Physical 3^4 Solved!

[Logan Maciejewski]

On 7/19/2024, I completed the first ever solve of the physical 3^4. It took nearly 10 hours with many pops along the way, but after correcting the errors made, I solved it. I then bought the puzzle and used it to develop my own method that is better optimized for physical solving. A few solves later, I cut my time down to just 56 minutes (with a few pops and mistakes during the solve). Better yet, this solve was recorded, and is now on my YouTube channel for anybody to watch. Hopefully there will be more solves in the future, as more are produced.

Link to the video for those interested: https://youtu.be/nTSYJjxtg4Q?si=OA0bN2mZw_Xk8LfK

[Marc Ringuette]

Sweet!   Awesome!   Fun!    Good stuff,  Logan!

These questions and suggestions are not just for Logan, but also for Grant (I presume this is one of Grant's puzzles or similar?) and the other folks working with the physical 3^4.   They're based on the kind of process I went through 6 years ago with the physical 2^4 as we were figuring it out.   I hope it's a bit useful.


1.   Mappings and moves

I'd be very interested in seeing some videos that would help us understand the mapping from physical to virtual, and what moves you use.   In particular, are there any moves you are using that do NOT correspond to a simple twist on the MC4D virtual 3^4?   This would not necessarily be a terrible thing -- on the physical 2^4, for instance, we found an important class of moves (Iy and related) that are easy on the physical puzzle but would require a multi-move (but legal) macro on MC4D.  I found those to be in "good taste", but some people may not have, and might distinguish a strict versus a less strict solve.

2. Scrambling

Scrambling is worth some consideration, too.   Back in the day (this was never in wide use) I asked Michael Gottlieb to host this 2^4 scrambler that I created, on his qqtimer page:
    https://mzrg.com/qqtimer/index%20-%202222.htm
which corresponds to this way of scrambling the physical 2^4
    https://www.youtube.com/watch?v=ymnvMzypUp8&list=PLNqYnY8yQrGOa9v24g7jJpDYh8H3z-msT&index=15

This didn't catch on; I don't know how people are scrambling 2^4's these days.   My purpose was to help people to avoid sketchy scrambles that fail to visit the full space of puzzle states even when scrambling a Very Long Time.   For instance, I don't think gyro-less scrambles can reach a monoflip position, without an equivalent way (such as the Iy move) to access those states.

3.  Side-by-side comparisons

The "gold standard" for a virtual-physical correspondence is a side-by-side video with the same solve being done on MC4D or other virtual program.   I'm not suggesting that you try to do that for the whole solve, but perhaps somebody could make a few videos showing a catalog of moves and gyros and algs and their MC4D equivalents?

Here's one of my proudest achievements in studying the physical 2^4, back at the dawn of time (June 2018).    Yes, this crappy old video really does give me a warm fuzzy sense of pride.   I was able to duplicate my monoflip algorithm (using semi-kosher sub-puzzle moves called ROIL Zero) on the physical 2^4 on the virtual MC4D 2^4 using some macros, showing it to be kosher in a stronger sense than previously.

    https://www.youtube.com/watch?v=k6ZSu0xOPbQ&list=PLNqYnY8yQrGOa9v24g7jJpDYh8H3z-msT&index=28

Note 1:  the "ROIL Zero" move set that I use in the video never did catch on; it allows "sub puzzle" moves on the R and L sub-cubes subject to parity constraints (you must use sequences of these moves that end with zero parity if you want MC4D to be able to follow along).    Although this video proves that it's kosher in some sense, the stricter move sets became the standard.    It's a matter of taste, and I bow to the consensus.

Note 2:  the link above is to video #28's spot in a playlist of all my 4D cubing videos from back in the day.   The next video in the list, #29, describes the Iy macro, i.e, a natural physical puzzle move that is a bit tricky, but possible, to solve on the virtual puzzle.   It is also an alternative to a gyro:  if you have access to the Iy and Iz moves, you can access all the puzzle states without any gyros.   This is the basis for my scrambling method earlier.    There are likely some similar moves on the physical 3^4, and this approach, side-by-side comparison via MC4D macros, is how I would propose making such a move "kosher", or at least more kosher.

Do y'all agree that this kind of side by side documentation is more or less the gold standard, if it can be achieved?   If so, there is a whole list of useful tasks that we can line up for a diligent nerd or three, involving a physical 3^4 and a copy of MC4D.   :)


I'm not willing to do the Discord thing, so if people could cross-post any relevant discussions and YouTube videos here, that would be perfect.


Cheers and congrats to Logan on a big first!   Even if it merits an asterisk later, like much of my early stuff, when we figure out how strict we want to be about our move sets.   :)  :)


Marc R.

On 7/29/2024 8:18 PM, Logan Maciejewski wrote:

On 7/19/2024, I completed the first ever solve of the physical 3^4. It took nearly 10 hours with many pops along the way, but after correcting the errors made, I solved it. I then bought the puzzle and used it to develop my own method that is better optimized for physical solving. A few solves later, I cut my time down to just 56 minutes (with a few pops and mistakes during the solve). Better yet, this solve was recorded, and is now on my YouTube channel for anybody to watch. Hopefully there will be more solves in the future, as more are produced.

Link to the video for those interested: https://youtu.be/nTSYJjxtg4Q?si=OA0bN2mZw_Xk8LfK

--
You received this message because you are subscribed to the Google Groups "hypercubing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hypercubing...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hypercubing/CAPWiAZD5iqjC1zkDfJ5Srn%2BDiYSATxw%3DF6bnK3C15GtEAVgmpg%40mail.gmail.com.

[Logan Maciejewski]

1. I do not use any moves in my solves that do not have a one-to-one correspondence to a move on the virtual puzzle. I do, however, do a few things that would not necessarily be allowed on the physical 2^4, the most prominent of these being the "half gyro." Since the gyro on the 3^4 requires manually reorienting pieces on both target cells, it is possible to only gyro one cell. Since this puts the puzzle in an "in-between" state, the cell can either be ungyroed or the other cell on the axis can be gyroed, after a move of course. All of this still only corresponds to one move, which is why I use it. I also do "slide moves" on the outside, which correspond to a Uz2, where the standard outer double move on 2^4 would correspond to a Uy2. Lastly, I do Iy2/Iz2 moves on occasion.

I hadn't thought of creating a video showing the correspondence, so thanks for the idea. I'll get that out sometime in the near future.

2. As far as I can tell, the scrambling method in your video is the same as what's commonly used today. I found that the same scrambling technique works on the 3^4, with a double endcap stack instead of a single one. It's also necessary to do a secondary axis gyro (though I don't think it's necessary to reorient the 3c pieces) in order to scramble all the 2c pieces.

3. I may actually do a step-by-step series of videos showing a full solve side-by side with the virtual puzzle. I was crazy enough to solve the puzzle that people thought was too inconvenient to solve, so might as well continue with that.

To view this discussion on the web visit https://groups.google.com/d/msgid/hypercubing/067bd9c5-116b-45a4-8a39-7c31d6e94a63%40solarmirror.com.