Rubik's cubes?

Bob Jewett

Bob Jewett

AZB Osmium Member
Staff member
Gold Member
Silver Member
No, not the member here who goes by a similar name. I'm talking about competitive cube solving. In some sense it is a competitor for cue sports in that it attracts the attention of very competitive young people. I bumped into the cube craze because a new student also competes in cubing. Personally, I haven't touched a cube since the middle 1980s.

It is startling how good people have gotten at it. If you take an average of 10 seconds in a competition (for a 3x3 cube) you won't make make the list of the top 1000, and you have to do one in less than 8 seconds to make the list of top 1000 single solvers.

Here is a list of the top 1000 personal times:

https://www.worldcubeassociation.or...ionId=&years=&show=1000+Persons&single=Single

Note that the association keeps track of great steaming piles of records. Note also how many competitions there are and how international they are.

Here is a kid solving a cube blindfolded (inspect position then blind when turning):

https://www.youtube.com/watch?v=-BX4iyprVi0

Technology: the fancy new cubes have magnets in them so they kind of lock into place when aligned.
 
PRED

PRED

AzB Silver Member
Silver Member
I read somewhere that a machine solved the cube in record time. Here we go-


What is the AI approach to solve randomly shuffled Rubik's Cube?
http://www.guinnessworldrecords.com/news/2016/2/rubiks-cube-robot
4 Answers
Michal Forišek
Michal Forišek, teacher, scientist, competitive programmer
Answered Dec 2, 2016 · Author has 1k answers and 7.2m answer views
There is no AI here.

First of all, the only thing stated in the cited article is that the robot solved the cube in record time. This statement does not mean that the robot used the optimal decisions. It only means that it was fastest in terms of wall clock time.

A huge part of this success is mechanical: their robot was able to rotate the faces of the cube in a way that’s both fast enough and precise enough.

Another significant part of designing such a robot is some basic computer vision: the robot must correctly recognize the colors on the cube in order to be able to solve it.

The actual sequence of rotations used by the robot is almost certainly computed by assembling a sequence of hard-wired patterns that transpose small subsets of positions — just like a human solver would solve the cube. This step does not require (and does not contain) anything worth calling “artificial intelligence”, it’s just a finite sequence of if-then-else statements.
 
Last edited:
Get_A_Grip

Get_A_Grip

Truth Will Set You Free
Silver Member
When I was in college I bought a book that showed how to go about solving the cube by solving various individual problems one at a time. I could usually solve the whole cube between 1 and 2 minutes. I always imagined that there must have been other, more complex patterns and solutions that could be used to speed the process even more. I can't imagine how to go about solving the entire cube in less than 10 seconds though.
 
Michael Andros

Michael Andros

tiny balls, GIANT pockets
Silver Member
Get_A_Grip said:
When I was in college I bought a book that showed how to go about solving the cube by solving various individual problems one at a time. I could usually solve the whole cube between 1 and 2 minutes. I always imagined that there must have been other, more complex patterns and solutions that could be used to speed the process even more. I can't imagine how to go about solving the entire cube in less than 10 seconds though.
Click to expand...

How about less than *5* seconds. The fastest ( human ) time is 4.22.
 
Bob Jewett

Bob Jewett

AZB Osmium Member
Staff member
Gold Member
Silver Member
Michael Andros said:
How about less than *5* seconds. The fastest ( human ) time is 4.22.
Click to expand...
Feliks (the record holder) points out in a video that part of setting a record is happening to get an easy scramble, sort of like making a ball on the break and having a shot on the one eight times in a row at 10 ball.
 
Michael Andros

Michael Andros

tiny balls, GIANT pockets
Silver Member
Bob Jewett said:
Feliks (the record holder) points out in a video that part of setting a record is happening to get an easy scramble, sort of like making a ball on the break and having a shot on the one eight times in a row at 10 ball.
Click to expand...

Makes perfect sense. But everyone in competition has "access" to those same "rolls", yes? Personally, I don't like to even THINK of solving one in under 5 seconds. It gives me a severe headache! :yeah:
 
9BallKY

9BallKY

AzB Silver Member
Silver Member
Get_A_Grip said:
When I was in college I bought a book that showed how to go about solving the cube by solving various individual problems one at a time. I could usually solve the whole cube between 1 and 2 minutes. I always imagined that there must have been other, more complex patterns and solutions that could be used to speed the process even more. I can't imagine how to go about solving the entire cube in less than 10 seconds though.
Click to expand...

I had one of those to and could solve the cube in just under a minute back then, but
no way could I even think about doing it in 10 seconds. In 4 seconds I'd be lucky to
make a single move.
 
BRussell

BRussell

AzB Silver Member
Silver Member
My daughter got into doing it and I started too. My fastest on 3x3 is probably about 30 sec, and it usually takes me about 45 sec. if I’m trying to go fast. We also do large ones (my biggest is 9x9, takes me about 15 min.) and some unusual-shaped ones.

To get faster like those going for speed records you need to memorize a lot more move sequences so you see a pattern and then know exactly what to do.

I always think it’s funny to see people playing with them like they can just figure them out and solve them. They’re impossible to solve just by picking it up and playing with one. Even the people who invented them couldn’t do it until they studied them and wrote down the different algorithms for different patterns. They’re just too difficult. The sub-5-sec. times are amazing.
 
M

MitchAlsup

AzB Silver Member
Silver Member
BRussell said:
I always think it’s funny to see people playing with them like they can just figure them out and solve them. They’re impossible to solve just by picking it up and playing with one.
Click to expand...

No, this is not correct. I got a RC for christmas in the early 1980s.
The first solution took me a bit over 20 hours.
The next solution took only 2 hours.
I have not done this in 20-odd years, but the last time someone handed me a randomized RC I fixed it in about 20 minutes.

I have never read a book on how to do the solutions. I'm not fast, but I did figure it out all by myself.

Even the people who invented them couldn’t do it until they studied them and wrote down the different algorithms for different patterns. They’re just too difficult. The sub-5-sec. times are amazing.
Click to expand...

The times are amazing.

My day job has a component of recognizing large scale patterns and then figuring out what to do about them.
 
iusedtoberich

iusedtoberich

AzB Silver Member
Silver Member
Interesting!

Take a look at this link, that shows the world records for each cube going back about 15 years. Notice, almost every single year, substantial time was shaved. And the difference between 15 years ago and current, is huge. What caused this? Is it really the equipment and that the cube can spin better? Are the players better than before? Is participation higher, and have simply resulted in finding more of the better players? I thought the heyday of this cube was in the 1980's?

LInk to records:
https://www.worldcubeassociation.org/results/regions.php?regionId=&eventId=&years=&history=History
 
Last edited:
iusedtoberich

iusedtoberich

AzB Silver Member
Silver Member
Well, I'll admit I had one as a kid and gave up after messing with it for a while:grin-square::grin-square::grin-square:

I'm not smart enough for this. ha ha.
 
BRussell

BRussell

AzB Silver Member
Silver Member
MitchAlsup said:
No, this is not correct. I got a RC for christmas in the early 1980s.
The first solution took me a bit over 20 hours.
The next solution took only 2 hours.
I have not done this in 20-odd years, but the last time someone handed me a randomized RC I fixed it in about 20 minutes.
.
Click to expand...

I’m sorry but this is a really well-known thing among cubers and a very common topic on cube forums. Everyone says “I solved it once just by looking at it and flipping it” but the odds are way worse than winning the lottery. Whenever people are challenged on it their memory turns out to be wrong.

You could spend weeks, figure out some algorithms, write them all down, and basically invent your own method, like the person who invented it and the others who pioneered solving it. But you can’t just flip around, looking at it, and intuitively solve it (unless you take it apart or take the stickers off). It’s just too hard. I don’t want to say it’s impossible, because someone can win the lottery, but not multiple times...
 
Rubik's Cube

Rubik's Cube

Pool Ball Collector
Silver Member
Best thread ever! :)

I agree with everyone's sentiments regards the sub ten-second times, simply astonishing. Another extraordinary feat I once saw was a fellow who could solve a cube as he juggled with three of them!
 
Bob Jewett

Bob Jewett

AZB Osmium Member
Staff member
Gold Member
Silver Member
BRussell said:
... You could spend weeks, figure out some algorithms, write them all down, and basically invent your own method, like the person who invented it and the others who pioneered solving it. ...
Click to expand...
That's how I solved it, but I don't recall ever writing down the algorithms. Much later I saw a discussion go by and noticed that while my algorithms were sufficient, they were a fraction of the standard algorithms. In the videos I saw there seems to be a whole lingo to describe the types of algorithms.

(Algorithm here means a standard set of rotations that accomplishes a particular change, such rotating a corner to complete a face.)

Edit: here's part of a discussion:

For A perms I need to do an x regrip anyways, so I can combine an x+y into effectively one regrip. So, the tradeoff is either an extra U + U' and an x rotation, or an x+y rotation.​

Kind of like, "I prefer to play that with inside follow if the cloth isn't sliding and then kill on the forth cushion."
 
Last edited:
Bob Jewett

Bob Jewett

AZB Osmium Member
Staff member
Gold Member
Silver Member
Rubik's Cube said:
... Another extraordinary feat I once saw was a fellow who could solve a cube as he juggled with three of them!
Click to expand...
Not nearly as remarkable as the guy on YouTube who memorizes 41 cubes and them solves them blindfolded. He forgets the last rotation on the last cube. :eek:
 
Bob Jewett

Bob Jewett

AZB Osmium Member
Staff member
Gold Member
Silver Member
sjm

sjm

Older and Wiser
Silver Member
God's honest truth. In the mid-1980's, I solved the cube in under 30 seconds several times and could average under 40 seconds, but doing it in under 10 seconds is truly inconceivable unless one were lucky enough to pick up a nearly solved one.

Today's solvers are simply amazing.
 
Michael Andros

Michael Andros

tiny balls, GIANT pockets
Silver Member
Bob Jewett said:
Not nearly as remarkable as the guy on YouTube who memorizes 41 cubes and them solves them blindfolded. He forgets the last rotation on the last cube. :eek:
Click to expand...

Nah, he didn't forget. No more than a good player who dogs the 9 at hill-hiil in the finals "forgot" how to hit it. It was PRESSURE!!! :grin: :yeah:
 
Last edited:
pt109

pt109

WO double hemlock
Silver Member
sjm said:
God's honest truth. In the mid-1980's, I solved the cube in under 30 seconds several times and could average under 40 seconds, but doing it in under 10 seconds is truly inconceivable unless one were lucky enough to pick up a nearly solved one.

Today's solvers are simply amazing.
Click to expand...

We had a pool player who got action in that time period...at three, two, and then one minute.
Too bad I wasn’t aware of you, I could’ve flown you in and robbed everybody...:)
 
You must log in or register to reply here.
Top