.

 

[pglewis]

Another discovery is the inverse... F2 r U2 r' which solves F2L case 31 from FR to FL.

Previously, I've been doing R L' U2 L R' or y' R' U R2 U' R' but I think tonight's find is better!

Since I have a sudden love of M-slices I've been doing M F2 M' for this case.

 

[h2f]

Is that link correct?

Yes. Maskow's comment under the movie. I dont know how to link comments.... I was thinkig about it and I think he is right. I count words not targets. And it's quite good: the longest memo wings I had took 13 words (+2 to normal) + twists - it was around 26 (46 means 23 words + 2 + twist + parity = a little bit over 26). I know he checked words needed to memo cube due to his mbld attempts. So his maths is quite ok. I like the idea of thinking about 4bld that it's 2 x 3bld + a little piece. And his math goes further - 5 bld is not 5x3bld (2,5 x 2) but 4 and a bigger pieces. It makes sense.

Ok. Let's get back to practice. Morning scramble and 57.55. I wish I could do this during comp...

 

[Logiqx]

@Logiqx: keep 'em coming, at this rate I may never get consistent again but you know how I love new F2L tricks right now!

If you like tricks then be sure to keep a record of them in one place. As you get faster you'll know which cases really need the speedups and what flows best in solves, even if algs are long. It's partly dependent on the ability of your fingers to do certain movements as well since they will learn to do some things well with practice. I will share the fact that I actually dropped a heap of tricks whilst pushing for sub-20 as simplicity was the key to my look ahead. It's only now that I'm re-introducing tricks for cases which frankly seem to take half a second longer than necessary.

I've mentioned it before but I'm sold on F l' U' l and the lefty mirror for hedgeslammer. Getting a fair bit of use out of it on 4x4 and 5x5 too.

I think that's a trick a lot of people do when they start to use sledgehammer but you should learn to do R' F R F' if you're to gain the full benefit. There was a time when I used to use F l' U' l but I stopped doing it a long while ago. Watch any fast person solving and you'll see they don't use that method of execution since it involves re-gripping with your left hand and if one hand should be fixed it's the left one. The wide l' also affects lookahead as it hides a lot of LL pieces at the back.

Since I have a sudden love of M-slices I've been doing M F2 M' for this case.

I used to try that as well but it is too awkward in the middle of a solve. It seemed good in theory.

I like that you're keen on tricks. Just be careful you don't over complicate your solves as your main priority should be to know everything blindfolded right now. That will lead to your biggest speedups as your look ahead will develop more quickly.

 

[defhacks]

Just be careful you don't complicate your solves as your main priority is to know everything blindfolded right now. That will lead to your biggest speedups as your look ahead will develop more quickly.

This is definitely something I struggle with, drawing lines between efficiency and f2l "bling".

One of the initial thoughts I had was to start looking at cubesolv.es reconstructions to see what actually happens in solves, and quickly started wondering how much "confirmation bias" (or something like it) happens.

Specifically, most of the solves I saw had really elementary F2L steps, which of course, leads to easy solves, which leads to fast times, which leads to being on cubesolv.es, then I thought about looking only at slow solves in cubesolv.es, like maybe the worst time in an ao5, but then started wondering if the slower solves are slower due to inefficiencies that would be bad to use an example, then I started to wonder if I was overthinking things, and now I'm thinking about how bad my penchant for run-on sentences is..

 

[Logiqx]

This is definitely something I struggle with, drawing lines between efficiency and f2l "bling".

One of the initial thoughts I had was to start looking at cubesolv.es reconstructions to see what actually happens in solves, and quickly started wondering how much "confirmation bias" (or something like it) happens.

Specifically, most of the solves I saw had really elementary F2L steps, which of course, leads to easy solves, which leads to fast times, which leads to being on cubesolv.es, then I thought about looking only at slow solves in cubesolv.es, like maybe the worst time in an ao5, but then started wondering if the slower solves are slower due to inefficiencies that would be bad to use an example, then I started to wonder if I was overthinking things, and now I'm thinking about how bad my penchant for run-on sentences is..

The best confirmation of specific tricks being any good is when they are used in the Friday Fours by Faz. Several of my favorite algs have turned up which is a good sign!

Sent from my SM-G930F using Tapatalk

 

[mark49152]

Yes. Maskow's comment under the movie. I dont know how to link comments.... I was thinkig about it and I think he is right. I count words not targets. And it's quite good: the longest memo wings I had took 13 words (+2 to normal) + twists - it was around 26 (46 means 23 words + 2 + twist + parity = a little bit over 26). I know he checked words needed to memo cube due to his mbld attempts. So his maths is quite ok. I like the idea of thinking about 4bld that it's 2 x 3bld + a little piece. And his math goes further - 5 bld is not 5x3bld (2,5 x 2) but 4 and a bigger pieces. It makes sense.

OK, I see the comment, but I don't know how to translate it . Not on a phone, anyway.

Anyway I get the point, but I still think the comparison is a bit pointless unless the quantification is objective. If it takes into account memo technique, it's subjective. So it will be different for different people.

Let's take another subjective view. For me a 3BLD takes 1 image location, 3 audio words and maybe some visual flips. A 4BLD takes 4 locations and 4 audio words. How do we even compare those?

Without an objective measure, the best we can say is it's "about" 2.2-2.5 but really the decimal place precision is meaningless when the numbers aren't derived with that level of accuracy.