What are your records for the 5x5x5 cube?

I want to work on 5x5x5 speedsolving. Specifically, I'm looking for a sub-3 average (eventually). My current average is about 3:45 on a good day, so this should be more or less feasible.

On the other hand, I can help out anyone who wants to get sub-5.

I have three big questions to those who think they can help:
- Rubik, Eastsheen, or Studio cube? If Eastsheen, how do I get it to not lock up so much? If Studio, where can I find one?
- How do you lubricate it well enough? [I've heard silicone on top of Cube Lube is really good, but I don't have any more Cube Lube. Anyone know where I can get it?]
- How do I improve recognition?

About recognition, I've been trying something called a 'theoretical' speedsolve, where you can start and stop the timer whenever you want, but you can only move while the timer is going. Best time is around 2:30, so I have something like a minute of lost recognition time in total. Try it - it's very interesting.

Thank you for your consideration.

Why not a 10+ option. I've done it in the 8 minute range before but most of the time it takes a few minutes longer.

As of right not, I am barely sub-3. I typically don't do 5x5x5 averages because they take up so much time, but I know that my times are going down. I can usually do centers in around 50-60 seconds, so I'm not too worried about that right now. However, edges take me way too long. I'm wlorking on learning all of the last 4 tredges algs listed on bigcubes.com, so I expect my times to drop somewhat once I do that. Final 3x3x3 solve usually takes me about 30 seconds. I use a Rubik's 5x5x5 by the way. At first, they are very tight and hard to turn, but they become good after a long time. I've had mine for around a year and a half, so its pretty good.

-Alexei

I'm sure I put that in. I don't know why it didn't allow it... is there any way to edit polls (I can't find a way to)? Is there a limit on the number of options you can have?



Mine too, but there are so many parts I've been having trouble adequately lubing it.

....like 20 minutes with like bigcubes.com....

lol

I'm afraid to speedsolve my 5x5. I broke a center corner piece off accidently while doing a slow slove (luckily cubesmith supplies those easily broken off pieces).

With that said, the fastest I've solved one is 11:17.14
I doubt i'll get much faster unless I risk breaking another piece again.

Whenever I've done mine, I've dont a bit at a time, throughout the course of a day. All using my own methods, except the difficult paritys.

My girlfriend can solve professor supercube faster than I can solve a regular professor.

I'm up near 10 minutes, it's weird though I'm in the low 2 min range for revenge.

Huh... I've
- always used a reduce-to-3x3x3 method (see below)
- never broken a piece on a 5x5x5 cube (although I did break one on a 4x4x4 once or twice)

Do any of you guys use a 'cage' method (centers last) or a layers method? For me, it seems like it's slow + a lot of algorithms to learn...

By the way, if you're worried about breaking the cube while speedsolving, you can always use a computer cube. There are a few floating around the 'net and I made one of my own (but I have yet to release it).

Cage can be really fast and you don't need any algs if you understand commutators. I'm not very good at it because I've started using my own method for big cubes, but Per gets some nice times with it...

I understand commutators pretty well, but I don't think of them as especially fast, except possibly in blindfold solving. I use commutators when there are only a few centers or edges messed up so I don't have to resolve the whole cube...

Is cage Per's primary method? If so... I guess I'll have to do some more experimenting!

I'm just starting to attack the 5x5x5 again. Had gotten sub-5's when I worked on it before. This time however, I am working on a different solution set to try to eliminate most searching time (looking for certain pieces) Unfortunately I almost always end up with a parity problem (when two edges need to be swapped). And I only know of one way to fix it, which destroys most of my previous work. Does any have an alg that will just swap the two edges? The alg I do know is

(RrU2)*5

Hopefully when I get this last piece figured, I'll be able to knock my time down pretty quickly.

I think you might be looking for the 4x4x4 PLL parity alg, or maybe the OLL parity alg... Other than that, I'm blanking out. Do you mean two "edges" or two "wings"?

You probably want the 4x4x4 orientation-parity algorithm. It swaps the two edge-wings on UF (works on 4x4x4 and 5x5x5 equally well):
r2 B2 U2 l U2 r' U2 r U2 F2 r F2 l' B2 r2

About 10 minutes, or even more, that's because I want to avoid to broke my cube, or some cubie pop out!
Speedcubing for me is only with 3x3x3 cube, for all other items I like the pleasue in solving and twisting my puzzle, and so I'm sure not to broke it

i do sub 3, usually have been averaging around 2:40-2:45, haven't been concentrating on it though, i've been drilling the 3x3 for the last month or 2.

That is the only way I can solve it. I do all of the corners first, then the edges and then the centers last. It's not the fastest, but I only run into the "edge parity" about half of the time and I have a modified solution for that. It's my "finishing" move for the edges. What takes the longest is putting the centers to their right positions.

Total time for me is around 15 minutes. I'm not a speed cuber though, I just do it for fun.

Don't laugh but my fastest time for a 3x3x3 is 1:10...

skeneegee, that's pretty embarassing....

Anyways, I am in the 3:00-3:30 range. Never really faster or slower.

Centers: 55
Edges: 1:45
3x3x3: 35
(approx.)

I am SO BAD at edges. I have no look-ahead.


Oh by the way you do cage method use commutators. It's like two algorithms when it's really zero.

You're bad at edges? It takes me about 3 minutes on those alone. 6 minute average for me.

Then again I put them in one at a time. I figured out how to solve it on my own though. I just haven't bothered to learn the bigcubes method for edges.

Edges takes me a while too... what I do these days is take an edgecenter, find the edgewings, bring them into the center layer and pair them one by one, then take the tredge out and get a new edgecenter (while looking for its edgewings). It's slow but maybe 10 seconds faster than what I was doing before.

I wonder if you can pair up multiple edges at a time... using a 4x4x4-type approach, for example, you could pair up edgewings, and then the rest of the edges is just pairing centers with wing-pairs, which is pretty fast (you can do a three-cycle of centers with (Rr)(Ll)'U2(Ll)(Rr)' or something). I have to try this...

i do something similar but i find my next center edge to work on next to the last piece of the tredge i'm finishing, so when i bring it out i can add one of the wings right away, and then go get the other, again noticing the center edge and looking for his buddies

I kinda solve like per only it's less of a cage, it's more of a box as the centres are filled in before i go commutator style, but it's pretty much the same. I used to use per's method but I'm awful at doing centre commutators fast. Also, I solve the last layer in 5-6 algs on average on the 5x5x5 and 3-4 on the 4x4x4 on average, thats before pairing up and depending on parity.

Ok, that's enough for one night.

Totally, you can do chain-solving with 5x5x5 edges, all though its alot more difficult. Man I really need take out my eastsheen 5x5x5 and practice some more... gotta beat that pathetic record of like 8:50 hehe

I have not timed myself with the 5x5 because I can't solve it completely
without looking at a sheet of paper. For the Eastsheen 5x5 do you just
loosen the screws a bit?

The thing about Eastsheen cubes is that they move very fluidly, but you can't cut corners in your turns. If you loosen the screws, you'll have the puzzle more loose (so it will move even more fluidly) but you still can't cut corners, and if you try you'll be much more likely to have a pop. Personally, I'd keep the tension, and tighten it a bit once it's broken in...

So should I lube it or not?

I ONLY do cage method! But i haven't been "big-cubing" since EC in Paris last year, at least not seriously. I need new cubes.

Commutators are fast. I can do a centers 3-cycle on a good day in about 2 secs max (8 turns). Commutators are generally slow if the cubing style is "triggering" not "wrist-action"

-Per

Wings is outer edges on a 5x5x5. It's not common to call the 4x4x4 edges wing edges (because it has only one kind of edges in the first place!). Edges can also be called "dedges" (4x4x4) or "tredges" (5x5x5). Derived from the words d(ouble)/tr(iple) obviously. Calling the edges on a 6x6x6 olympicube quedges is maybe stretching it a bit far

Chris' : r2 B2 U2 l U2 r' U2 r U2 F2 r F2 l' B2 r2 or
Frederick's : r' U2 l F2 l' F2 r2 U2 r U2 r' U2 F2 r2 F2

Both have no side-effect and work well on both 4x4x4 and 5x5x5. However, conjugated versions may have a side-effect on centers

-Per

Exactly what do you do, Per? I agree that commutators are fast, but how do you do them fast enough for use in a speed solution? Your records are near the top of the list, and maybe your method has even more potential...

Hi

i use a variety of commutators for various purposes so i cannot really give a fully fledged answer. I will list a few. As i mentioned already these algs can be very fast if u are a typical wrist-action-cuber and not relying mostly on clever fingertricks to speed up your algs for you

First layer : no commutators used normally
Middle layers edges: R' {u} R [U] R' {u'} R [U'] (*)

(* u can be replaced with e' or d' to make up "similar" 3-cycles working on different edges. One can also use a combination of these u/e'/d' to create block cycles or multi-cycles. These are useful for solving more edges at the same time or remove "bad" edges from the middle layer. Smart tricks here work miracles for your times!!
U can be replaced with U', this creates inverse cycels!!)

Last layer edges(just few of those i use-all are conjugates)
r2 l D2 l' U l D2 l' U' r2
R2 l D' l' U2 l D l' U2 R2
(Ff) R' d' R U2 R' d R U2 (Ff)'
F' R' u' R U2 R' u R U2 F ("PLL" alg)

All these have simple variations. Recognition is the key. You must see fast what 3 edges you want to permute (cycle). 2 of them will be in same "orbital" while the third is in different orbital and is moved from U to D (the conjugation move!!)

Centers (adjacent faces):
r' u r U2 r' u' r U2
r' e' r U2 r' e r U2

Centers (opposite faces)
(Ff)' l u l' U' l u' l' U (Ff)

if the 3-cycle is written (1,2,3) it's all about making up P Q P' Q' such that P moves 1 to 2. Q is a single turn of a layer such that 2 (taken out by P) returns to 3, not back to 1 again. That would actually cancel ALL moves!! You end up with P P'=I (identity).

This looks very throretical but is surprisingly easy with a little practice. Recognition (and quickly seeing the setup move for conjugation) is what makes it fast in a solve!!

-Per

In your three-cycles, you always solve two pieces at a time, right?

I don't know if you already use this algorithm (for centers), but you didn't list it... it's my favorite at the moment.
r U* l' U*' r' U* l' U*' (where * is either clockwise or counterclockwise)

Does the first layer involve doing the center as well? When do you do corners (especially the last layer corners)?

Do you happen to know the approximate amount of time you spend on each step (first layer / middle edges / LL edges / centers)?

chain solving 5x5 edges? i'd like to see what you're talking about, can you demonstrate a solid way to do this? i've never been able to make it work sucessfully everytime, i can get some easy tredges if i try but the wasted time hasn't been worth it so far