There's also a card game called Accordion, where the player moves stacks of cards backwards either one or three positions. The cards must match the stacks they are being moved onto, in either suit or number.
Finally, there's a game called Q-bert, where the objective is to turn every tile to a certain colour, by jumping from point to point. This test is a combination of all three ideas.
The objective of the game is to change every tile to a specific colour, by moving a bullfrog -- a leaping piece -- from tile to tile. Each time the bullfrog lands on a tile, it changes colour, and either increments or decrements based on the move taken to that tile. A move does not need to increment or decrement; in fact, a puzzle may include tiles of only one number, colour being the only restriction to movement. Alternately, a puzzle may include many numbers, with colour being unnecessary.
The puzzle's board is a setup of tiles, in a grid of 10x10 squares. Squares can be 'blocked off' by the creator as needed, replaced by slate. If this is done, the number of raeli tiles required for loading is decreased, and the tiles can be unloaded as needed. When the puzzle is opened for judging, however, all tiles required for the puzzle are consumed, and cannot be removed even after the puzzle's designer changes.
At the beginning, the puzzle's designer designates each square in the grid with a colour and number, from the raeli tiles he or she has loaded. Then, the designer creates a number of 'hopping moves'. These cannot be rotated or reversed, so it is necessary for the moves to include variations for any given type. For a knight's move, for example, eight variations would be necessary, one for each direction. Moves can be any number of spaces, horizontally and/or vertically.
A move has two options besides direction. The first is restriction: a move may require the player to hop to a tile of the same colour or number, or a specific 'different' colour or number. The second is change. As stated above, a move may either increment or decrement the tile the bullfrog lands upon by any given amount, and it can change the colour of that tile to one specific type.
The designer of a puzzle then creates a specific 'target' result using the same grid. This can be viewed at any time during the puzzle for easy reference. The target may ignore either numbers or colours, but has to use one or both. When all squares in the grid (other than slate) reach the target result, the puzzle has been completed.
The moves were varied and strange, but Heket was quickly able to sort through the loops and bounds, the bullfrog carrying but a slender thread to show the path taken. As it lept, the land changed and reformed, the valleys and craters catering to seemingly random whims, and at the end, Heket found a strange pattern in it -- a volcano. Kauket laughed, "See? Even with your 'vast intellect', you yet fall to chaos!" However, the string remained, tightening as Heket took her bullfrog...and, with a roar, the volcano became ensnared, collapsing on itself within seconds. Kauket stared for but a moment before leaving Heket to her own, more gentle laughter.
For this, the Test of the Leaping Bullfrog is named, in honor of Heket's intelligence and cunning.
When designing the test, the student begins with a board of a variable size: Either 5x5, 10x10, or 15x15. Each is set up in an array of raeli tiles, initially black; the building costs include both black and white tiles equal to the number of squares on the board: 25, 100, or 225. Other materials may include boards, slate, or similar items, making the initial setup somewhat similar to a raeli mosaic.
After the board is created, the designer can allot an equal number of tiles for 'other colours'; up to eight other colours may be loaded into the building, thus providing a puzzle of up to 10 colours for the truly difficult. The initial board's design consists of three points:
1. The 'initial setup' of the board is placed. Each square may have a particular colour, from those loaded into the board; they may also have a number set to them initially. The numbers appear either as a particular 'height' of stacked tiles, or simply as a carving. Numbers are entirely optional, intended merely to add further depth to the puzzle.
2. The 'ending setup' of the board is determined. This is changed much like the initial setup, with tiles holding a colour and, optionally, number. When the initial setup is manipulated to match the ending setup, the puzzle is considered solved.
3. The pieces are placed. The designer may place up to six bullfrog figurines, of up to three different colours, onto the board. For 5x5 arrays, this is limited to two bullfrogs; for 10x10, there may be up to four, and for 15x15, it ranges to six. These are the pieces by which the player manipulates the board.
Next, the designer is asked to set up 'moves' for the bullfrogs. This is the variety which allows this test to range from extremely simple to exceedingly complex. Moves incorporate the following features:
1. Initial placement. The move may require that the bullfrog be of a specific colour, or that the bullfrog begin on a tile with specific characteristics, either in 'height' or colour. This is an optional requirement; should the designer agree, the move may allow any initial placement. (Suggestion: Allow the designer to specify an 'always off' option in the building, for interface simplification.)
2. Initial changes. When the bullfrog 'jumps' in the move, the player may alter the tile the bullfrog is on, or any other tiles adjacent to the bullfrog. Changes may be colour and/or 'height'; if height is selected, the player must choose an amount to increment or decrement those tiles. They may not choose a number to directly change the tile to. Numbers range from 0 to 9, with 9 and 6 underlined for simplicity of viewing. All changes to the tiles are aligned to the *north*.
3. The bullfrog moves a certain number of spaces in a given direction, landing on another tile on the board. The designer may, again optionally, specify that a given move *must* land on a specific type of tile, e.g. that the target tile be a certain relative altitude or simple colour. Altitude, again, may not be specified exactly; it can only be relative to the opening one. The necessary colour, however, may be specified exactly.
When the move is created in the interface, a number of additional options are suggested. First is reflection, allowing the player to reflect the move horizontally, vertically, and/or diagonally. (Consider the Knight's move in chess; it can be two up and one left, or two up and one right, or one down and two left...and so on.) This would allow a smaller move list, easier to work with. Second, a suggestion is for only *legal* moves to be listed for any given bullfrog by default, similar to the standard mode of Empty Hand Towers; a 'learning mode' would be accessible to list every move.
4. The bullfrog influences the tile it lands on, and any tiles *surrounding* it, similar to the second point in this list.
As stated above, once the puzzle is manipulated by the player to match the 'ending setup', it is won, and eligible to be voted on.