Tbh, Chess has an update every few years. There's that one goblin chess game where the rules differ, you have that interdimensional time travelling one, and a frick ton of fairy pieces. Hell, I'm making a Chess card game; there's a lot of chess updates if you know where to look
true future versions of this should have "solving the cube" be a wining move. I would suggest the cube starting with a checkerboard pattern on all sides, then chess masters and "Rubikscube" masters can battel it out
I gotta say, having the pieces take the longest route possible feels like an over the top flex on the opponent. (IE the bishop going all the way around just to kill the queen, the squid just zooming around in the color,etc) This game, if developed further, would definitely be extremely complicated… and definitely should have ways to mark which tiles are being attacked. I would never be able to win otherwise @~@
At 2:45 the knight could technically also move down one square onto red and then onto the adjecent green square followed by moving in that direction onto the next green square. In regular chess it doesn't matter wether you allow the knight to move two squares in one direction and then one square in any orthogonal direction or first moving one square then two squares. However, in the chess on the surface of a cube it does seem to make a difference which exact type of movement you allow. That being said, the third version of defining a knight move: first going one square in any of the four cardinal directions followed by one diagonal move does seem to be entirely covered by the first two mentioned versions.
I think a reasonable definition is "move three spaces through two axes", which collapses to one axis getting one move and the other axis getting two moves (although it allows more freedom when one axis might, for example, change because of the one move through the other axis)
Also, even with the "two in one direction, then 1 orthogonal" definition, it should have one more possible destination by moving two into green and then one orthogonally into red
I was initially mildly disappointed by how the knight interacted with the corner, as there were other ways an L shape can be folded and laid around the corner. I also wondered how a bishop would handle a direct hit on the corner. As that case was not in the video, pulled up the website myself. the answer is that it just stops at the corner. But this gives some logic that also explains the knight's movement as well. When unfolding the cube always have it be a cross centered on the side the piece is currently on. The Bishop would be moving off the board if it continued straight into the corner, and it blocks any double edge crossing paths the knight would otherwise be able to make, allowing only the ones shown.
Alternatively, you could decide the bishop moves by picking a first direction and a second direction perpendicular to the first. The bishop can land on any space where the number of steps in the two directions is equal. A cornered bishop could pick 1:down 2:right. 1d1r would put it in the right face's near corner, 7d7r would put it in the right face's far corner. In this way, I believe it is justified to make a cornered bishop able to travel along any of the center diagonals of the 3 adjacent faces.
Yeah, interactions at the corners where the meaning of "diagonal" becomes muddied was what I was curious about starting the vid, and I was kinda bummed to see how the knight resolved it.
This is a really fun concept. One criticism I have is that I think the white and black squares need marked edges or arrows indicating what is the opposing side at the start of the game. This way pawns aren't omnidirectional AND can do their two-long initial move. I hope to see this playable as an actual multiplayer thing at some point instead of a solo chess toy at some point.
What happens when the pawns are moved around by the hand rotating slices of the cube? I don't think there's any good way of keeping track of their direction here
@@squiddler7731 Hmm, I guess the alternate way to go about it is borrow a page out of shogi's book and have the pieces have an arrow shape to indicate directionality. Interestingly you could theoretically have diagonally oriented pawns, where their normal move is along a diagonal, and then capture along the grid.
I'd have the centre of edge face be the equivalent to the classic opposite edge. Have some sort of crown symbol on the black and white tiles representing a destination square the pawns must move toward as close to orthogonally as possible. This would keep the directional movement and account for shifting board positions. I'd also set up the pieces in the centre of their coloured face, rather than along an edge as he's done.
This is reminding me of my perennial desire to play Wizards’ Sudoku Chess from MSPA’s Problem Sleuth. Only recently, I obtained wooden checkered polyominoes to play “Broken Chess,” a game of my own invention. But an 8x8x8 cube would be the perfect 3D sudoku grid! And if the pieces could rotate after being placed, that adds the Rubik’s element… it also reminds me of Nitrome’s “Numbskull” Flash game. Of course, the Problem Sleuth version also includes Mario Kart and incomprehensible symbol systems, which would be… difficult to “take seriously.”
10:34 what appears to have happened is youve searched the possiblebspace depth-first? Or something like that, when it should be searched width-first (do multiple passes, on each pass only add tiles which are adjacent to tiles we already know are accessible (and bonus, are the same number of tiles as the pass #, from the inkling), and then do multiple passes, until there are no more accessible and unprocessed tiles) By doing that in explicit passes, the linked list will be ordered in such a way that each accessible tile will point backwards to the one space that is closer-est to the inkling (unless there are two, in which case the choice depends on how you ordered the checks within each pass and doesn't ultimately matter), at large resulting in the inkling always choosing the shortest path, because we searched the whole move space in a distance-saving way I know this because I did this for a grid-based fire-emblem-like game prototype with turn-based movement like this. Another trick is to do two passes with the same tiles to get particularly desirable movement behavior. First, add all the orthogonally-accessible tiles, and _then_ add all diagonally-accessible pieces, to the linked list (and the "to check next pass" list) This causes the inkling to "move diagonally first", AKA it will align itself with holes-of-accessibility from afar so that it can moge like a rook, rather than like a bishop, through them. This isnt important here because the inkling has an infinite move speed, but this is important jn my game prototype where your pieces have a limited "speed" and I wanted them to take sane, human paths, and splitting the one pass into two passes in series was the only way I could make the linked list turn out that way. As for the queen's movement, instead of using the linked list each step to determine movement, have the linked list instead determine how you will determine the queen's movement (AKA not a linked list at all, but a list of orthogonal or diagonal directions), where a direction will tell the game which way to "trace backwards until it finds the original piece", carving out the piece's visual movement path, allowing the piece to pass over the same tile in different ways if you choose different destination tiles
I think that there are arguably four additional moves rather than the stated two in that knight corner case. In traditional chess moving 2-then-1 is indistinguishable from moving 1-then-2, so one could argue that both should be allowed. This would mean the knight could access the squares diagonally adjacent but around the corner by moving one space onto one side followed by two into the other.
So awesome! Have you thought about adding another game-option that allows players to use their turn to twist one of the rubiks cube slices INSTEAD of moving a piece? That way you could also use this to set up a "normal" game of chess, with just the addition of a cubic board, moving pieces around corners and manipulating slices instead of moving. I could imagine lots of chess players making content about that version!
9:07 In standard chess, you can assume the player being "checked" passes their turn (even though pass is not a valid option in Chess) for the purposes of determining check. There are plenty of positions in standard Chess where blocking check is the defender's only legal move (even without assuming the "you must block check" rule didn't exist). In short, you would only need to check 1-ply to determine if an enemy is in check. 2-ply is only needed for determining which defender moves are legal. However, for a reasonably optimize script even that 2-ply check should be doable quite quickly.
2:47 Along with the 2 options you mentioned you can also go 1 down the red and 1 across. This is because you could go 1 down on green and 2 right towards red. But again... code lol
as someone whos been a huge fan of yours for a long time, and, unrelatedly, super into speedcubing, this was a super cool video to watch. this is a super cool concept and i cant wait to see what people do with it :D
It would be fun if, graphically, if a piece had multiple ways to get someone it got split up and parts of it went one way while the others went the other way
I like how you made this version of chess, although I had a similar idea where it was just a Rubix Cube puzzle, and the solved state is 5 of the 6 sides show different forms of checkmate.
I have been obsessed with cubing, Splatoon 2 and Chess at different points of my life. I never could have possibly imagined they can be combined. Just wow I'm impressed you even managed to build it
One feature I think would be really good is to make a custom rubiks cube size, like, what if I want an 8x8, or an 8x16x4. Although you would probably have to add some checks for if the pieces hit a backfase.
This is really similar to a version I was trying to make in Unity several years ago, except my "extra sauce" was the board is volumetric. Meaning the pieces can traverse INSIDE the cube, making all kinds of weird 3D diagonals and such, with new pieces being better suited to that, like ninjas etc. It's so cool to see a (mostly) working version of this with a whole different twist! (Literally 😂)
2:48 - A problem with how the Knights move. A Knight on the corner should be able to see more squares. If we're going by the traditional chess rules of 2 squares along, 1 across, then when dealing with corners the geometry should allow a Knight to access more squares than you seem to have programmed. For instance... In the timestamp provided, the Knight should be able to see the Red and Green squares two squares down from where it is. By moving two down on the Green, then one across onto Red, it lands as if it just moved 2 down on Red. By moving two down on the Red, then one across onto Green, it lands as if it just moved 2 down on Green.
@@wirelessbaguette8997 unfortunately, as a chess player, I cannot read anything other than digital clocks and chess notation. (I completely missed that lol. Tx for pointing it out. Ik he said he just whipped this together pretty quick, I was just tryna point out something I thought was overlooked.)
How about giving an option to also scramble the cube randomly at the beginning? Also, some reward for solving the cube during the game? I don't know what that would look like, but I find the idea interesting that splatoon characters get better and better the more solved your board state becomes.
Hey if you polish this up a bit and make each side checkered with their respective color, I could see this being the new chess! Naviary is making an infinite chess game and it’s so fun to see even cooler chess versions!
I would love to see this with other variations of "chess-like" games, like Hnefatafl (viking chess), Martian chess, tank chess, etc. and seeing how people could provide unique variations to those games
