Why flip sprites instead of creating extra assets?

Why flip sprites instead of creating extra assets?

Why is the preferred method for changing the direction a character is looking in a 2D game a sprite flip? Why not just create the extra asset and show that one on screen?

Solutions/Answers:

Answer 1:

Flipping an asset at render time is relatively trivial to do, even on the CPU.

Creating an entirely new asset that is just a copy of an existing asset is essentially a way to double your asset footprint for no gain.

If you have the budget in memory, disk, et cetera for unique assets per direction, they can provide extra visual fidelity (for example, Super Metroid provided unique sets of sprites for both directions Samus could face, and because she is a very asymmetrical character it made the game look a little better). But often a flip is good enough, and cheaper.

References