Video data?

The broadest we can describe 'video' data is "anything that you can
watch," but we all know theres more to it than that. Take e.g. Flash
movies, some people will say "Although you can watch it, it's not
video."

So I've come to the conclusion that the definition of "video data"
would be any moving pictures encoded as a series of macroblocks or
(uncompressed) pixels. Examples are .mpg, .avi, etc.

And Flash mainly encodes using object-based techniques, right?

I was wondering how many different major techniques exist for video
encoding? Object-based techniques obviously compress way better than
rigid macroblock technique MPEG uses, but would be understandable that
we are not yet at the level to create an encoder that would recognize
objects and other redundancies in natural real-life graphics movie.
Not to mention it would take fuckin FOREVER to encode, if MPEG takes 6
hours to compress a 4.7 GB DVD to 700 MB then object-based codecs
would take far longer, possibly an impractical amount of time
(thousands of years.)

Another method I can name (ineffective for a real life graphics movie)
is the one used in storing the "movie" of a videogame. The movie is
basically the input of the player playing the game and is stored as a
record of which buttons were pressed and is extremely tiny (100 KB for
12 hours worth of gameplay.) But if the movie was encoded with MPEG
techniques it would be over 1 GB, the difference: you can play the
movie from any point while with 100 KB worth of information you can
only watch it from the beginning, no rewinding.

Mainly, I'm just trying to distinguish between rigidly-encoded video
data and the discrete, specific types of video with their independant
complex algorithms.

.