Published on
I started this post back in June actually, and was planning for it to be specifically about the differences in how I do yau compared to others, but it started getting really close to the end of the month so it’s all going together.
Cubing Stuff
Since I’ve gotten back into going to speedcubing competitions, I’ve started messing around with learning some more algs and methods. Specifically the idea of switching from reduction to yau5 for 5x5+ and learning a bunch of ZBLLs. I’m never going to be the fastest in the world, I missed that boat a long time ago lol, but I would like to get faster still. I’m a competitive person, so I’d like to maybe get back to where I can get to finals and do well at competitions again. 3x3 and 5x5 are my favorite events and these seem like decent steps to getting fast that I can experiment with. I have a long term goal now to shoot for because there’s a big competition I’m probably going to next year!
World Championships 2025 and US Nationals 2026
I wasn’t glued to the screen for world championships like I thought I was going to be. I was hanging out with friends for most of the weekend, so I only caught a few solves here and there for worlds. I was actually planning on going to worlds because my friends and I were looking for a boys trip and Seattle seemed really cool. We didn’t end up going because I was super dumb and didn’t understand the way the qualification times worked, so we ditched Seattle. I thought everyone had to have a sub-7 3x3 average to compete, but I didn’t realize that they were going to have more groups where they relax the qualification times each time. I thought maybe they were limiting the competition to only the very best. It wasn’t until the spots had already filled up and we had planned our Toronto trip that I finally realized how it worked. Oh well, Toronto was still really cool and watching it was super neat, I was glued to the final round of 3x3 and 5x5. Tymon getting a world record in the final round to win 5x5 was insane, and Yiheng having to clutch the final solve to become world championship was so incredibly hype. Congrats to them both!
Even though I didn’t get to go to worlds, I was happy to watch it. But what really excited me, and I didn’t believe at the beginning, was the US Nationals 2026 announcement. It’s going to be in Raleigh, North Carolina! It’s been forever since I’ve even seen a competition in Raleigh, and I’m so much closer to Raleigh than Charlotte where most of the NC competitions are held these days. Out of the 17 competitions I’ve been to I’ve only been to 3 competitions in the Raleigh-Durham area. I’ll be able to wake up late and drive to the competition without rushing! I’m so excited. Since this was announced exactly a year before it’s going to take place, so I have a long time to decide how and what I want to practice. My goals are to get a sub-10 average and maybe a sub-8 single for 3x3 with ease at Nationals or before, and to maybe be able to get a sub-1 minute 5x5 solve or average at Nationals too. I feel like these are realistic goals because I can get sub-10 averages at home easy, and I usually get at least 1 sub-8 single in an average of 100 for 3x3. I don’t do as well at competitions (like most people), so it’s hard to get good averages sometimes. 5x5 is a little harder, as of right now I’ve only gotten 1 sub-1 5x5 single, and my best average of 5 at home is around 1:03. But these will be fun challenges to see how I can improve!
Switching to yau5?
It’s obviously possible for me to get faster with reduction, but I’ve used yau on 4x4 for at least a decade now so I’m comfortable with the method, and it seems like a lot of people prefer it for bigger cubes as well as 4x4. I’ve tried about 50-60 solves total for yau5, but something seems awkward. First 3 cross edge pairing is probably the most awkward, but centers don’t feel great either. People make the distinction between yau and yau5, so I decided to go look up a few videos about yau5 to see if people had some tips. After a few videos I realized I do yau pretty differently than how a lot of people do yau.
I’ve known that I do yau differently compared to some people for a long time. During the 2nd step where you solve the cross edges, most people will solve 3 correctly in relation to each other so when they place the last one into the empty slot the cross is fully solved. I don’t know exactly where I learned yau to teach me this, but I solve the first 2 cross edges correctly, and then I solve the 3rd one into the incorrect spot. It doesn’t matter which edge it is and it doesn’t matter what order I put them in, I only need 2 in correct relation to each other. The reason I do this comes after center pairing. After 3 cross edges are solved I put them on the left with the U layer having the free space. Then I solve the last 4 centers. Some people will solve the last cross edge and put it on the bottom right side of the cube, because you don’t technically have to use R moves. I find this doesn’t work as well for me because I use Rw moves to do centers instead of only using inner slices, and I don’t want to have to do R/R’ or M/M’ moves to keep the edge on the bottom. I see the benefit in doing it, but it doesn’t work as well for me.
Once all of the centers are done this is another place I’ve seen the method differ. I’ve heard of people going ahead and pairing up the rest of the edges before solving the cross. I don’t think this is quite as popular though, but it does keep your F face free to do some unique stuff with edge pairing. Instead, I (and a lot of other people) will move on to solving the last cross edge. I’ll usually keep my already solved cross edges on the left hand side like I had them when I solved my centers. I’ll match up the edge using Rw and U moves, but I won’t undo the moves to fix the centers back yet. I’ll first switch out the incorrectly placed edge with the newly created edge using a U or a U’ move, this will correctly place the newest created edge in relation to the 2 cross edges that were placed in correct relation to each other from earlier. This takes the edge that was incorrectly placed earlier and puts it into the M layer, and temporarily messes up the centers on the U layer because they are still askew from pairing up the 4th cross edge. Then I’ll do either an L or L’ move to swap the empty slot to the U layer, I can then do a U or U’ move to move the previously incorrectly placed edge into the right spot where it’s in correct relation with the rest of the cross edges now. It sounds sort of complicated, I promise it isn’t. I’ve always done it this way. It makes placing the last cross edge easier in my mind, and it saves me moves.
Other people will pair up the edge, solve the mismatched centers, and then put the edge directly into the right spot. This way I guess has better look ahead because you typically probably flip the cross to bottom and then pair up the last edge, so you have a better view of the cube when you go to start doing edges. Mine seems like maybe it saves some moves, but I do lose out on that view of the rest of the cube so I’m probably slightly slower to start edge pairing in comparison.
After I’ve placed the last edge I will then put my cross on bottom. Because I haven’t fixed the centers yet from pairing up the last cross edge, I can go ahead and begin edge pairing. Way back in the day, I would intentionally have the centers be 2 turns off (Uw2) so I could do 6-2 edge pairing. I would place my edges where I could do either a Uw or Uw’ and have 3 edges solved, then I would move all the solved edges to the U layer while setting up for the next slice. Then I would simply have to do a slice-flip-slice to fix the last 2 edges (once they were mirrored to each other, I could do (Uw) (R U R’) (F R’ F’ R) (Uw’)). That lead to some weird cases where it’s basically too many things going on, and I’m losing out on the benefits of doing it that way. Nowadays I do 3-2-3 edge pairing, which is kind of the standard, I don’t see too many people differing from that. Then I do normal 3x3 except for parity.
In either case if I decide to use yau for 5x5 or not, I should learn my last 2 edge algorithms finally. I know a few of the good ones, but I’ve never committed to learning them. I know I’m losing at least a couple of seconds on some solves to those stupid algorithms. That should be my main focus, but here I am contemplating if I should switch methods entirely.
ZBLL
“Jackson”, you say, “you just said you haven’t learned 5x5 L2E, you say you’re gonna learn 493 algorithms instead? What’s wrong with you?”.
How about you leave me alone.
On a serious note, yes I’m learning some ZBLLs, and yes it is silly but I’m having fun. It is a ginormous amount of algorithms, but there’s ways to significantly reduce the number. For a start, if you don’t plan on learning the sune and antisune cases because it’s almost faster to do sune + PLL, the number drops to 328. Then if you know COLL you already know another chunk for when those cases happen to solve edge permutation as well as corner permutation, so it gets closer to 300. It’s still a lot of algorithms, but there’s a few tricks like that to get the number smaller and smaller. Mirrors can also knock you down by a lot, but that’s not the best approach sometimes and you should probably learn the better algorithms.
I’ve decided to begin with 2GLL U and T. 2GLL is a popular subset of ZBLL where the corners are already permuted, so you’re doing edge permutation and corner orientation. They can all be solved with 2 gen algs that only contain R and U moves. 2GLL is only 84 cases, so it’s a good intro for people who want to start.
I’ve started learning 2GLL in the past a few times, but the algorithms never stuck. It was always too daunting, and the algorithms didn’t make sense or weren’t easy to memorize. This time around I’m learning them pretty fast for what I have in the past, and it’s been really fun. I’ve been using (this trainer)[https://bestsiteever.ru/zbll] and it’s great. I can select which algs I want to practice, I can save them in sets, and it has really good algs on the site in my opinion. Right now I’ve learned about 20 algorithms in about 2 weeks. I’ve been learning them in pairs. Take the 2GLL U set for example, I’ll learn the case that has the Ua Perm edge swap on the side without the headlights along with the one that has the Ub perm. This way has been working for me really well so far. I’m planning on learning the last 3 2GLL U algs in the next few days. I’ve already learned the 2GLL T set and I periodically practice both to make sure I still retain them since I’m learning them pretty quick.
It’s been nice getting some of the cases in actual solves. My recognition in real solves is a lot slower than in the specific practice. I’m normally like “Hey this looks familiar. Do I know this case? Yes, let’s try it.” then I slowly do the algorithm and end up with a worst time than if I had just done (C)OLL+PLL. That will have to come with more time and practice. Practicing ZBLL in real solves kind of stinks because the cases you don’t learn don’t come up that often. I use CFOP and will try to force a cross on top anyway, but the chances of getting one of the 24 cases I know out of the 493 possible ones that there are is kind of annoying. When I know a full set then they’ll show up way more often. I thought about trying to learn ZBLS first, but a bunch of people on forums kind of agree that it isn’t a great idea. It seems like learning ZBLL is putting the cart in front of the horse, but I’m gonna trust the experts on this one.
I don’t know if learning ZBLL will save me any time in the long run, recognition is looking like it might be tougher than I was expecting, but it’s a cool thing to say you know ZBLL. Plus I’m having fun, so that’s what matters.