I came up with an idea for a game a couple of days ago and was hoping for some feedback on whether it could actually be fun. You can read about in more detail here but here is for your benefit: You start off with a simple platformer. In this platformer, the map is constantly moving and you must use your 4 commands to avoid obstacles. These four commands are jump, duck (slide if you keep it held down) spin and mouse grind (I'll explain in a second). Once you have built this, speed up the movement of the map so it is unplayable without lengthy memorization of the entire course, sounds terrible doesn't it? Then, you add the music. In theory it should work for any song. You craft the map around the song. The map is moving to fast for the players to play it blind, or deaf in this case. Instead, they must get in to the rhythm of the music and predict what they are going to have to do. This is probably the tripping point of the game. It is all very well if the programmer knows why he put such a block in, but the game will become very frustrating if you have to keep repeating the level to learn its foibles. The solution, at least in theory, is to stick to a logical implementation. If the note goes up, jump, if it is a low note, use down. Spin on beats . However, the second problem comes in two forms. Firstly, if you picture a jump animation, can you see the character landing before the next beat? And what if it is more extreme, with continuous variation? That is where the mouse grinding comes in. With continuous change in pitch, such as during a guitar solo, the user would use the mouse to keep the character in a broad path (that would rise and fall with pitch) Obviously, the best way to see if this idea would work (and be fun) is to implement a level. However, in the meantime, what do you think? Could it be fun or would it be too difficult to translate the music in a way that would be readable for everyone.

Thanks for the replies. With the gravity, do you mean that depending on the beat of the song, the gravity would change, or would it change as the song went though?

Two other things I was thinking of to get round the how-does-the-user-predict-the-type-of-move problem. Firstly, what if it was just one button (spin, for example) and you just had to do it to the rhythm, would that be too simplistic to be fun?

Also, if it was just moving up and down with the pitch and spinning to the beat, would that work, or would that be too far removed from the platformer idea and basically just be Guitar Hero.

I don't have any insight on the subject, but have you tried Music in Motion? It might have some ideas you can borrow

Certainly similar to what I was thinking of, thanks for the link. The problem I felt with that was a) you were out of sync with the music, so if just felt like a platformer with a nice music visualisation in the background. Also, it never really felt like it was flowing. You know in Guitar Hero when you get a huge note streak, that's the kind of feeling I am after when you make jump after jump in time with the music.