I think one might divide design thinking into three levels.
1. Child level – Basically reciting stories of what you want the player to do.
2. Balancer level – Having understood that level 1 is broken, with an understanding that cross-cutting concerns, balance, and decision-making are important, seeks to complexify and balance and diversify decision-making to produce elegant and deep game systems. Can create well-balanced, abstractly interesting mechanical systems. The problem with level 2, though is that it works on abstract, mechanical-level analysis of systems, which doesn’t relate directly to what players actually want from their games: emotions. People working at level 2 can end up ‘balancing out the fun’ of a great goofy design (happened during the dev of Magicka; they fixed it), or complexifying a game until it’s so deep that it becomes intimidating and drives players away, or working endlessly on details that players just don’t care about, or ignoring easy wins because they depend on mechanically-irrelevant pushing of emotional buttons.
3. Emotioneer level – Having mastered level 2, with an understanding that peoples’ emotions respond to stimuli besides balanced gameplay, seeks to pull heartstrings as much as possible while asking the minimum of cognitive effort from the player. Uses level 2 skills effortlessly to produce the foundation of balanced, elegant gameplay, on which to generate varied, emotionally-relevant experiences which aren’t necessarily balanced or complex or deep or mechanically elegant. Requires broader thinking than level 2 because while maximizing balance/elegance/crosscutting concerns is a relatively narrow problem, human emotion triggers are absurdly complex and varied.
It took me years and years to solidly reach level 2 and I am in an ongoing effort to understand level 3. To the point where I’m not even totally sure I can properly define level 3, but I can just sort of feel that there’s something there that I don’t fully get yet.