Older cubers discussions

[mark49152]

I need to learn what the algorithms look/feel like as I am voicing them in my head to do them so they only come out as fast as I can say words in my head.

Drilling algs is a good idea and helps develop speed and dexterity. Just repeat each alg over and over. Eventually they become muscle memory, and you get the opposite problem. My algs flow fluidly and automatically from muscle memory, but as soon as I think about what I'm doing, I'm lost, and have no idea how to recover! For some algs, like say R-perms, there's no way I could even do them step-by-step and write it down. The only way I can do it is without thinking. Lookahead or thinkahead actually comes in pretty useful as deliberate distraction, because to stop myself thinking about what I'm doing with the alg, I will instead think "what colours did I see when I started this" and try to predict which way to AUF at the end . I still use 4-look on big cubes because I can't do the algs as fluidly and am more prone to catching and getting lost.

I like memoing corners first because if there is parity, I swap UB and UL in memo to avoid it.

That's an interesting idea, but isn't it prone to mistakes? Certainly it would be nice to avoid an annoying PLL parity as the final step

When you don't have a success what is the most probable cause? Mine is often center execution I might look for that thread you were talking about, I'm improving my system to execute always the same way also. Will tell you if it works ok for me.

Can't answer because both my full 4BLD attempts were successes . I had plenty of failures when practising the individual steps though. With centres, I forgot to switch A/C, or shot to the same location twice, or just did the wrong setups for some pieces. For edges I would sometimes mix up the S/W, A/Q and C/I pieces or forget to undo the setups for C/K/W, but my most common mistake was to mix up the stickers on each edge when tracing. So for example, I would shoot to J, but then on the J dedge look at the upper rather than the lower sticker/wing to continue tracing.

Ollie's thread is here.

I'm getting excited to have my first sub 20 solve. I had a lucky 25 the other day. I remember my first lucky 59 just a few weeks ago! Fun stuff!

My first sub-20 was a 19.70 PLL skip on 17 Jan 2014 when I'd been cubing about 16 months and my PB ao50 was around 28.5. I didn't get another one until 5 Feb 2014 when out of nowhere came a full-step 18.13. That was pretty special

 

[h2f]

That's an interesting idea, but isn't it prone to mistakes? Certainly it would be nice to avoid an annoying PLL parity as the final step

Noah Arturs solves parity that way. In fact, you dont avoid parity, but you solve it this way add one extra target to the edges I guess. I dont know if it's better than switching UB/UL alg and doing last corner with OP.

 

[mark49152]

Noah Arturs solves parity that way. In fact, you dont avoid parity, but you solve it this way add one extra target to the edges I guess. I dont know if it's better than switching UB/UL alg and doing last corner with OP.

I like it. Presumably there must also be a parity alg that would swap the two OP corners without swapping any edges. I wonder if any good algs exist for that.

 

[h2f]

The way I solve parity looks like this. I solve corners first, and when I got parity I do an extra target UBR to last 3-cycle. This swaps UBL and UBR. After solving edges I do M2 y L2 Tperm L2. Done.

 

[mark49152]

I like it. Presumably there must also be a parity alg that would swap the two OP corners without swapping any edges. I wonder if any good algs exist for that.

https://www.speedsolving.com/wiki/index.php/4x4x4_Parity_Algorithms#Opposite.2FDiagonal

The alg labelled FB14 seems pretty good, with the first R cancelled. Not sure if it's any better than a Y-perm and adjacent parity fix though.

The way I fix parity today is to solve all corners OP as usual then x' L2 B' [PLL parity] B L2

 

[Jason Green]

then
I may be coming to FW in a couple of weeks, my grandad is not very well... Is there a group of you there at all?

If there are any groups I have not found them. Maybe I should do as post specifically for that... Maybe you can help me find them.

Sorry about your granddad.

 

[h2f]

https://www.speedsolving.com/wiki/index.php/4x4x4_Parity_Algorithms#Opposite.2FDiagonal

The alg labelled FB14 seems pretty good, with the first R cancelled. Not sure if it's any better than a Y-perm and adjacent parity fix though.

The way I fix parity today is to solve all corners OP as usual then x' L2 B' [PLL parity] B L2

I've lost myself if we are talking about parity in 3bld or 4bld.

But if we talk about 4bld Robert Yau found very nice r2 parity alg: https://www.speedsolving.com/forum/...D-Discussion&p=1057684&viewfull=1#post1057684

 

[mark49152]

I've lost myself if we are talking about parity in 3bld or 4bld.

But if we talk about 4bld Robert Yau found very nice r2 parity alg: https://www.speedsolving.com/forum/...D-Discussion&p=1057684&viewfull=1#post1057684

We're talking about fixing the edge swap that results from doing an odd number of corner swaps in 4BLD. The alg I pointed to is a pure corner swap that does not affect edges.

Rob's r2 parity alg is very nice.

 

[h2f]

We're talking about fixing the edge swap that results from doing an odd number of corner swaps in 4BLD. The alg I pointed to is a pure corner swap that does not affect edges.

I see. The idea of using this alg is very interesting.

Rob's r2 parity alg is very nice.

I use some of his algs and all of them are very nice to fingertrick.

 

[mark49152]

Here's my first foray into corner comms, for anyone interested. Twelve comms for the price of one!

Cycles A to (M, N, O or P) to (I, U or S).

1. Set up first target to P location.
2. Set up second target to A location (buffer).
3. B' R2 B - easy as M2-ish
4. Undo A setup
5. B' R2 B - again
6. Undo P setup

Should also work for A to (L, H, T) to I but I haven't tried it.

 

[Logiqx]

LOL, that was awesome to read.. I literally was laughing out loud. I am glad to see you are improving that fast Jason. I do not know how long it took for my first sub 20 but it was more than a year.

My first sub-20 took just under 13 months... somewhere between 2036 and 2087 timed solves.

 

[Jason Green]

My first sub-20 took just under 13 months... somewhere between 2036 and 2087 timed solves.

It will be hard to say how long I've been cubing. If you read my intro post, I learned Petrus several years ago. I could solve in a couple of minutes or so, but did not care about speed past that. I started wanting to speed solve a little over two months ago. I was so out of practice I had to do review before I could even do Petrus once. Still I'm sure that previous experience helped the learning curve some, even though it feels SOOO much simpler to solve now!

 

[MarcelP]

Okay, I have a challenge for all of you. I enjoy watching cube video's of people I know. So here is the challenge. If you think you average xx seconds, try to make an Ao12 video with achieving that result. I just did my first 12 solves of the day (with many days without cubing), and tried to solve as slow as possible without getting real bad times. The result is a pretty steady average with a standard deviation of 1.1. That is quite good for me. This average sounds quite alright as my global average btw.

Spoiler: Ao12 of 19.79

Time List:
1. 21.07 F L2 F2 L2 D2 U2 B U2 F2 U2 F D' R' B2 L' U B F U L' D'
2. 19.34 L2 B2 U B2 D' F2 U F2 D2 F U R' F2 D' L' B2 U' B R2 D'
3. 19.56 B2 L2 F2 L2 U2 B2 D2 F2 L' D2 U2 B U' B2 R2 B2 L' B L
4. 19.96 L2 U' F' L U2 L' D B L U F2 D2 F2 D' R2 L2 D2 B2 U' R2 L
5. 21.49 B' R2 D' L2 D F2 L2 F2 D' U2 L2 D F' L D' B' F2 R' F' L' D'
6. 17.26 L F2 U L F2 B' L2 D R2 U2 R2 F R2 F2 D2 L2 D2 F2 U'
7. 17.11 U F2 L' F2 D2 L R2 F2 L' B2 F2 R2 D' R2 F' D' L2 D B' L2
8. 21.34 L B R2 B D2 F' U2 B2 U2 R2 L F2 D2 U' B' L' D' L' F'
9. 19.19 R U2 L2 B2 R F2 U2 L' F2 R2 F2 U' L B F L D B F R'
10. 19.51 F2 L2 R2 D' U2 R2 F2 D2 U F2 R2 B U2 L' U R F2 D2 B2 R D2
11. 22.57 D' B' L2 F2 R2 F' D2 F' U2 F' D2 U R' D L2 B2 D L' U2 L2
12. 19.18 L' U' F2 D' R2 U' B2 U' L2 R2 B2 D2 F L2 D' R U2 R2 D L' F

 

[Schmidt]

I won't bore you with that! My avg is around 25s
5 cubes of 2,3 & pyra each week is sadly not enough to improve.

 

[MarcelP]

I won't bore you with that! My avg is around 25s
5 cubes of 2,3 & pyra each week is sadly not enough to improve.

I wish you would get back into speedcube practice. It was really nice motivation for me when you where faster than me.

BTW, I would also love seeing just an Ao5 (although it is much easier to film a good Ao5 )..

 

[Logiqx]

Ok. Just done a sub-20 Ao12 from scratch (rather than do a big session and pick the best bits). I'll upload shortly...

Funny how much harder it is when you know you're recording and doing exactly 12 solves. Similar feeling to a competition.

Edit: You beat me by 0.05s. lol

 

[MarcelP]

Edit: You beat me by 0.05s. lol

Cool.

 

[muchacho]

I'll do an ao12 and upload it later, I'm hoping to average 40 seconds.

 

[Jason Green]

I'll do an ao12 and upload it later, I'm hoping to average 40 seconds.

I'll do one also hopefully tonight. Ao12 if it's not too long! Sub 40 is my only goal, or else I'll do it again.

 

[Logiqx]

Ok... it' uploading but predicting 40 minutes remaining. I've added a few easter eggs to make it a little more interesting!

It seems we have a sub-40 challenge later today.