Skewb Diamond Tutorial

Here she is, in the foreground. Maybe not a girl's best friend, but a Skewb Diamond nonetheless. Behind it from left, as I mentioned here, are the original Skewb, the Skewb Ultimate and three "Skewb-adjacent" puzzles, the Dino Cube, Ivy Cube and Redi Cube.

One of the reasons I'm burning through these 3D twisty puzzle tutorials is because of cases like the Skewb Diamond. It's a puzzle so easy to solve that it tends to get pushed into the second row from the back, ahead of puzzles I haven't mastered yet or don't dare to scramble because they're so hard to turn, I'll never get them solved. And let's be honest, my collection no longer fits on one tray table so it really ends up parked on the edge of a shelf on my living room bookcase.

It tends to get left out of the bag, like a toiletry bag or soft-sided lunchbox, that I take with me on overnight trips, filled with as many puzzles as will fit in it, which is somewhere around 16. Because space is limited on the tray table where I park the puzzles I'm actively working on, it tends to be left out of the group that I scramble at the start of an evening spent solving them. I have to make mindful choices and the Skewb Diamond, while it has its fine points, often doesn't make the cut. And then I come back to it after neglecting it for a while and realize I've forgotten how to solve it, and I can't make heads or tails of whatever notes I took the last time I re-learned it. This has happened multiple times.

So here, once and for all, to help me as much as to help you, here's the clearest I can make this. The Skewb Diamond, a.k.a. the four-axis octahedron, is another octahedral twisty puzzle, like the FTO and the CTO. That is, it's a perfectly symmetrical, 3D shape made up of eight identical, equilateral triangles – well, identical other than being eight different colors. It has the color scheme of my original, stickered FTO (that's face-turning octahedron, in case you don't feel like looking it up). Its sides are cut into four smaller triangles: three corners and one center. Like other puzzles that bear the name of Skewb, each twist moves half of the puzzle with respect to the half being held in place.

Because of the unique symmetry of the octahedron – something true of only this one of the five Platonic solids – pieces from any one side can only mix with pieces from three other sides – the sides that it touches only at the tips of their vertices, not along an edge. So, looking again at the example above, a you can twist that front white corner toward the green side or the blue side, but you have to keep going to the next side over before it can stop. You can't put a white piece on a blue or green side; but you can mix up pieces on the gray and white side, for instance. So, on this octahedron, the white, gray, red and orange sides can interchange pieces, as can the yellow, green, blue and purple sides. So in a way, that cuts in half the complexity of scrambling and solving this puzzle.

It has all the same possible moves as the outer layers of the FTO – no slice moves or wide moves, though. We're talking good old R and R', L and L', U and U', B and B', D and D', F and F', BL and BR and their prime moves. All of which are handy to know when you scramble it. You could use an FTO scrambler for it. Since the Skewb Diamond and I go back farther than when I had an FTO, I've always just pulled up a 3x3x3 cube scrambler and applied the steps as best I could, which means ignoring 2s (as in R2 or U2) because, with triangular faces, those would just end up being prime moves (like R' or U'). But now that I think of it, an FTO scramble would do nicely.

However, during your actual solve you really only need to know four moves, six on the outside. Always holding it so the triangle on top points away from you, here's R:

And R' (arrgh-prime, matey):

Here's U:

And U' (ewe-prime):

Sometimes, but not always, you may need to do an F move:

Or perhaps an F' (eff-prime):

OK, so we're scrambled.

STEP 1: Get all three white corners on the same side, taking care to match up their side colors as well. Don't worry about the white center just yet. So here's a white corner on top. The camera angle doesn't show it, but the white corner in the lower layer also has purple on one side, so it's good to dial in next to the one on top.

You may need to do some creative twisting around the bottom layer to get your corners into position to dial up into place; you may even have to knock one off the top layer if its side colors are misaligned. But here, this blue-white-green corner on the lower layer is good to go.

STEP 2: Solve all other corners. Start by putting the side with white corners at back. (It's not easy to show this in a photo. Trust me, it's back there.) The yellow corners will naturally gravitate toward the layer opposite to white. You'll either find yourself looking at all three yellow corners at front, or just one. If there's just one, rotate the whole puzzle so that one yellow corner is at the upper left corner of the front side:

Then do the algorithm R U' R' U.

Chase this pattern with an F or F' as needed to fix the sides so all their corners are correctly permut(at)ed and orient(at)ed.

STEP 3: Solve all centers. Take stock of your eight sides.

This will show you if there are any pairs of two sides that need to swap centers, like (in the example above) gray vs. red and white vs. orange. If there is at least one pair, there will always be two. It will also show you if there are three sides that need to cycle their centers, like green, blue and yellow.

To do a "two-swap" – actually, this will solve both pairs of sides with swapped centers – put one pair of sides with swapped centers at front (F) and back right (BR), like the red and gray sides shown below; then do R U' R' U three times. (Again, I stress – in case the camera angle is deceptive – that's with the top-side triangle pointing away from you.)

How about an even more awkward camera angle? The first one, below, is an attempt to show all three sides whose centers require a "three-cycle." The second one is an even more dubious attempt to show how you position those three sides with one facing down (D), one back left (BL) and one back right (BR).

That's when you do the algorithm R U' R' U', twice. And this time I want to stress that that last step is U' (ewe-prime); unlike the previous algorithm, the top layer keeps turning counterclockwise instead of alternating back and forth. Also, because this is a 3-cycle, you may have to repeat it to get this final result:

Of course, that's because the centers may need to cycle either clockwise or counterclockwise; but the algorithm will only cycle them (I think) clockwise, as viewed from the back. So, don't give up if it still looks like a three-cycle is needed after you've done R U' R' U' times two. Maybe, just maybe, fourth time's the charm.

I'm not done enjoying my Skewb Diamond, but I think you can see how fair my assessment of it is. It's that puzzle that puts up so little resistance to being solved, if you do those R U' R' U(') patterns with the right sides facing in the right directions, it almost isn't worth the effort it takes to scramble it. Or worth the space on your crowded tray table of puzzles to solve tonight, or in your twisty puzzle overnight bag. It's another one of those warmups to a fine evening of puzzling, maybe. One that (in the model I own) feels interesting in the hands and moves with satisfying, audible clicks unlike any other puzzle, and that delivers its endorphin hits without too much ruckus. I feel I should play with it more often, if for no other reason than to avoid having to relearn the solution each time I pick it up again. And I should take it with me more often when I visit my folks, who absolutely do not understand my interest in this hobby but who might be able to wrap their heads (and hands) around this simple, straightforward puzzle.