Older cubers discussions

[pglewis]

Yeah, I'd agree it's definitely more a "nice to have" than a big deal because the timers are so simple. I wouldn't trade mine in for another puzzle though .

 

[One Wheel]

Yeah, I'd agree it's definitely more a "nice to have" than a big deal because the timers are so simple. I wouldn't trade mine in for another puzzle though .

I guess if you've got enough puzzles, and I have more than enough for most sane people, but how does that story go about the beggar asking the rich man how much money is enough?

 

[mafergut]

There are cheap timers, such as the Yuxin one. I bought it but I never use it, just because I have to set it up every time (I cannot leave it always out because I need to use my laptop for other things than cubing

Also, finally 5x5 Ao50 sub-3 minutes: 2:58.xx. Now continuing to Ao100

 

[newtonbase]

Great to hear you're getting to a comp! I'd get a Stackmat timer for home if you can; they're very simple to use and compact/portable, nearly zero learning curve. At the very least be familiar with the location of the reset button and press it every time you sit down to solve as it's easy for the judges to forget and you don't want to realize it won't arm itself at 12 seconds into your inspection time.

Otherwise, I figure just a deep breath before inspection and remind myself that I've done thousands of solves. Just do what I always do without "trying" to go fast.

I think you are approaching the competition with the right attitude. Just enjoy the experience and set targets for next time. Enter everything that you can solve no matter how slow you are. Try to get involved too by running or judging once you are comfortable with the process.
Timers are very easy to get used to but nice to own. Things like letting the judge uncover the puzzle seem a little strange at first and I still forget I have to sign my scorecard after 6 comps.

 

[newtonbase]

Anyway, I got some new puzzles.
Fangshi Jieyun - it has capped pieces and is small so I planned to make a magnetic OH cube but there's no space for the magnets and it's too rounded so it's on the shelf
Mini Weilong - another OH option. Much nicer this time but I'll need to loosen and lube it. A lot.
Moyu magnetic pyraminx - it's quick and cuts well but it just doesn't feel right and I don't like the sound. Much preferred the regular Moyu that I bought for my nephew. Maybe he'll swap?
Shengshou Master Pyraminx - this is a proper fun puzzle. Not quite the quality you need for speed but it's still pretty good. I couldn't quite finish the solve by myself but found someone who could. Highly recommended.

 

[pglewis]

Spent a good part of my evening practice time playing with the stickerless WuQue. A little catchy on the inner slices at times and not quite the smoothness of its 5x5 cousin but I definitely like it. No pops and it seems like I'd probably deserve it if it did. It's nowhere near as smooth and fast as my G4, but the G4 was catchy and very sluggish until break-in, so I'll definitely reserve judgement there for a while. It's definitely more stable which seems helpful in the 3x3 stage. I would prefer 60mm to the 62mm size but it's not a big deal to me.

I haven't been timing 4x4 for weeks, definite improvements with a 2:28/2:48 single/ao5 with OLL parity on every $#@$ one. 7 solves in a row that needed OLL parity fixes . Should be targeting a sub 2 min single next go-round but it's going back on the cube stand for 3x3 focus for now.

 

[One Wheel]

Should be targeting a sub 2 min single next go-round but it's going back on the cube stand for 3x3 focus for now.

For what it's worth, my biggest improvement in my 3x3 times was a direct result of a lot of 4x4 practice. Of course there is a limit: I think my 34/1:55 averages are a little bit out of balance, and my best chance for improvement on either is 3x3 practice and learning full OLL.

On another topic: I was noticing today that OLLs on my Weichuang seem to flow really nicely, but PLLs get really catchy. Has anybody else noticed this kind of phenomenon? Is this something with the cube, or just something I'll have to practice my way out of?

 

[Jason Green]

I signed up for my first competition: Montreal Open Winter 2017. I registered for 3x3, 4x4 and pyraminx.
I don't want to stress myself out with goals, so here is my list of goals

1. get a WCA id. (I think that should be not too hard )
2. see how competitions are organized and connect with people to see how I can help organizing in the future.

And I have a list of "nice to have":
1. Don't freak out too much during solves.
2. Get a sub-30 single in 3x3. (I currently average around 25-26 at home, so I think it's realistic)

Do you guys think I should buy a timer to practice with one? How difficult is it to adjust to it?
Will anyone go to that competition?

My first comp my only goal was to get an average. I didn't want to mess up and DNF. I had a +2 on my first solve I think but that was it so I was happy.

I like practicing with my timer sometimes but they are simple to adjust to do if you don't want to spend the money I wouldn't sweat it either. My times are basically the same with it without timer (actually usually better with which is weird, my PB's are all with it).

On another topic: I was noticing today that OLLs on my Weichuang seem to flow really nicely, but PLLs get really catchy. Has anybody else noticed this kind of phenomenon? Is this something with the cube, or just something I'll have to practice my way out of?

I had as similar experience with the GTS, in that my f2lb is bad because it's too fast, but I love last layer. In the end I did not use it as main because of that though.

 

[mafergut]

4x4 PB single out of nowhere when trying to get back onto the race to sub 1:20. Almost 2 seconds off my previous PB and with PLL parity. A pity because it would have been pretty close to sub 1 min otherwise. It happened on the 2nd scramble of round 47b:

1. 1:05.44 L B F Uw B' R2 U L Fw F2 Rw F' Uw2 F U' R2 Uw2 Rw L2 U2 R2 L2 Fw2 F R L' F B' U' L' B U' L' R2 Uw' R B' F Rw U

 

[pipkiksass]

For what it's worth, my biggest improvement in my 3x3 times was a direct result of a lot of 4x4 practice. Of course there is a limit: I think my 34/1:55 averages are a little bit out of balance, and my best chance for improvement on either is 3x3 practice and learning full OLL.

Amazingly, I didn't really get into 4x4 when I used to cube, it was all about 3x3 (and only 3x3). This time around, I've fallen in love with 4x4 in a big way! I think I do more 4x4 solves than 3x3 at the moment.

I've resolved not to time any 3x3 solves until I've learnt full OLL, except for forum comps. I spent about 2 hours last night going through some DIY alg sheets that me and @Logiqx knocked up about 2 years ago, trying to find the best algs for the 28 OLLs I haven't already committed to memory. I got about 18 cases through, and had to go to bed! Once I have the 'best' algs (for me) all chosen, I'll knock up a nice clean cheat sheet of just the cases I DON'T know, and then do untimed solves focussing on lookahead, and refer to the sheets where necessary. I need to work on my lookahead anyway, and it seems as good a time as any to put the timer away - then when I get to OLL I won't be tempted to 2-look, because I won't be worrying about squeezing the solve in as fast as possible.

Not sure if this will work, but it seemed like a good idea at the time!

4x4 PB single out of nowhere when trying to get back onto the race to sub 1:20. Almost 2 seconds off my previous PB and with PLL parity. A pity because it would have been pretty close to sub 1 min otherwise. It happened on the 2nd scramble of round 47b:

1. 1:05.44 L B F Uw B' R2 U L Fw F2 Rw F' Uw2 F U' R2 Uw2 Rw L2 U2 R2 L2 Fw2 F R L' F B' U' L' B U' L' R2 Uw' R B' F Rw U

Very nice. I love the fact that, with 4x4, there's SO much room for me to improve. For example, I entered the race to sub-2:30 race a couple of weeks ago, and my average to graduate from that race was 1:35.xx - i.e. within a couple of weeks I'd shaved a minute off my times!

I think it helps that I went straight for Yau with 323 edge pairing, rather than learning a 'beginners' method or plain redux. Edge pairing feels natural for me now - I think the stages where I have big improvements to make are first 2 centres and first 3 cross edges. My lookahead in these steps is RUBBISH - if I get an easy F2C case, and fluke cross edges at the front so I can see them, I'm scraping in solves sub 1:20.

Just sat down and did about 30 4x4 solves, set PB single (1:11.02 U Fw2 F2 R2 Fw' D' Rw' F' B2 D' B Rw' D' B' U2 F U Fw2 Rw2 U Uw2 F2 B' Fw2 U' Uw Rw F2 Fw Uw2 U2 D R2 F R' Uw U' R2 B' R'), Ao5 - 1:25.17, and Ao12 - 1:30.10:

Spoiler

Generated By csTimer on 2016-12-8
avg of 12: 1:30.10

Time List:
1. 1:37.34 U' L F Fw D' B' L' Rw' D Fw L' U' F' U' Fw' L D2 Fw R' B' Fw F D2 Fw2 B' F2 D2 B' U2 Fw Rw' L2 F2 U2 R2 D Rw2 U' Fw' R'
2. 1:21.41 R2 Fw2 D' B U R2 Fw' U2 D2 Rw2 Uw2 U' D L U2 D Rw2 D2 Uw' Rw2 Uw2 D2 B Uw' Fw2 U' L2 B' R2 L' D2 F L2 Fw2 F' Rw2 Uw2 Fw Uw F'
3. 1:27.65 Uw' R' Uw' L2 R Fw' B2 U2 F Uw2 F' D2 Uw2 Rw2 Uw2 Rw2 R2 D2 B' F2 R L Uw' B' Rw' Uw2 F' U D' R2 Fw' Uw Rw' L2 R2 Uw2 R2 D' F L'
4. 1:26.45 U2 B' Rw' L2 Uw2 R U2 Uw' F' R' B2 Rw2 U2 L' U2 L' U' Rw2 D2 F' Fw2 L' Rw D Rw' R' F2 Rw' R' Uw' R' Rw2 U Uw L Uw2 F2 Uw' U2 F2
5. (1:12.45) B Uw R U Uw2 R' D B2 D2 Fw D B U L' D2 L2 R' Fw U2 F2 Fw2 R2 D Rw2 Uw' D' U2 Rw2 L' B' U2 D' L2 R' Rw2 Fw2 Rw' R' U' R2
6. (1:44.20) F Uw' Fw' L F L Fw U' Uw2 D' B2 R F2 L' Fw' R Fw2 F L' R B R2 F' D' L' Uw R B Uw' L F' Fw' L' Rw D2 U F Rw2 D2 B2
7. 1:32.24 Fw Uw Rw' R D' F2 D' Rw' R' L B L' Fw2 F Rw' Uw' B F' U2 Uw D Rw B' R2 D' F B' D' Rw2 Uw R' Fw2 D Uw' R Fw2 F' R2 Uw Rw2
8. 1:40.06 F' L' Rw' Fw' L Uw' U' Fw2 U2 F2 R D F' Rw L D2 Rw B' Rw2 F2 Fw' D' B2 Uw2 R' F2 D Rw2 Uw R F Rw' Uw' Fw' F2 Rw2 U Fw2 Rw' Uw
9. 1:28.80 L D F Fw' U2 Uw2 R2 F' Uw' D F' Uw' Rw' R' Fw' Rw2 B' F' Uw2 Rw U2 D' R D' U Fw2 R B2 F2 D L' Fw D' Rw F2 Uw' B2 L2 R' Fw'
10. 1:40.66 F' L2 B' Rw2 Uw' D Fw' D B2 Rw' Uw2 F2 D2 B2 Fw D2 B' Rw2 D' Rw2 B F' Fw' Uw' Fw F2 D2 U2 L2 Fw2 Rw' Fw' L' R' Uw L' B D' Rw2 F'
11. 1:23.98 D2 R L2 Uw F' Rw' Fw' U2 Fw Rw2 R' F' L2 Rw R' Fw' F D R F Uw' R' U' R2 B' U D2 Uw' Fw D' F' Fw2 U' F L2 B' F2 R2 Uw F
12. 1:22.38 D2 Rw Uw' Rw' Fw2 Rw F U Rw2 D' Uw2 F Fw' Uw' L2 Rw' Fw2 Rw F Fw2 B L Fw2 Rw Fw R' F2 Uw Rw' B' Fw2 F' Rw' Fw2 Rw2 Uw D2 R2 Fw2 L2

I think the fact I have a 1:12 solve followed by a 1:44 (double parity, bad edge pairing, etc..) is evidence that I need more practice!

Off to learn some OLLs...

:rolleyes:

 

[mafergut]

@pipkiksass, You just started to solve 4x4 some weeks ago and you are already chasing me... and soon passing me, I'm sure.

My weakest parts are L4C and also F3E. My F2C are good I think, even though I don't always make an effort to plan the whole 2nd center during inspection, just the 1st one and, depending on how easy the scramble is, maybe just the 1st bar of the 2nd center. I should practise with unlimited inspection to improve this.

I can only start with either white or yellow center (and always do white cross) as I don't know the colour scheme well for other F2C combinations. That's what I like about 5x5, you always have the centers there to guide you. It's a pity, because I very frequently see easy centers in blue/green or red/orange that I cannot take advantage of.

 

[megagoune]

Thanks for all the answers @pglewis, @One Wheel, @muchacho, @mafergut, @newtonbase, @Jason Green.
I think I will buy a timer, but my reason will not be because I need to practice with it, but because I just want to try.
Which one works well with a Mac and cstimer?

Shengshou Master Pyraminx - this is a proper fun puzzle. Not quite the quality you need for speed but it's still pretty good. I couldn't quite finish the solve by myself but found someone who could. Highly recommended.

I also received mine. 4 weeks in transit, that's always the hard part. I really like it. It's fun to solve. I am waiting for Shengshou to release a 5 layer version now. I am not sure I want to buy the Meffert's one.

Quick photos of the puzzle and its introduction to the rest of the tetrahedron family.

 

[Jason Green]

The stackmat pro should work in theory, but I did not have the correct input ports on my iMac to make it work.

 

[pipkiksass]

My weakest parts are L4C and also F3E. My F2C are good I think, even though I don't always make an effort to plan the whole 2nd center during inspection, just the 1st one and, depending on how easy the scramble is, maybe just the 1st bar of the 2nd center. I should practise with unlimited inspection to improve this.

I can only start with either white or yellow center (and always do white cross) as I don't know the colour scheme well for other F2C combinations. That's what I like about 5x5, you always have the centers there to guide you. It's a pity, because I very frequently see easy centers in blue/green or red/orange that I cannot take advantage of.

I'm also white/yellow centres and white cross. I don't think this is necessarily a bad thing, but I sometimes see 3/4 of a blue centre or whatever during inspection, and wish I were colour neutral!

I only plan first centre during inspection. I've never tried planning the second, unless the scramble is SUPER easy. I'll sometimes try to track 1/2 pieces of the second centre, but I know the 'pros' can plan and insert F2C blind. I think I could definitely improve F2C, which might help me chasing the F3E?!?

Try

for L4C. I use half-centres and, 90% of the time, this gives me a really good case. On a good solve, I think I probably do L4C sub 10. It can literally be done in under 10 turns if you have a few 2x1 blocks after F3E. Takes a little getting used to, but (once again) I'm fortunate that I went for 1/2 centres and 323 from the start - once you get used to them, they're ridiculously easy.

I need to get in to 5x5!

Cubing is eating my life.

This is a good thing.

Edit: carried on with 4x4, and got sub-1:25 Ao5 (1:24.67) and sub-1:30 Ao12 (1:27.77)

 

[CLL Smooth]

I'm also white/yellow centres and white cross. I don't think this is necessarily a bad thing, but I sometimes see 3/4 of a blue centre or whatever during inspection, and wish I were colour neutral!

I only plan first centre during inspection. I've never tried planning the second, unless the scramble is SUPER easy. I'll sometimes try to track 1/2 pieces of the second centre, but I know the 'pros' can plan and insert F2C blind. I think I could definitely improve F2C, which might help me chasing the F3E?!?

If you haven't given Hoya a shot, I suggest that you do.

 

[pglewis]

I haven't been timing 4x4 for weeks, definite improvements with a 2:28/2:48 single/ao5 with OLL parity on every $#@$ one. 7 solves in a row that needed OLL parity fixes . Should be targeting a sub 2 min single next go-round but it's going back on the cube stand for 3x3 focus for now.

Just wanted to know what a decent time for me is now without OLL parity and had to wait like 7 more solves still lol. That's kinda crazy. Anyway, the answer was 2:07, so sub-2 not so far out of my reach.

 

[efattah]

I'd like to introduce a new metric called 'lookahead fraction' (LF). LF is the fraction of your solve that you spend pausing, examining, recognizing cases, deciding what to do, etc.

The procedure to measure your lookahead fraction is simple: take a scramble, solve it legit and time it; then solve the same scramble several more times using the exact same series of moves until you execute it smoothly without pauses. And time the solve that has no pauses. The ratio gives your lookahead fraction. Ideally the scramble should yield a full step non-lucky solve, and even better if you can take an average over several solves.

As an example I did the test myself. The first solve was 15.7. Reconstructing the solve after I found it 46 moves STM which is right about average for the LMCF method I am using; it was a full step non-lucky solve. The second try ended up at 11.50 and this quickly fell below 10 seconds until after several tries I made it through without any lock ups and got an 8.77. So theoretically the time spent actually turning was 8.77 / 15.7 = 55.86%, and the rest of the time (100.0 - 55.86 = 44.1%) was spent in dead time. So my LF = 44.1%, and this is the fraction of my solve spent recognizing stuff, and it is the theoretical amount of time I would 'save' on a solve if I had flawless lookahead. This number is very useful because many people wonder if they should learn new algorithm sets and so on, or how fast could they get with 4LLL as an example. This test measures your raw TPS (in my case the original solve was 46/15.7= 2.93 TPS, and the RAW TPS with no pauses is 46/8.77 = 5.24). So I know from this that my hands can turn a whole solve at 5.24, but my real TPS collapses to a horrid 2.93 when you incorporate pauses. This tells me that rather than learn new algorithms I am still probably better just working on lookahead and recognition speed. To be fair I should try to calculate the LF on a series of solves and I hope to do that in the next few days. Here is the solve after a few runs through when pauses were eliminated, I should have taken a video of the original solve as comparison! Next time.

 

[Jason Green]

@efattah I think that's a cool idea, nice solve also! The only other thing which might very slightly improve besides pauses for lookahead would be your transition speed between the turns themselves? In other words going from first slot to second slot for example you would be practicing the same combination of moves while you do this, so basically it would be like drilling an alg and your actual execution time might improve too. Although I guess you could argue that you've done those transitions in thousands of other solves so maybe not? In any case I think it would be insignificant so the LF would still be pretty accurate.

The way I know I have lots of room to improve lookahead is watch a video of myself. Holy crap I still pause that much!?! Haha

 

[mark49152]

@efattah Yes it's an interesting experiment that was tried on this thread 2-3 years ago and also turned into a sort of mini-competition IIRC. I saw similar huge improvements, and used the results to conclude that with enough practice I should be able to reach at least sub-15 global even if my fingers never get any faster.

 

[JohnnyReggae]

Had a few more days to play with the Valk and I must say that it's kind of growing on me ... like a fungus Even with the stock sticker shades my recognition is great and I seem to be getting consistent lower times than I do with my AIR. Not sure if it's the puzzle or the sticker shades. Initially I wasn't a fan of the stock shades, but they do contrast well and my recognition is still spot on and sometimes better. I do however prefer the feel of the AIR. It just feels nicer in the hands so I have been swapping between the 2 regularly to train both the change in colour recognition and very different feeling in the cubes. I have ordered some of my prefered shades for the Valk, however I'm considering ordering some Valk shades for my AIR to see how that goes.



Also unboxed another package yesterday. 5 in total, Master Pyraminx, 2xCrazy Winmill, 1xAxis, 1xFisher

The Master Pyraminx is awesome. There are no ball bearings to click things into place which feel different but I don't mind that at all. The puzzle moves nicely, feels great. The solve is also interesting, but I suppose there are a few ways of accomplishing that. I'd say that it is a good addition to the collection



The other 4 are going to be Frankenstein'd between themselves and we've yet to see what is going to come out of that. Should be interesting as there should be some Ghost cube properties brought in somewhere.