Yes, I like the "caramenti" name. The Mexican name for tic-tac-toe is "el gato" (a sharp or pound sign is "el gatito"), so there might also be another name possibility in there someplace. The more memes the merrier!
I just wrote this new page in which I took a deep dive into how the game works using the memetic modeling techniques I've been working on. It might or might not be interesting, or shed light on how the game works. Hopefully it appeals to computer science types. I am working on another page where I'll try to analyse strategies. Ultimately I should knock up a programmed interface where one can play against the computer, and maybe the computer against itself.
I devised the game as a simple model wherein you can look at strategies based on the state of the game and also beliefs about what the other player is trying to tell you. They can be trying to cooperate, deliberately lying, or random (or a mixture).
The point of the game is that you can "signal" the other player with your "show" coin (on the board) and your "cover" coin, to give them an idea of what your "secret" coin is, which is what determines the outcome of a bump, i.e., each player gets two points, or a "conflict" in which both flip coins to arrive at a random outcome (see below)
The game is played in "rounds" of whoever gets to ten points first. Cooperation is where each player gets as close to ten as possible each time. How well you're doing could be expressed as a ratio or a rolling average, e.g., 7/10, 9/10, 10/6, 10/8, 10/10, 10/10, 10/10, as opposed to where you just try to beat the other player as badly as possible.
I don't have this on the page (I may add it) but you can play in "passive", "even money", or "aggressive" mode where a "conflict bump" is different depending upon what you do if both players flip coins. The bumped player always gets three points on both tails. On a double-heads, the bumping player either gets zero points, one point, or two points, and one point for one heads, one tails. In "even money" the average result is the same as cooperating and each player getting two points every time. In aggressive mode, there is an average advantage to a player trying to bump regardless of what they think is going to happen. In passive ("simp") mode, one tends to lose over time with devil-may-care bumping.
In any case, if you bump without any idea of what the other player is trying to tell you (assuming they are not just playing randomly) there is a 25% x 25% x 25% chance that the other player will get right up to nine and you've practically lost the round.
Would it be useful for me to put up another page showing some samples of play? Like a few moves with some bump results?
No comments:
Post a Comment