Learning the right lessons

The computer games industry is a bit of a strange one. Everyone is convinced somebody else has the secret of success, so every single time something succeeds unexpectedly, you can guarantee a flood of imitators. Currently, the recent unexpected success is Minecraft, so we're seeing a flood of games which, to put it bluntly, look like the bastard offspring of Tetris and a random noise generator.

This is largely thanks to the industry's astounding capacity for learning the wrong lessons. In the case of Minecraft, they saw the presentation, and decided "Blocks are what people want!", when in fact the lesson they should have taken was "people want sandboxes they can really affect".

Now, I don't own Minecraft, and probably never will, but it's my understanding that the exceedingly cubic look is a consequence of its treating the world as a set of discrete chunks, each cubic unit of the world being homogeneous. The appeal isn't in the cubes; it's in the ability to mess with said cubes, to affect how the world is put together. That sort of player agency, coupled with accessibility, is what made Minecraft the huge success it became. Dwarf Fortress has been offering the ability to drastically remodel the game world for longer than Minecraft has existed, but its interface is opaque and it relies on AI routines following player orders, rather than the player doing it himself.

So, for success, offer something that takes the right aspect of Minecraft: it allows the player to do things that weren't previously possible, not the wrong aspect: it's made of cubes.