More than once I’ve been asked what advice I have for people who want to design games, and everytime my first piece of advice is make sure everything in the game supports the story/experience.
I have usually just said “story”, but not every game’s main thrust is the story–hence, story/experience. What does that mean? Even if your game is open world exploration or a MOBA fighting game (Multiplayer Online Battle Arena) or an FPS, and not an adventure game (as is my wont), there’s still a specific experience you want players to have. A mood, a frame of mind, a social interaction, or whatever else. I would add that I believe most people unconsciously add their own story to games that don’t lay out a specific one (ask me sometime about the narrative I’ve decided to paint when my husband plays World of Tanks!), but even if they don’t add their own, there is still the experience.
That’s what it means in theory. In practice? It means don’t add anything to your game that doesn’t support the story you’re telling. Don’t do it because when you do, you waste your time (and money and resources) and you waste the player’s time on something that is superfluous, unimportant, and may even remove them from the gameplay experience. This is why you should never add a puzzle for the sake of a puzzle. Only add a puzzle if it reinforces a point about the characters, the world, or the plot. If it’s there just to kill time, then that is all you’ve done. Killed time. And with the modern game audience, that’s all it takes for them to move on to the next thing.
My favorite examples of these are, of course, from adventure games. In that community, it gets called “moon logic” a lot, or puzzles where your true goal is to “figure out what the designer was thinking.” At the peak of this in my book are three puzzles from King’s Quest V: cheese in the machine, pie at the yeti, and emeralds in honey.