The Pyraminx (/ˈpɪrəmɪŋks/) is a regular tetrahedron puzzle in the style of Rubik's Cube. It was made and patented by Uwe Mèffert after the original 3 layered Rubik's Cube by Ernő Rubik, and introduced by Tomy Toys of Japan (then the 3rd largest toy company in the world) in 1981.[1]
The Pyraminx was first conceived by Mèffert in 1970. He did nothing with his design until 1981 when he applied for a patent on 27/03 (EP0042695 on 12/30/81) and brought it to Hong Kong for production. Uwe is fond of saying had it not been for Ernő Rubik's invention of the cube, his Pyraminx would have never been produced.[citation needed] Somewhat earlier (for 40 days) in the Soviet Union, the chief technologist of the Kishinev Tractor Plant, Alexander Alexandrovich Ordynets, filed his application for an invention (patent SU980739 dated 12/15/1982, with the filing date 02/18/81), because[clarification needed] of that, in Russia many people call puzzle "Молдавская пирамидка" (Moldavian pyramid).
The Pyraminx is a puzzle in the shape of a regular tetrahedron, divided into 4 axial pieces, 6 edge pieces, and 4 trivial tips. It can be twisted along its cuts to permute its pieces. The axial pieces are octahedral in shape, although this is not immediately obvious, and can only rotate around the axis they are attached to. The 6 edge pieces can be freely permuted. The trivial tips are so called because they can be twisted independently of all other pieces, making them trivial to place in solved position. Meffert also produces a similar puzzle called the Tetraminx, which is the same as the Pyraminx except that the trivial tips are removed, turning the puzzle into a truncated tetrahedron.
The purpose of the Pyraminx is to scramble the colors, and then restore them to their original configuration.
The 4 tips can be easily rotated to line up with the axial piece they are respectively attached to, and the axial pieces are also easily rotated so that their colors line up with each other. This leaves only the 6 edge pieces as a real challenge to the puzzle. They can be solved by repeatedly applying two 4-twist sequences, which are mirror-image versions of each other. These sequences permute 3 edge pieces at a time and change their orientation differently, so that a combination of both sequences is sufficient to solve the puzzle. However, more efficient solutions (requiring a smaller total number of twists) are generally available (see below).
The twist of any axial piece is independent of the other three, as is the case with the tips. The six edges can be placed in 6!/2 positions and flipped in 25 ways, accounting for parity. Multiplying this by the 38 factor for the axial pieces gives 75,582,720 possible positions. However, setting the trivial tips to the right positions reduces the possibilities to 933,120, which is also the number of possible patterns on the Tetraminx. Setting the axial pieces as well reduces the figure to only 11,520, making this a rather simple puzzle to solve.
The maximum number of twists required to solve the Pyraminx is 11. There are 933,120 different positions (disregarding the trivial rotation of the tips), a number that is sufficiently small to allow a computer search for optimal solutions. The table below summarizes the result of such a search, stating the number p of positions that require n twists to solve the Pyraminx:[2]
| n | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| p | 1 | 8 | 48 | 288 | 1728 | 9896 | 51808 | 220111 | 480467 | 166276 | 2457 | 32 |
The world record fastest Pyraminx solve is 0.73 seconds, set by Simon Kellum of United States on 21 December 2023 at Middleton Meetup Thursday 2023. The world record fastest average of five Pyraminx solves (excluding fastest and slowest) is 1.45 seconds, set by Ezra Shere from the United States on 23rd September 2023 at Washtenaw Fast 'n Late Fall in Ann Arbor, Michigan.[3]
| Number | Name | Fastest solve | Competition |
|---|---|---|---|
| 1. |  Simon Kellum |
0.73s | Middleton Meetup Thursday 2023
|
| 2. | Elijah Brown |
0.75s | Berkeley Winter A 2023
|
| 3. |  Dominik Górny |
0.91s | Byczy Cube Race 2018
|
| 4. | Jakub Hibszer |
0.92s | Cube4fun in Gorlice 2023
|
| 5. |  Jasper Murray |
0.96s | Sydney Summer 2023
|
| Number | Name | Fastest average | Competition | Times |
|---|---|---|---|---|
| 1. | Ezra Shere
|
1.45s | Washtenaw Fast 'n Late Fall 2023
|
(1.82), 1.42, 1.42, 1.50, (1.38) |
| 2. |  Lingkun Jiang
|
1.46s | Chongqing Open 2023
|
(1.40), 1.40, 1.58, (1.59), 1.41 |
| 3. | Jasper Murray
|
1.47s | Lyttleton Spring 2023
|
1.64, 1.60, (1.82), (1.08), 1.18 |
| 4. |  Sebastian Lee
|
1.51s | Melbourne Summer 2024
|
(1.19), 1.23, 1.55, 1.76, (2.70) |
| 5. | Michael Nielsen
|
1.53s | Virginia Still-Open Winter 2024
|
1.56, 1.38, (2.18), (1.29), 1.64 |
| 5. | Simon Kellum
|
1.53s | Side Events Minnesota 2023
|
1.54, 1.42, (1.34), (2.61), 1.64 |
There are many methods for solving a Pyraminx. They can be split up into two main groups.
1) V First Methods - In these methods, two or three edges are solved first, and a set of algorithms, also called LL (last layer) algorithms, are used to solve the remainder of the puzzle.
2) Top First Methods- In these methods, three edges around a center piece are solved first, and the remainder of the puzzle is solved using a set of algorithms.
Common V first methods-
a) Layer by Layer - In this method, a face with all edges permuted is solved, and then the remaining puzzle is solved by a single algorithm from a set of 5.
b) Algorithmic L4E and Intuitive L4E - L4E or last 4 edges is somewhat similar to Layer by Layer. The only difference is that only two edges are solved around three centers. Both of these methods solve the last four edges in the same step, hence the name. The difference is that Intuitive L4E requires a lot of visualization and "intuition" to solve the last four edges while algorithmic L4E uses algorithms. Algorithmic L4E is generally used more at higher levels, although there are very fast Intuitive L4E users. It is also easy to transition between Intuitive L4E and Algorithmic L4E.
Common top first methods-
a) One Flip - This method uses two edges around one center solved and the third edge flipped. There are a total of six cases after this step, for which algorithms are memorized and executed. The third step involves using a common set of algorithms for all top first methods, also called Keyhole last layer, which involves 5 algorithms, four of them being the mirrors of each other.
b) Keyhole - This method uses two edges in the right place around one center, and the third edge placed elsewhere on the puzzle. The centers of the fourth color are then solved using the slot formed by the non-permuted edge. The last step is solved using Keyhole last layer algorithms.
c) OKA - In this method, one edge is oriented around two edges in the wrong place, but one of the edges that is in the wrong place belongs to the block itself. The last edge is found on the bottom layer, and a very simple algorithm is executed to get it in the right place, followed by keyhole last layer algorithms.
Some other common top first methods are WO and Nutella.
Many Pyraminx speedsolvers learn several methods, particularly top-first methods, and use the method that is best for the given solve.[6]
There is no consensus among pyraminx speedsolvers regarding whether top-first or v-first methods are faster, although v-first is more common in the present day.
There are several variations of the puzzle. The simplest, Tetraminx, is equivalent to the (3x) Pyraminx but without the tips (see photo), resembling a truncated tetrahedron. There also exist "higher-order" versions, such as the 4x Master Pyraminx (see photos) and the 5x Professor's Pyraminx.
The Master Pyraminx has 4 layers and 16 triangles-per-face (compared to 3 layers and 9 triangles-per-face of the original), and is based on the Skewb Diamond mechanism. This version has about 2.6817 × 1015 combinations.[7][8] The Master Pyraminx has
In summary, the Master Pyraminx has 30 "manipulable" pieces. However, like the original, 8 of the pieces (the tips and middle axials) are fixed in position (relative to each other) and can only be rotated in place. Also, the 4 centers are fixed in position and can only rotate (like the Rubik's Cube). So there are only 18 (30-8-4) "truly movable" pieces; since this is 10% fewer than the 20 "truly movable" pieces of the Rubik's Cube, it should be no surprise that the Master Pyraminx has about 10,000-times fewer combinations than a Rubik's Cube (43 quintilion in the short scale or 43 trilion in the long scale). The Master Pyraminx can be solved in numerous ways: one is layer by layer like the original one or reducing it to a Jing pyraminx.[9]
