# How is this even possible? (Rubik's cube video)

If you watch the video, there are people in the background.

In fact, this guy has a number of videos on youtube, and they are all basically the same, except for the number of cubes. But if you look at the activity behind the guy, there is no way it can be reversed. People cannot walk backward and look like they are walking forward when you reverse the film. It’s not possible.

With that said, if you are asking me if he is cheating in another way, well. I guess anything is possible. However, I son’t see it. Not only does he put a mask on, but his girlfriend (I assume) puts a placard between the guy’s head and the cubes, and it would seem that she is blocking his line of sight.

Again, I am not saying there isn’t some massive trick to this, and I don’t know this guy. I stumbled onto this video and thought. I’d share it.

I just don’t see how this can be done with regular memory tricks. Some folks have given some decent guesses at how it might be possible, but I really don’t know what the answer is.

To me, this is almost like a “rainman” type of trick, where an autistic savant can retain and memorize things that a normal person cannot. However, this guy doesn’t strike me as having autism.

For those of you that have watched the video, there are a few periods within each stretch where he is going through the cubes where he stops looking at the cubes and does elaborate hand-arm waving gyrations that I would guess is part of his memorization routine. It’s odd to watch, but he does it every time in every video I’ve watched. If anyone else has seen this, maybe you can give us some insight into what he’s doing there.

All I have learned about is guy is that he is Polish, he claims to hold the world record in what they call “multi-blind” cube solving, and he seems to be involved with a website, which I wont list here, but ends in the .pl suffix, which. I assume stands for Poland.

I did watch the video, and it is most certainly possible. It would obviously take practice, but it could be done. In fact, if I were to try it, I would speed up the playback so that the jerky motions of the sped up actors would hide any otherwise unnatural looking motions. Oh wait…

You can solve any Rubik’s cube by memorizing a sequence of algorithms based on starting position. He probably examined the cubes, found what he was looking for, set them up so that when he grabbed them they would be in position to execute the solve, and ran the algorithms.

I’ve not spent any time memorizing the patterns, but my understanding is that once you have them down solving a cube isn’t any problem at all. Maybe I should, although since I’m already married and old I’m not concerned with impressing the hot geek contingent.

You are looking for a conspiracy here that just doesn’t exist.

I didn’t put the link in the last post, but he also appeared on a program that looked like “Poland Has Talent!” Or some similar type show.

His trick on this show was to have 25 or 30 cubes in front of him, all numbered, and each judge got to pick at random whatever cube they wanted him to solve.

What was really cool about this one is that he first made a design on the sides of the cube, and then he solved it.

So the first cube, he made a “cross” on each side, the second cube he made a “T” and I can’t remember what the third one was.

There is no way these are faked by running the film backward. So, of all the answers given in this thread, your answer is the one i can say is incorrect.

This isn’t exactly true.

While it is true that you can solve the cube with algorithms (and that’s how everyone speed solves these things), i know of no algorithm that tells a person how to solve a cube from scratch from a random starting position.

Here’s the thing… If you have one cube in front of you, that one cube has 26 individual cubies

6 centers
12 edges
8 corners

So let’s say you want to solve your one cube… You have to not only memorize the starting position of each cubie on the cube (which can be done), but when you begin solving it, you have to know in your head where each cubie will be after the algorithm.

This part is extremely challenging. However, again, for one cube, I think with practice it can be done.

But to be able to line up 100 cubes and solve 95 of them blows my mind. His memorization technique, whatever he is using, should be marketed.
Of anyone actually has a published method for blind solving, I’d love to take a look at it. Clearly there is a trick to this that I’m not understanding. If it is just pure memorization, fine. I can’t believe it, but fine. My brain couldn’t memorize the starting position for 100 cubes, but that’s just me. And solving a rubiks cube isn’t exactly hard when you have a billion youtube tutorials that show you how to solve it.

However, when I solve the cube, I have to look at the position of the cubies on the last layer and the algorithm I pick is based on a specific starting position. Since i use my own algorithms, i have a lot less to memorize, but I still have to look at the last layer before I start. I cannot see how i would be able to keep track of the location of each cubie that was still in the unsolved position as i work through the solution.

Perhaps his solving method permits an easier way of knowing where each cubie goes.
Ah, well. I’m impressed. This is an amazing skill, although I’m not exactly sure what you can parley it into for future earning potential.

For anyone who thinks he is doing this backwards, this should put that theory to rest.
Poland Has Talent

This is worth watching because he only does three cubes out of the 25? Cubes on the table. And not only does he solve them, but he solves them into a design before actually solving the cube.

Clearly this guy has the ability to know where each cubie is after he memorizes the starting position. He can follow that in his mind. I find this ability amazing.

Probably because i misplace my car keys on a regular basis.

Since I’m the only one suggested this, I assume this is directed at me. You need to re-read my posts (#7, #20, and #23).

Your original question was “how is this possible”. I offered one feasible explanation. But I concluded that post by saying I didn’t really think it was the case in this example.

My second post simply questioned how you could be so sure it wasn’t fraud. I also re-stated that I didn’t believe he was cheating.

My third post was responding to your claim that it was “impossible” to fake a video in this way. I simply wanted to make the point that it was possible. I did neglect to reiterate my position that I didn’t think he was faking it though.

At no point did I believe (or state) that this was a video being played backwards.

First result for a google search of “How to solve a rubik’s cube blindfolded

From the introduction of the linked article:

It describes a method for reducing a scrambled cube to a string of numbers which dictate what algorithms to apply and in what order.

This technique describes a method by which someone can recall a long string of numbers.

So if you can convert a scramble cube into a string of numbers (skill #1), recall that string of numbers (skill #2) and then apply the proper cube solving algorithm for that number (skill #3), you could do it too.

Hope this is more what you were looking for, I apologize for my inadvertent hijack re: backward videos.

Thanks for the link, (i guess I could have googled the same thing. Goes to show you that you should actually search for something instead of assuming it wouldn’t be out there, like i just did)

Oh, and you didn’t hijack the thread at all, so no apology necessary.

As you said, i was asking for possible ways this thing was done. And I’ve seen folks have their 3 or 4 year old kid screw up a cube, run it backwards and post “4 year old solves Rubik’s cube!”

The only reason I said your suggestion wasn’t possible in this case was because it is clear (to me, anyway) that when the video is run in normal speed, everyone has normal movements, not ones that look like they are moving backward.

That’s all. Your suggestion was fine for the thread. I only responded to your comment “how are you positive this wasn’t done?” (Or something in that vein). My answer was and is that in all of the clips I’ve watched, I’ve seen no evidence that the video was run backward. Am I positive? Well, I wasn’t there when they filmed it, but I’d be willing to make a wager on it. I guess anything is possible, but i’d give it a less than .001 % chance.

But that’s just me.

And seriously, thanks for the link. Not sure WHY i didn’t google it, but i didn’t. :smack: that’s embarrassing. But I am going to read up on what the links say and see if I can learn to do this with one cube. If I can, then i would assume i can learn more than one at a time with practice.

I don’t think this part is particularly impressive. You said you were self taught to solve a cube (I consider that impressive, BTW) but my method of solving (the “layer” method, learned from the instructions that came with the cube) involves making a T on all sides first. In fact, I don’t think I could solve it without first making the T.

If I’m not mistaken, other methods could result in a “cross” on each side first (basically, solve the edges first, then the corners). Once the edges are placed, there are algorithms that will only move the corners without disrupting the edges you have already solved. So he is versed in multiple techniques of solving, which is not at all surprising.

How do you solve it, if not by making recognizable patterns on the faces first?

Well, based on what I’ve read so far in the link you provided, I’m starting to think this may be a bit easier than I initially thought. Don’t get me wrong, I am a long way from solving even one blindfolded, but there seems to be a couple of clear methods and strategies that have been published over the years. This guy clearly has an amazing memory, but I don’t now believe he is some sort of savant.

The “T” on all sides was impressive to me because I have never seen that from a blind solve. Same with the cross. Before I got your link, I was under the impression he had memorized each cube’s cubie layout. I am no longer sure this is necessary. As Airman siggested, there are a number of algorithms that this problem can be broken down into. Based on what i have read, there seems to be much more flexibility in how you attack a cube from memory. I am not exactly clear yet on how this will work, but you are not tied to one color for the top, and another for the face of the cube, for example.

Making the “T” or the cross aren’t hard tricks with your eyes open, but with them shut, I would think they would be difficult, especially if you were memorizing the starting locations that are important on each cube.

Let me try to explain why i thought this was a hard thing to do.

Lets say the starting position of the cube is called X, the T configuration is Y and the solved state is Z.

I can see perhaps understanding how to get from X to Z, and then turning the cube to the Y position. THIS should be straightforward, because going from Y to Z is shown to be only a few moves. However, he goes from X to Y first, and then to Z, which would further imply that he is aware of each piece’s location at all times.

However, it may not be as hard as that. I’ll know more when i work through that paper.

The way i solve the puzzle is a layer method. Top, middle and bottom. I don’t make a cross or any other pattern on the top. When i solve it, i just know where each top piece goes based on the colors on the sides. I know many books teach the cross first method, but I never learned that from a book, so it never occurred to me to start with a cross. I usually start with the white face, and when I see a white cubie, it immediately goes in its proper location.

The second layer is solved with a couple of algorithms I came up with that basically require you to dismantle the first layer temporarily to build the second layer. There is a method referred to as Fredrich, or F2L (which stands for first 2 layers), which many speed cubers use because it permits the solving of the first two layers at the same time. I have just started to try and figure this out because it will shave a lot of time of my current solution time.

The third layer I am able to solve by placing the edges in their correct spots first, and then i finish by placing the corners in their correct spots.

I’ve looked at a number of solutions, and it seems the top-down method is fairly common, since it is the most intuitive for most people. But my algorithms are a bit more flexible since I don’t require each case to be accounted for. Many cases are exactly the same, you just need to know what to look for.

One of my algorithms is longer than a similar one i’ve found on-line, but I continue to use mine because it makes more sense to me. Most of the algorithms are similar in that they perform the same function (like cycle three edge pieces) but my move sequence might be different.

I’m sure this isn’t exactly clear, but i hope it answers your question.

One of the things I am hoping to get from the blind solve paper you linked to is a better solution to the cube regardless of whether or not I’m blind-folded.

One of the things i noticed in the video of this guy is that he uses the center layer a lot. I NEVER use the center layer for solving. The blind-fold method also seems to use the center layer quite a bit to flip edge pieces, so I think there may be a link to what he is doing and what this paper is outlining.

To be continued…

I would say he:

[ol]
[li]Has very visual memory[/li][li]Has practiced the ‘language’ of solving the cube until he can think in that language - thus, he’s able to solve the cubes ‘fluently’[/li][li]Has practiced the whole thing until he can ‘do it with his eyes shut’ - this idiom is often treated as an empty or figurative phrase, but it can be literally true for some people.[/li][/ol]

I stumbled across this thread and thought I’d give you some pointers. If you don’t want this magic trick spoiled then don’t read any further.

The 12 edges and 8 corners are solved seperately. Pre-memorised algorithms known as conjugates and commutators allow pieces to be “shot” to their target locations without affecting other pieces too much. Essentially he is quickly memorising where pieces need to go and then using algorithms to move them into place as fast as his memory recall and finger execution will allow.

There are various blind solving methods with the simplest ones moving one piece at a time (e.g. Old Pochman). Maskow uses BH for corners and TuRBo for edges aiming to do two pieces at a time as efficiently as possible:
http://www.speedsolving.com/wiki/index.php/Blindsolving_Methods

There are a number of methods for memorisation:
http://www.speedsolving.com/wiki/index.php/BLD_Memorization

Most people use “letter pairs” (creating words and images) but Maskow uses something different. He made a video of the images he uses for memorisation.