Monday, February 17, 2014

Curvy Rhombohedron Skewb

The first couple solves I just started putting pieces together that looked like they needed to be put together. Eventually a 3-cycle or a twist or two needed to be done to finish up and I drew from my Skewb Family Arsenal of Algorithms (I just made that up).

But then Kevin asked how I solve the Skewb, so I got out the Skewb Cube and solved it. Then the Tetrahedral Skewb. Then the Squished Skewb. Then I intentionally tried to use the same strategy to solve the Curvy Rhombo Skewb.

Documented Scramble #1: DNF
After getting the first 8 pieces solved, leaving only 6 to go, the thing still looked scrambled! After using a setup move and the 3-cycle to place 2 of the 6 pieces, and not being able to see how to undo the simple setup move, I decided that although it is possible to solve in the same way, a more intuitive approach would be better.

Documented Scramble #2:
After solving the yellow layer, including the green and purple triangle center pieces, one twist of the black layer was needed to solve the 3 remaining triangle center pieces. That left the 3 remaining edges out of place and the remaining corner twisted. After using Double-EPS to place the edges instead of the pure 3-cycle, both of the diagonally opposite corner pieces needed twisted. I don't think it is possible to twist them together, so I twisted each one with a triangle using an algorithm that hadn't been used in any of the previous posts' solves. I will not even attempt to type it out.

So it turns out I not only demonstrated the algorithm. I also showed what happens when unexpected results happen, and one way to solve mysteries! :D

Documented Scramble #3:
Very difficult to get the yellow layer this time. The first several times I tried it this way it seemed almost trivial to get a layer solved. But this time, either it was a different scramble, or my mind wasn't in sync with the puzzle as before. This can be a problem when solving intuitively rather than with a systematic strategy. I ended up using Triple-EPS a couple times to swap pieces into place on the yellow layer.

After solving most of the black layer as described in #2, once again I ended up with the 2 small corners twisted. Two lessons to learn. First, if I would use the same steps as I do with other Skewb puzzles, I wouldn't have this problem at the end. So is it worth it to work through a more visually challenging solve at the beginning to make the endgame go more smoothly?

Secondly, it dawned on me that one conclusion I made in #2 above is incorrect. I said "I don't think it is possible to twist them together," because one is in the corner role and the other in the center role (thinking of the Tetrahedral Skewb). But what did I learn in the video above? I can twist 1 corner and 3 triangles, in the unshapeshifted puzzle, and 2 of the triangles are centers! So I can twist a corner and a center together! Yay!

In conclusion... well, is it ever really concluded? Let's just say that after several more solves I have decided that at this point I rather like solving it using the same basic approach taken with the other Skewb Family puzzles. That is, get the 2 small corners and the 6 triangles first. This only takes a few simple twists. Maybe 1 EPS. So with half a dozen twists or less more than half the puzzle is solved! Then 3-cycle the 6 big pieces home.

If a corner needs twisted without disturbing triangles, up twist down replace up twist-the-other-way down go-back works well and is pretty simple. Let's see.

Bottom line though, with this puzzle, as with any puzzle, there is no right or wrong, easy or hard, better or worse, way to solve it. I may decide on a favorite way to solve it tonight, but 6 months from now pick it up and take a whole different approach. I suppose I should demonstrate the Preferred Method applied to this one, so

Documented Scramble #4:

Sunday, February 16, 2014

Squished Skewb Solve

I think the Squished Skewb is great! The shapes you can make with it are fun and interesting. The quality of the puzzle is very good (thank you LanLan). Here is a solve.

In this solve the corners only took a couple twists plus one EPS. None of the corners needed to twist after being placed. That scenario was covered in the Skewb Cube video and works the same with the Squished Cube.

Meier-Halpern Pyramid (Jing's Pyraminx)

In the previous post I showed a solve of the Skewb Cube. The Meier-Halpern Pyramid corners can be solved like the first 4 corners in the Skewb video, with just a few simple twists. If the centers are not solved they can be solved with one EPS. Or, if preferred, the centers can be solved later using a Triple-EPS.

Now the edges may be solved with the same algorithm as the Skewb center squares, but there are easier ways. For example, the EPS can be used to place them if centers are going to be solved last, or the Double-EPS can be used to place them if the centers are already solved. So it just depends on whether you like having the centers and corners all solved first, or not. Why can this simpler method be used here, but not with the Skewb Cube? Because it twists the centers of the Tetrahedron, which doesn't matter, unless orientation of the centers matters to you. It doesn't to me. I ignore the grain of the stickers. If it matters, the longer algorithm demonstrated in the Skewb Cube video works.

Finally, if a pair of edges need flipped, without disturbing the rest of the puzzle ,,,

How I Prefer Solving a Skewb

Now that I've watched it, I have one thing to say. When twisting two corners, they have to need to twist opposite directions. Don't try to use this method to twist two corners in the same direction. It won't work. But you probably knew that already.