[Help Thread] Big Cubes Discussion

[cubeshepherd]

Nitpick: Freeslice is part of a reduction method. ("Reduction" isn't a shorthand for "reduction with one-at-a-time edge pairing"; it literally just means any type of reduction method, usually understood to be with freeslice for F8E + cycles/slice-flip-slice for L4E.)

Yes! You are correct on that and I should have clarified that in the previous post. What I had meant was instead of just focusing on one edge color at a time and reorienting the centers once you are finished with a edge pair, to instead leave the centers misaligned and focus on solving all the edges before to realign the centers. I am not positive if CubeStack_Official is doing that or not, but if he is not then he should practice it. Thank you for letting me know.

 

[Space Cat]

I have a competition in like 2 months, so I have some room to improve. Unfortunately, I don't know exactly what to do about the slowest parts in my solve.

Here's a recent solve (note that I average around 3 minutes and that this solve does not represent my entire average)
Another thing to note is that I use Yau5 and learned from Cyo that inserting the first two f2l pairs during edge pairs is efficient.

2:34.809
First two centers: 22.4
First three edges: 22.5
Last 4 centers: 29.6
all edge pairing + 2 f2l: 1:06
3x3 (2 f2l and LL): 13.4

 

[One Wheel]

I have a competition in like 2 months, so I have some room to improve. Unfortunately, I don't know exactly what to do about the slowest parts in my solve.

Here's a recent solve (note that I average around 3 minutes and that this solve does not represent my entire average)
Another thing to note is that I use Yau5 and learned from Cyo that inserting the first two f2l pairs during edge pairs is efficient.

2:34.809
First two centers: 22.4
First three edges: 22.5
Last 4 centers: 29.6
all edge pairing + 2 f2l: 1:06
3x3 (2 f2l and LL): 13.4

I use redux, so it's not a straight across comparison, but that would be a slightly above average solve for me (I'm assuming this is 5x5?). I don't think there's any way I would get close to that time for 3x3 stage, even with cross and 2 f2l pairs done. I haven't done splits on 5x5 in a while, but from catching the timer out of the corner of my eye I think I run about 18-19 seconds on f2c for a good solve, and another 28-30 on l4c. It's hard to say from splits on a single solve, but my hunch is that you've got room for a few easy seconds on centers, and I wouldn't be surprised if you could save enough on edge pairing to make up for the slower 3x3 stage if you switch to redux.

 

[Space Cat]

(I'm assuming this is 5x5?)

Sorry, I just assumed people knew. That's entirely my fault.

I think I run about 18-19 seconds on f2c for a good solve, and another 28-30 on l4c

Could you describe your steps? I thought centers weren't usually split like that.

 

[One Wheel]

Could you describe your steps? I thought centers weren't usually split like that.

I think what I do is fairly standard reduction:
1. 2 opposite centers
2. Last 4 centers (1, 1 adjacent to the third, last 2 adjacent centers simultaneously)
3. First 8 edges (generally 4 solved and put on the bottom, last 4 on top.)
4. Resolve centers (I'm learning to avoid this step occasionally)
5. Slice-flip-slice back to get L4E down to either solved or edge parity
6. Edge parity (if necessary)
7. Cross
8. F2L
9. OLL (mostly 2-look)
10. PLL

 

[xyzzy]

this solve does not represent my entire average

Well, that does make it harder to give specific advice.

I used to average sub-1:50 with Yau5, although I haven't practised with it in the past few months and my times are up to around ~1:55 now. Comparing our splits, I think your weak points are in F2C and edge pairing; these steps are very much all about tracking pieces while you're solving other pieces, so doing untimed solves might help with these.

(For reference, my splits, averaged over around twenty solves: 14.5 F2C / 45.3 cross + L4C / 43.2 edge pairing + two slots / 11.1 last two slots + LL. I didn't do finer splits like you did because the mere action of splitting the times disrupts lookahead a lot for me.)

 

[UnknownCuber]

Anyone has a edge pairing tip for 5x5? (I use yau5/freeslice hybrid)

 

[One Wheel]

Anyone has a edge pairing tip for 5x5? (I use yau5/freeslice hybrid)

With a question that vague, are you expecting an answer more specific than "practice?" Here you go: practice a lot.

 

[UnknownCuber]

With a question that vague, are you expecting an answer more specific than "practice?" Here you go: practice a lot.

I do yau 5 up to doing all the centres, then freeslice for first four edges (cross edges not counted) , then continue with yau 5 for last 4 edges.
Is that efficient/viable?

 

[xyzzy]

I do yau 5 up to doing all the centres, then freeslice for first four edges (cross edges not counted) , then continue with yau 5 for last 4 edges.
Is that efficient/viable?

But that's just standard Yau5???

Generic advice still applies: don't turn so fast that you can't keep track of any of the pieces and just practise a lot.

 

[Didi Stone]

For 4x4 I use multiple methods... Stadler, redux, not yau, K4, meyer...

I’ve never heard of K4, stadler, or Meyer. What do they involve?

 

[UnknownCuber]

Should I use yau 5, or freeslice? I'm aiming for fairly consistent times (within ±10 seconds)

 

[One Wheel]

Should I use yau 5, or freeslice? I'm aiming for fairly consistent times (within ±10 seconds)

.
Yau5 vs. Freeslice is pretty much a matter of personal taste. There are a lot of guys who like Yau5 for faster 3x3 stage and better look ahead during edge pairing. I prefer redux because of the ability to take advantage of lucky cases during edge pairing.

As far as consistent times, I'm not sure what to tell you. I presume by +/-10 seconds you're probably not talking about a range between 30 and 50 seconds, more likely 3:10-3:30, or something like that. At that speed you're likely pausing a lot, and off the top of my head I would guess that you might lose a fair bit of time during f3e with Yau5 on some solves, so redux might be a little more consistent. It's an unusual goal, and more detail might help. personally I average about 2:40, but any given solve can be anywhere from 2:10 to 3:05.

 

[EvilGnome6]

I don't think the method will have much influence on consistency. All methods will have solves where everything falls into place, solves where nothing falls into place, and solves that are in between.

 

[Sicira]

Does anyone have an alg sheet for yau5-specific l2e? I can't seem to find any algorithms that keep cross and the two f2l pairs solved.

 

[Sicira]

Does anyone have an alg sheet for yau5-specific l2e? I can't seem to find any algorithms that keep cross and the two f2l pairs solved.

Never mind, I found cyo's video

 

[EvilGnome6]

Does anyone have an alg sheet for yau5-specific l2e? I can't seem to find any algorithms that keep cross and the two f2l pairs solved.

This is my 5x5 Edge Pairing alg sheet:

https://docs.google.com/document/d/1MgO30LaIjjZPHwv8N0JHbxHs4qkEp1sp3ySauvDYaAA/edit?usp=sharing

I execute all of them except for the single edge parity by putting the L2E on the F face. I execute the single edge parity with the remaining edge on the U face. I haven't looked at this sheet in a long time. I think there are a few cases I avoid altogether now by orienting the last edge.

 

[Trexrush1]

Does anyone have an alg sheet for yau5-specific l2e? I can't seem to find any algorithms that keep cross and the two f2l pairs solved.

just do F M' (alg) M F'

 

[xyzzy]

just do F M' (alg) M F'

Or [x U : alg], which should be faster. A few of the standard L2E algs won't completely preserve the cross edge, but usually it's only one move off.

 

[mencarikebenaran]

yesterday on one of my 5x5 solve, after the center is done, i got 3 dedges already paired, i super surprised....