Fred's Thread

[Fred]

2:00.37 megaminx single. So close to sub 2-min! It's coming soon. I also found out that my method is called Omega LSLL, though I'm neutral as to which last pair I solve.

Also lubed my clock, and it's fast now.

 

[Fred]

Got my sub-2 mega single: 1:53.29. Also had a terrific 3BLD result in the weekly comp. First sub-5, first sub-4, and first mean (also sub-5): 4:59.23, 3:57.61, 4:50.46

 

[Garf]

2:00.37 megaminx single. So close to sub 2-min! It's coming soon. I also found out that my method is called Omega LSLL, though I'm neutral as to which last pair I solve.

Also lubed my clock, and it's fast now.

This is a really interesting approach to doing megaminx last layer. However, I don't see how this is faster than doing normal 4LLL or LL with less steps.

 

[Fred]

Megaminx discord seems to think it's better than 4LLL and worse than 2LLL. Mostly I didn't want to have to learn a bunch of algs just for megaminx. I've messed around with Heise on 3x3, and megaminx is much better suited to that sort of approach due to edges and corners each having even permutation parity.

 

[Fred]

I learned OBLBL just for kicks. I got a 1:26.84 single and a 1:31.99 Ao5, moving it up to 3rd place among my 4x4 methods. It feels like a cross between QTPI and Yau5, so it was pretty quick for me to learn. The main differences are that for building the slices, I should use wide moves instead of always slicing and I can use the R layer to manipulate edges, and when I get to edge pairing I don't already have 4 edges built in the D layer so I just do a F or F' after pairing the first two edges to get my fourth paired edge into the D layer. The edge pairing is L4E-ish, but there a 7 of them to pair up, so I'd miss out on freeslicing 3 edges on 5x5+. On 4x4 it's just 2-2-3 edge pairing instead of 3-2-3, so no big deal. The 3x3 stage is traditionally done with Petrus, but solving the last cross edge and doing CFOP/CFCE works just as well. I'm not switching to this method; I just enjoy learning new methods.

 

[Fred]

Got a sub-20 Ao100 on 3x3.

 

[Fred]

Tried a few ZZ solves. Got a 17.07 single and a 21.77 Ao12. EOcross is still hard, but I can plan it out some of the time, and the rest of the solve can go really fast.

 

[Fred]

Some FTO progress: 1:10.30 single, 1:26.09 ao5, 1:32.19 ao12, and the logo sticker fell off.

 

[Fred]

59.50 FTO single! Also had a great 3x3 session today with PB ao5 (17.33) and ao12 (17.73). I'm doing simpler F2L solutions for better lookahead until last slot, when I can break out my cool tricks.

 

[Fred]

First sub minute 4x4 solve! at 55.81! No parity. A couple days ago I had a 1:03 with OLL parity, so I knew sub-minute was coming. After videoing myself for Garf's comp I found a couple things I could improve, most notably not doing an x rotation to do sledge/hedge thereby blocking my lookahead during edge pairing.

 

[Fred]

Competition coming up this Sunday with 2x2, 3x3, mega, clock, and skewb. I'm doing CLL for 2x2, just my 3x3 algs with slight modifications on a few of them to avoid D moves. 3x3 I'm hoping for sub-20. For mega I'm still doing a lot more solves over 2 minutes (the cutoff time) than under, though I had a nice Ao5 (2:07.94) with 2 sub 2s: 2:06.12, (1:52.35), 2:19.48, 1:58.22, (2:24.09). For clock, I recently learned Alyz method because 7simul was too much. Alyz has four easy things to memorize and does four simultaneous moves with the rest just single moves (and no D moves), so it's like a beginner version of 7simul. I got a 16.95 average with it in the weekly comp, compared to 22.14 the week before with plain old flip. I know it's slow, but I only bought my first clock because of this comp. Skewb I'm doing Sarah's intermediate, but for some of the cases I'm still shaky with the recognition. Still, other than 3x3 this is my first time competing in these events, so any result will be a PR.

Meanwhile, I'm still doing a lot of big cubes practice. I slightly blame Garf's comp, but mostly I just like big cubes. I got a really nice 5x5 result in the weekly comp with a 1:42.37 pb single and a 2:00.90 pb average, both by several seconds. For 4x4, I just realized I've been doing PLL parity wrong fingertrick-wise (Rw2 R2' instead of just slicing), so I'm trying to fix that, but so far my brain just automatically does my old, bad fingertricks when PLL parity comes up in a solve.

Edit: Starting to get the hang of new PLL parity fingertricks. Down to sub 1:15 ao100 for 4x4 and sub 2:20 ao100 for 5x5

 

[Fred]

I made the megaminx cutoff! And sub-20 (sub-19 even!) 3x3 average!

2x2 was first and I got a 6.85 average and 4.96 single. Not the best, but close enough to my global average that I'm happy with it. I think I had two solves where I struggled to plan out a layer in inspection and got a slow time as a result.

3x3 was next. 18.82 average and 17.91 single. Each solve got faster than the last: 21.00, 19.22, 18.80, 18.43, 17.91. I figured out if I tell myself to turn slowly, the adrenaline will still make me turn reasonably fast, but I can actually look ahead in f2l and not lock up (as much) in the last layer. If I try to go full speed, I have a lot of pauses. I'm really happy with this result.

Next was megaminx, and I decided to continue that slow turning strategy, and it worked. 2:02.17 on my first solve, just over the 2 minute cutoff felt rough, but I got the sub-2 on my next solve: 1:57.50! 2:00.79 average (only 0.02 off my at home PB) and 1:57.07 single, and my slowest solve was 2:04.25. Consistency! (Also I messed up when judging because I didn't realize the time limit was on the score sheet, so I didn't know it was lower than 10 minutes. One of my competitors took over 6 minutes and the cutoff time was 4 minutes. Luckily there were still other judges available, so I don't think I single-handedly held up the competition.)

Clock came next, and it didn't go as well as I would have liked. 21.43 average and 17.07 single. The first couple solves I struggled to figure out how to start on a stackmat with the clock standing up. I had one solve where I started with it lying down. I only had one solve where I messed up the execution, and the fifth solve was good, so overall I'm still pleased.

Finally, skewb. I got a 10.71 average and 8.18 single, which is quite good for me. I had a counting 8, 9, and 13, so I was regretting the 13 a bit. But I talked to another guy in my group with a similar average who had a counting 5 and 15, and a non-counting 24. Inconsistency! Overall it's well below my global average, so it went surprisingly well.

 

[Fred]

Learned bpaul 7 simul flip, and it's not that hard! I got a 13.37 pb single in the weekly comp, but a bad average (23.22) because I'm inconsistent. Got a 15.39 pb ao5 in the cubers.io comp though. I'm memorizing positive numbers in English and negative numbers in Chinese.

 

[Fred]

My clock was uneven, and I finally figured out it was due to one of the centers having the magnets pushed in more than the other center. So I poked the sticking out magnets in a bit to make the centers match and it's much better now!

 

[Fred]

Got my first sub-10 clock single (9.61), and my most recent ao100 (18.77) is no longer DNF! I've also been doing a lot of slow solves on 3x3 to try to get better at making decisions based on EO, and also two-sided alg recognition.

 

[Fred]

I watched Tymon roast The Cubicle's solves and got inspired to learn some more F2L solutions. More slow solves incoming to get used to them. Here's an interesting one:

 Twizzle link 

 Setup 

z2 R U R' F R' F' R2 U' R'

 Moves 

R U R2' F R F' R U' R'

Three move setup into the upside-down pair case.

 

[Fred]

My non-W/Y 5x5 cstimer session caught up to my W/Y session (mostly white) in number of solves (Yau solves only). Note that these sessions do not include weekly comp results.
Non W/Y:

W/Y:

(Though I also have a 1:42.37 single and 2:00.90 average on W/Y that I did a while back in the weekly comp.) 3x3 is further behind with 24.49 best Ao100 compared to 19.03 for W/Y. I also need about 1000 more solves.

 

[Fred]

Got some new puzzles, with more still shipping. Aosu v7 triple track 4x4 I enjoyed greatly straight out of the box. It's much better than my old MGC 4x4. After spamming a ton of solves, I realized that the inners do feel too heavy, so I'm going to order some weaker magnets to replace the ones that are not glued in, like Lai S Cube

I also got a puppet cube v2, which was much more reasonable to solve than the puppet v1. I could manage to get a layer and then attempt CLLs from various angles. If all of them were blocked, I'd make a different layer. After solving the corners, the edges could be solved as in corners first. I then looked up some algs and finally solved the puppet v1. Such a crazy, stupid puzzle. I was using some RU-only 3-style method, with many algs that were 40 moves long. The thing that made this puzzle even harder is that the tensions were too tight, so there was some catching that I thought blocked off even more moves. And it's pretty challenging to adjust the tensions

I also got a Qiyi smartcube. I mostly got it for alg training, since cstimer lets you practice algs on a virtual cube without needing to scramble in between. I also tried out some full solves and got this 12.12 PB single:

 Twizzle link 

 Setup 

F D L2 R2 D2 B L2 U2 F L2 U2 B2 D2 F2 R' D R' B U' B2 R

 Moves 

z2 y B' D' U' L2 U' L' F U F2 //xxcross
R U' R' //3rd pair
U R' U' R //4th pair
R U R' U R U L' U R' U' L // CLL
U R U R' U' M' U R U' r' //ELL

Like my previous PB single, it was a planned xcross that I turned into an xxcross during the execution.

Coming soon are Dayan FTO and Qiyi mirror blocks.

 

[Fred]

Dayan FTO is here and it is good! I got double PBs in the weekly comp (50.60 single, 1:05.44 avg), beating my old ones by several seconds. It has better corner cutting than the Diansheng, and a wide range of screw depths that are usable, ie loose enough for good corner cutting and tight enough for no popping. The size and weight are a bit larger than would be ideal, but after using the Dayan, the Diansheng feels too small. With my tensions and fast lube, the turning is fast enough for my liking, but it's also easy to slow it down with tighter tensions, slower lube, or the extra springs. I tried the extra springs and I had to loosen the screws to where popping was possible, but there are a couple people in the FTO discord who do like the extra springs.

Qiyi mirror blocks is pretty good too. It's very cheap, so my expectations were lower, but it's magnetic and turns pretty well. I tried and failed to find a way to adjust the tensions, so I ended up stretching the one side that was way too tight by propping my screwdriver under that layer and leaving it overnight. That helped a lot. It's fun because it ought to be easy and familiar since it's just a 3x3, but it's surprisingly hard because the recognition is so different. I ended up using Petrus because it's easier to tell with blockbuilding if I'm solving pieces correctly, and the EO mid solve is easy and helpful.

 

[Fred]

I got really excited after seeing Max Park's name on the competitor list of a comp I just signed up for. Then I looked further and saw Tommy Cherry too, which is also exciting except that there are no blind events or clock. Meanwhile, there's 3x3, 5x5, and 6x6 for Max Park to shine.

I got some extra magnets for modding my Aosu v7 triple track. I swapped out the unglued magnets for weaker magnets and broke a few PBs. The slices are a lot lighter now, but I'll also want to add some weak edge-edge magnets since the slices don't stay together as well in 3x3 stage. Maybe I'll also replace some of the smallest radius slice magnets too if I can figure out how to weaken the glue.