Alex Wheldon wrote about reducing the sprawl of game mechanics and a focus on core mechanics to improve game designs in his Bene Factum blog post Density, not volume.
I agree that ignoring the minute mechanics underneath a game while focussing on a high concept is dangerous (and tempting). However a lot of AAA games fixate on the given mechanics of a genre be it platform/FPS/RTS and just bulk up the art and story around it – making dense gameplay but really just clones of old design ideas.
Usually I come up with a game mechanic, then fill it out with story, levels, upgrades, scores, bosses, achievements etc. It is this kind of volume that actually adds length and fun to the game – compared to psychologists lab tests of skill or abstract puzzles.
However I agree very much that non-essential vagueries of mechanics are unhelpful – both for the player and for the developer. Unfortunately some genres seem to need a lot of little mechanics to play out – Simulations & RPGs are particularly bad for this, as the ‘High Concept’ clearly precedes the mechanics. Still it is an interesting challenge to minimise the mechanics and still create an engaging simulcrum of the concept.
Exactly the king of challenge I am working out on the Spice Road design. As there are a wide range of activities the player can carry out and lots of stories to engage with – it is important that the basic interactions remain consistent and I should minimise adding new mechanics that will hardly get used. To this end I am focusing on four mechanics that will drive the vast majority of the gameplay – Travel, Combat, Trade & Conversation. The minor mechanics like party management, city policies, inventory management will all share a consistent dialog style interface, while the complexities of politics, diplomacy and questing will fit into the Conversation system.
So, while not as minimalist as Mario – Spice Road will be a lot simpler and cleaner than a lot of Simulation/Strategy/RPG hybrids.