Message #3025

From: Ray Zhao <>
Subject: Magic 120-cell solved, yay
Date: Thu, 04 Sep 2014 18:03:32 -0400

Hi all,

I got pretty bored one day so I decided to retry solving the 120-cell for
the heck of it. My previous attempts got me to layer 4, but I decided to
start from scratch again because my 4D CFOP, 120-cell edition, improved a
little since then (2011). =P Method’s here (for 3^4), though I might edit
it in the next few days or weeks.

It was finally solved on Aug 30, 2014. Started Aug 13, 2014. Including the
1k move scramble, it’s just over 20k moves.

Of course, the fact that I didn’t solve it a long time ago meant that I
couldn’t break the age record of 15 since I’m 16 now, but at least I got to
beat the shortest. I also beat myself; by the time I was on layer 4, the
solution seemed to have become almost 40% more efficient than the 2011
attempt though I probably have to check that number. Half of the reason for
the efficiency was the use of RKT, or limiting moves to only two cells,
pretending one of them is a megaminx (a bit hard to explain. the wiki link
mentions it as <RK, A->). The other half would be because I strived to
learn how the 3c and 4c turns work, and pressed undo many times just to
find the smoothest insert of an f2l pair.

Because CFOP uses pretty intuitive F2L, what’s macros? Roice’s program
works nicely, and I have to admit it probably has the easiest and most
comfortable controls, but that’s probably because I used it for weeks
straight. Its similar colours have got me into trouble multiple times,
though, but how do you pick an arrangement with 120 contrasting colours so
that you don’t get a case with a 3c grey-grey-grey piece? (I hated that)
Perhaps even a variation of 4-colour rule would work here. ^_~

As for during the solve, the beginning was pretty frustrating since pieces
would seem to be everywhere; most of the time was spent finding the piece
since I’d rotate the puzzle, turn on the layer, and forget the exact
position that the piece was in.

By the time layer 4 was half-solved, things felt extremely restricting
since some turns would have the same effect as exposing a solved slot on
the 3^3. In fact, I screwed up multiple times and had to refix multiple
pairs because I had taken them out by accident. That by itself probably
added around 2k moves.

The worst part was the very end, when there seemed to be only one flipped
edge; the other one was actually wedged between two greys in layer 5. @_@
At the same time, it was kind of exciting to be able to find a solution to
that case; filled up a little piece of scrap paper with the steps and all.

So that’s pretty much all about the solve. .-. Roice encouraged me to post
so yeah. Not sure what puzzles I’ll go for next (aka improve on),
considering that Uni registration is this year (w00t) though duoprisms and
penteracts sound nice. :)