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. Don't just take the game that's inspired you, and add stuff to it.
Deciding what your game isn't, is as important as deciding what it is.
There are things that are common in other games in your genre. In your game, some of them will be 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, 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.
In Magic, creature combat is about two numbers — power and toughness. Combat is a numerical trade-off 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 and health amounts.
Where is the complexity?
It's not just about adding and removing things. Some things will be amplified, while others are minimised. You need to choose.
Where is the complexity in your game? In the combat? In the movement? In the trading? It can be anywhere, but it should be somewhere, and not everywhere.
If some things are complex, others must be simple.
In my adventure game, combat is thematically necessary. But there's no combat system. You just draw a card, and if it's a bandit, you must pay him one coin, or lose five health. That's it. You might think that sounds extremely simple, and not complex enough to make an interesting game. That's correct. The complexity is elsewhere, not in a combat system with "block" stats, bonuses, and convoluted mathematics.
Design into the good parts
What's the most interesting phase of your game? The beginning/end? A certain place, or condition? Few or many resources?
This is a decision that is ongoing, during development.
There's something I call "zoom". You can just zoom in or out, mechanically and thematically. This is a great way to make your game different.
I've tried to make a lot of farming games. One of my attempts zoomed out, and was more of a game about a whole village. There was war, and money, and exploration. Animals and crops, the usual farming systems, were massively simplified.
In Lords of Waterdeep, you go around collecting adventurers, and doing quests. You do this by going to spaces on the board. There's also the option to add spaces to the board, by buying buildings. Whenever someone goes to a building you built, you get a bonus. I really liked that, and made a zoomed-in game that was ONLY about building buildings, and having other people use them. I made the resource system much more complicated.
A story about adding things
We were talking about who should go first in Radlands. Roxley suggested flipping a 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 (you didn't get to choose). After the game was published like this, 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.