Unreal. 5 seconds was not broken too long ago. Crazy to see the 3 second barrier looking like a possibility. Pretty astonishing. Is it even possible without skipping some steps?
Using computers, it has been found that there is a limit of some 21-23 moves to solve any configuration of the cube, so in theory if you knew all the billions++ possible states, you could just do the required 2x steps as fast as possible and then work on motor skills from there, but of course reality is that you can't memorize them all so it is some kind of tradeoff between knowing a lot of patterns and applying the shortest solution you know to take it closer to solved.
All the people solving in seconds aren't even using the same methods, so some will be better at planning while they look at the cube before the start, some will have faster motor skills.
The diameter of the cube group is 20, but speedcubers go for natural/memorized sequences of moves they can execute extremely quickly ("fingertrick") over technically shorter but awkward move sequences that require grip adjustments, have hard to visualize side effects, and so on.
Like you said, none of these folks are deriving God's Algorithm[0] in their heads in the planning stages. I haven't been in the game for a while but I'm pretty sure people are all still doing CFOP[1] but getting better at lookahead, execution, and some of the advanced tricks like x-cross[2], and a bunch of gnarly last-layer tricks[3]. Some of which include memorizing thousands of algorithms.
Really curious to know the guidelines for the state of the unsolved cube. Clearly some configurations are faster to solve than others discounting the speed of the solver.
The scrambles are computer generated and yes you can get lucky cases, which is why they are often taking averages as well.
Also, not everyone will solve every scramble the same way, since there's an element of decision making involved there, and future steps depend on previous ones.
The WCA regulations (which any competition follows) specifies a uniform distribution over all possible states, excepting ones that take 0 or 1 moves to solve.
I was obsessed with Rubik in my youth, and I had a few of those solution tutorial books that told you all the patterns to take to solve the cube from any possible scramble.
My personal best time was 63 seconds.
I owned both official and knock-off cubes, and I found the official cube far superior, merely for reasons of coloring, and also solid construction; it was smoother and it was much less likely to explode in my hands as I cubed at top speed.
I spent inordinate amounts of time in the upstairs bathroom lubricating my cubes with a big tub of Vaseline. My parents never confronted me or even seemed to wonder what I was doing in there; I suppose it was still a bit too early for us to worry about anything morally questionable. But I did go through a lot of Vaseline.
Interestingly enough for a while the knock-off cubes were markedly better than the official one. The official mechanism used to require a quite good alignment to allow turns. The knockoff mechanisms had a spring under the center pieces essentially allowing the cube to "stretch" some, which in turn allows you to force a turn with just about any misalignment up to 40 degrees. Of course these days rubiks official cubes use that same mechanism.
That's Max Park. I'm not a cuber but I recognize him from the excellent short documentary The Speed Cubers, on Netflix. It's nerdy, engaging, and heartwarming — perfect for HN.
Fun short story: I was the owner of the PuzzleProz ebay store in the mid 00's which was, for awhile, the most popular source of "knock off" speed cubes in the US.
I actually sold Max's family the Red speed cube that you can see him using as a kid in the Netflix documentary "The Speed Cubers"
I have that same cube (model / color) on my desk right now as its one of my favorites.
It was exciting to see the cube in the documentary and to be sure I went back through my order history and found the order.
It was really fun to be part of the speed cubing / puzzle cube community during that time.
17 comments
[ 3.0 ms ] story [ 50.6 ms ] threadAll the people solving in seconds aren't even using the same methods, so some will be better at planning while they look at the cube before the start, some will have faster motor skills.
And any position can be solved in as few as 20 moves, aka “God’s number”
https://www.cube20.org/
The site states "20 moves or less" which is stating a maximum.
[0] https://en.wikipedia.org/wiki/God%27s_algorithm
[1] https://jperm.net/3x3/cfop
[2] https://jperm.net/3x3/cfop/cross
[3] https://jperm.net/3x3/cfop/ll
Also, not everyone will solve every scramble the same way, since there's an element of decision making involved there, and future steps depend on previous ones.
https://www.worldcubeassociation.org/regulations/#4b3
My personal best time was 63 seconds.
I owned both official and knock-off cubes, and I found the official cube far superior, merely for reasons of coloring, and also solid construction; it was smoother and it was much less likely to explode in my hands as I cubed at top speed.
I spent inordinate amounts of time in the upstairs bathroom lubricating my cubes with a big tub of Vaseline. My parents never confronted me or even seemed to wonder what I was doing in there; I suppose it was still a bit too early for us to worry about anything morally questionable. But I did go through a lot of Vaseline.
I actually sold Max's family the Red speed cube that you can see him using as a kid in the Netflix documentary "The Speed Cubers"
I have that same cube (model / color) on my desk right now as its one of my favorites.
It was exciting to see the cube in the documentary and to be sure I went back through my order history and found the order.
It was really fun to be part of the speed cubing / puzzle cube community during that time.