What is Your Game Not About?

A game can't be about everything. Players' time, comprehension, space, and money are all limited. You cannot make a game of unlimited scope. You shouldn't just take the game that's inspired you, and add stuff to it.

You need to decide what your game isn't about.

There are things that are common in other games in your genre. In your game, some of them will absent, simplified or merged. This gives space for the things that you'll be amplifying or adding. Or, perhaps, you're just creating a simpler game.

In Magic, creature combat is about two numbers — power and toughness. Combat is a numerical tradeoff puzzle. Some creatures have abilities that affect the combat, but most don't (some have simple or ignorable abilities.)

Radlands has people in it, similar to Magic's creatures. However, I really like abilities and special powers, so the people in my game have lots of abilities, but I intentionally omitted numerical strength or life amounts. People in Radlands don't have a power or toughness number at all.
In Magic, some abilities and spells can be played in response to what the opponent does, and will actually do their effect first (unless the opponent also has such an ability/spell, and continues the chain of response.) You can play these at any time, including during the opponent's turn, or in the middle of combat. While these add a lot of surprise and tension to Magic, they're massively complex, rules-wise. I omitted them when making Radlands. I only allowed players to play cards during their own turn. I felt that the rules complexity cost of these kinds of abilities was far too high.

Don't just add things

We were determining who should go first in Radlands. Roxley suggested flipping the water token, and if you won the flip, you'd get the choice of whether to go first or not. I told them the flipping was a good idea, but that if you won the flip, you should just go first. After the game was published like this, numerous people asked me why I wouldn't want the player to be able to choose. My response was "would you prefer to play 20 games of Radlands, and 20 games of 'work out whether you should go first or not', or 21 games of Radlands?" Everything you add to your game has a cost.