I can’t even tie my shoe in five seconds. Some people are just amazing, in a good way.
I used to be able to do it in 8-10 minutes. That was 20+ years ago. Got bored with it.
I’ve never solved a Rubik’s cube without taking a screwdriver to it and splitting it apart. The spatial intelligence required to do this in such a short time simply boggles my mind.
I consider myself a fairly intelligent person. Given some time I can fix, or at least diagnose, problems with household appliances and cars. I am rational and can figure out just about every IQ type question, given enough time.
I cannot solve a Rubik’s Cube. I cannot solve it even if you gave me a month. My brain just doesn’t work that way. Yes, some people are astoundingly spatially aware in a way I will never be.
I think the same thing about musicians. To have the ability to play the piano, for example, having one hand do something completely and utterly different from the other hand is astounding to me.
I haven’t touched one for years, so I’ve probably forgotten the algorithm. And I doubt that I could ever have solved one without seeing the solution first. But up to ten years or so ago, I could solve one in about five minutes. I deliberately used a non-optimal method that emphasized visualization over memory, because that was what I wanted to improve.
So my method just had a series of steps that were fairly general, i.e. they would work with several different configurations of the same type, so I only had to learn four or five different sequences. My understanding is that the speed champs who do it in 30 seconds or less have 50 or more sequences at their fingertips, so they can do in one step what I would do in five or six. Plus both they and their cube are physically a lot faster than me and my cube.
But 5 seconds? That has to be a fluke. Sometimes, even with my slow method, the starting configuration was such that just by chance, I could skip a step in the solution – the corner cubies would already be oriented correctly, or something. I think that something like that must have happened here.
I’m not saying he isn’t brilliant and super fast, and I’m not saying the starting position wasn’t sufficiently scrambled, but I’m saying I have to believe that most of the time, it takes him quite a bit longer. I just can’t believe that someone can recognize patterns that fast.
If you have a good memory, you don’t have to be spatially aware to solve it. The number of permutations are actually smaller than most people realize, and a lot of positions are just rotated versions of the common positions in the “fast” algorithms.
There are plenty of computer programs out there that will solve a Rubik’s Cube. They aren’t spatially visualizing anything. They’re just crunching numbers.
Yea, doing it in five seconds is a feat of dexterity, not spatial reasoning. The kid memorizes the algorithms, and than the real test of skill is how fast he can get his fingers to execute them.
I can do a Rubik’s Cube in about three hours. Great speed record, huh? I can’t even get one blue side (I always start with blue) in five seconds, and I know perfectly well how to do that. (The rest is a little iffy.) Actually I probably couldn’t even twist the cube twice in that time.
Watching him do that makes me feel completely inadequate as a human being.
He’s amazing.
Yeah, when I was younger I would take my cube apart and grease it up so I could spin it faster. I won a contest at school and solved the cube in about 35 seconds.
Solving a Rubik’s cube is dead easy, if you get someone to teach you how to do it. It involves zero thinking. All you have to do is memorize a few steps.
Coming up with a solution for solving one? That probably takes a planet-sized brain. But just solving one, if you know how? Your potted plant could probably do it.
I can do one in… I dunno. A couple of minutes, or something? Not sure. Not very long. I learned it from a YouTube video posted by some fourteen-year-old. I would need to brush up now, though. I haven’t done one in ages.
Of course, solving one in five seconds? Yeah, now we’re probably back to planet-sized brain again.
I don’t play the piano, but I do play the guitar some. I started learning it at past age 30. Before that, I was convinced that it was impossible. But, hey presto, I got pretty decent at it pretty quick. It was like a part of my brain that was specialized for the purpuse, and that had been gathering dust for my whole life, suddenly got cranked up. Give that piano a shot, maybe you won’t suck.
Probably not. Haven’t done it myself, but I’d guess that anyone with a M.S.-level grasp of abstract algebra could do the proof. It should be a pretty straightforward application of group theory.
Like I said, planet-sized brain. 
Any given sequence of moves will do three things:
It’ll move some subset of the cubies in a way that you want to move them.
It’ll move another subset of the cubies in a way that you don’t want to move them.
It’ll do absolutely nothing to the remaining subset of cubies.
Solving the cube is just a matter of knowing a few sequences of moves, and doing them in the right order, such that the cubies you’re scrambling at any given step (the second subset) are always ones that are scrambled anyway (so you don’t care if they get more scrambled), and that the cubies you’ve already gotten where you want them are all in the third subset, so they stay where you want them.
For instance, in the method I use, I start by putting all the edge cubies into place on the top and middle layers (this is fairly easy to do; you can probably figure out how to do it yourself). In the process, all of the corner cubies, and the edge cubies on the bottom layer, all go who-knows-where, but I don’t care about that, because I haven’t solved those yet.
Now, I need to solve the bottom-layer edge cubies. I have a sequence of moves that’ll move three edge-cubies without affecting any other edge-cubies (and again, which scrambles corner cubies), except that the three edge-cubies this sequence affects aren’t all on the same face. Two of them are, but the third is on the opposite face. So I have to first use a sequence of moves to put another one of my bottom edge cubies in that position, then use my three-edges-sequence, and then undo the sequence that put that third edge there. There are actually two different sequences I can use to put that third edge there; use the right one (and possibly repeat the process one time), and I’ve got all of edge cubies done.
From there, I just need to solve the corner cubies, and I have yet another sequence of moves that will move three corner cubies around and do absolutely nothing else. If I’m lucky, then the three corner cubies are all in such a position that they’ll be moved where I want them to. More often, one or two of them are going where I want, and the third will be going someplace useless. So again, I have to take some other sequence to put a “bad” corner in that position, use my three-corner move, and then undo that sequence. Repeat this enough times, and eventually there are only three corners out of place, and they are in the places that will be fixed by that three-corner move.
There are, of course, many different variations on this, but they all follow the same pattern. Some people solve the corners first and then use something that moves the edges while leaving the corners intact. Some people use a method that solves the top entirely, then solves the middle layer while leaving the top intact, and then solve the bottom while leaving the top and middle intact. But in all cases, you’re solving some things, while leaving the parts you’ve already solved intact.
Now, for speed-solving, you don’t just want a method that will work, you want a method that uses as few moves as possible (and then you make those moves extremely quickly, but that’s just a matter of manual dexterity and a well-lubricated cube, not of brainpower per se). Part of this is knowing multiple methods and seeing which one would be shortest, and part of it is taking the sequence of moves you’re making, and removing redundant steps in the middle.
For instance, the last step in getting one subset solved might be to turn the top counterclockwise 90 degrees, and the first step in the next subset might be to turn the top clockwise 90 degrees. If you know you’re going to go straight from one to the next, you can just not turn the top at all, and never actually “finish” that previous step, but it doesn’t matter. For this process, I suspect that having a strong verbal memory might even be more important than having a strong spatial sense, since you’re not thinking of the cube as movements in space any more, but just as a string of symbols that can be converted to other strings of symbols according to some rules.
Is there a minimum for how mixed up the cube has to be to determine records
I could solve it even faster if it was just one turn away
See that look on his face when he finishes it? That is surprise.
A great algorithm, a ton of practice, a lot of lube… and yeah, probably quite a bit of luck. Then maybe you finish in seconds. Just like my sex life, really.
I have no doubt that he’s a consistently super-quick solver, though.
Yes. There are official rules for ensuring that the cube is properly scrambled.
This is such a happy video. That kid is cute and all the congrats are very cool to see. I don’t even care how he did it. Brightened my morning!
Back in the get-off-my-lawn heyday in the '80’s, I remember my best was 90 seconds. There was a little book of algorithms published at the time. The one I used required you to get 3 sides by your own logic, and then you just followed the algorithm to bring it the rest of the way home.
There were a few books like that back then. Mine was fairly thin. I memorized the patterns for the last sides and could consistently solve it in under 3 minutes.
Those patterns are long forgotten now.
Good for him! Seems every couple years the record is broken with someone doing it faster than the previous record holder.
More often than not by a teenager.