I wanna speculate and say that I should have access to my pc sometime later this month. If not, then hopefully by at least March. Until then, I could use some help.

Most of the gameplay of a fnaf fan game I already fundamentally have a technical understanding as to how to implement. However, the ai is one aspect that I do not. Along with that, it’s not being made in Clickteam. So I won’t necessarily have a reference to go off of.

However, throughout my hiatus, I’ve thought on and off about a potential few ways in which it could be implemented.

The first would consist of how the original game is actually designed. Or, at least believed to be designed, which is in a way that is predictable.

Characters follow a path, most likely through a timer, and upon the timer expiring, the character is signaled, or, I guess in this case, the texture is signaled to be updated in order to symbolize that result.

So effectively, it wouldn’t really be ai, but a continuously looping system of prediction. And this, of course, assumes that the original game actually works in that exact way.

The second would be more or less how an actual fnaf game works. Essentially, you have a span of numbers that correlate to either the chances of a move, or the chances of a draw, (halt) a timer is constantly at play, where, upon expiring will check the result of those numbers, and only if the numbers correlate to the range for movement will the characters actually move accordingly, and optionally, to a random location at that.

This would make the game less predictable, but it would also serve to be less accurate. (Again, assuming the actual game is not randomized, but instead a linear procedure.)

So what would you all rather see? Best case scenario is that we stick to the “as close to 1-1 as possible route,” and only after the game is finished, can we try to maybe expand on it with original artwork, and more robust design systems.