It was during a boring lecture at NCRA that Debsankha and I came up with the idea of this game. We were wondering how to pass the time, when he suggested we play a 4X4 tic-tac-toe. I had tried it by myself once, and it wasn’t too good, so I suggested a 3X3X3 version instead. We drew out the grid, and played it, and believe me it was fun.

Let me give a brief idea of how to play this.

The tic-tac-toe is essentially being played in a 3X3X3 grid. All the rules are same as in the normal tic-tac-toe, except that now you have many more lines which you can complete to score. You have all the edges, face and body diagonals etc. of the cube, along with the normal 2D tic-tac-toe lines on each face of the cube. This gives you a total of more than 40 lines (haven’t calculated it exactly yet), as opposed to only 8 for the normal tic-tac-toe.

But the trick is, you cannot play this on a 3D cube. You must draw three normal tic-tac-toe grids, corresponding to each ‘floor’ of the cube, like so:

Take any to be the lowest floor, doesn’t really matter (as the rules of the game are reflection-symmetric). Play as in the normal tic-tac-toe, but don’t cross off the cells once you complete a line. This is because first, only lines that belong entirely in one floor can be crossed off in the old style. Second, and this is interesting, more than one of your lines may pass through a single cell. So instead, you must add to your count of lines. What’s more, before you add to your score, you must explain to your opponent how exactly you are completing a line. It may so happen that in one move you have completed more than one line, which you must also explain. But here’s the catch: if you fail to notice that you have completed a line, you will not report it, and hence miss that point. Here’s a 3D tic-tac-toe game in progress:

I’ve pointed out the completed lines with the same colour. The circle with the red/blue gradient belongs to both the red and the blue line. The yellow and blue lines are verticals, for example, the green is a body diagonal, and the red is a normal line on the top floor. The black ones do not complete any line.

In a normal game there won’t be any of these colours, so you cannot keep track of which lines have already been drawn. That might make it difficult for you to diagnose whether or not your opponent has really completed a new line when they are claiming to. Fortunately, there’s no need to keep track of this. Just make sure that when a player claims to have completed a new line, that line must involve the last cell which they have filled. The reason for this is simple, figure it out.

We noticed that the centre cell of the middle floor is vital, because many lines pass through it. If a player places their symbol there, they’ll block many lines. This doesn’t matter so much in 2D tic-tac-toe, which isn’t too smart a game anyway, but here the benefit is increased a lot. This means the starting player has an unfair advantage, because they can block that cell. Moreover, since there are 27 cells in total, the starting player will have one more move than the opponent, which is also a little unfair.

To simultaneously take care of both the problems, we decided to make that cell ‘democratic’, meaning that we shall assume that each player has a symbol in that cell, and their lines can pass through it. This greatly increases the number of possible lines in the game and makes it more interesting. Also, now there are only 26 cells to fill, half of which is filled by each player.

Therefore, we shall put a little symbol in that cell that signifies that it has already been allocated. What better symbol to put there than the ‘trademark’?

3T stands for Tic-Tac-Toe. It’s just like the centre white cubie of the Rubik’s cube or the Ace of Spades in a deck of cards. It’s special.

Note one more thing. Since now we have many lines to complete, we can even have three players at this game, and everyone will have some lines to claim for themselves. We actually did play with three players. But we needed a third symbol now. So just to increase the geek quotient, we used the initials of our surnames in Greek letters as our symbols. So we had δ, μ and ɣ. You can try it too, if you want onlookers to be awed.

We thought of extending this to even higher dimensions, but then we’ll be left only with a tensor-like index notation, and no good way to visualize the grid. I thought that’s what would make it interesting (what makes this one interesting is that you must visualize the 3D grid in your mind), but then I thought that it’s too complicated.

Overall, I think it’s a fun game, and a good exercise in mental visualization and spatial manipulations. It also requires brains, unlike the normal 2D 3T.