Newbies' first game designs tend to range from medium, to epic kitchen-sink games. They have no idea of the amount of work such a game will take, and no ability to execute such a game. Don't put all your ideas in one game. Just pick one thing, and make a game about that.
Your first game should be something very small. Being small is not a downside. Games do not sell more as they increase in size — probably the reverse.
Start with something very simple (remember that it needs to work two-player, so you can playtest it.) By simple, I mean really simple. Medium-light (Splendor, Carcassonne, Ticket to Ride) are all way too big. Try something more the size of Love Letter. (It's sixteen cards and a few cubes.) Do not include a board. Use maybe 20 cards, and 20 tokens, and a single A4 sheet of rules, at most. Properly executing even a small game like this is a big job. It will take at least 50 revisions to complete, and many months. It will be far more interesting and challenging than you might expect, and you'll get to experience the full arc of designing a game.
In the game Coup, players have two character cards. On a
player's turn, they use one of their characters to do something very
simple. Using the Duke gains three coins. The Assassin spends three
coins to eliminate a character. The Contessa blocks this assassination.
The twist to this game is that each player's character cards are hidden,
and you can pretend to be whichever character you like, until other
players challenge you.
This is all the game is. It's a very successful game, and an ideal size of
game to design as your first game.
I've just designed a long series of medium-light weight games. The publishable ones took over 100 revisions, and 1-2 years of design each.
As I write this, I'm just starting design on my 21st game. It's my first medium-weight game (comparable to Catan.) It is not an easy task. I can't get my head around the whole thing. Everything interlinks into everything else, and I need to work on only parts of it at a time, so I don't change too much and disrupt the larger structure. There's a lot of work in just creating the components so many times. The game will likely take 200 revisions, and up to 2 years. This is despite me working on it very regularly, and my improved design and playtesting methodologies. Update: I've abandoned that project.
Don't think "Hey, I can do it."
Design a very small game as your first game.
If you've started designing something medium or heavier as your first game, stop working on it. It will not succeed. Not only will it not succeed, it will almost-certainly not be finished. I've seen these projects go back and forth (and sideways) in endless design, and sometimes be reincarnated multiple times. Many of them have expensive, professional art prematurely done. The faster you abandon this project, the better. It's an important and painful learning experience. Hopefully, it doesn't chew up all your formative years of game design. At the very least, start working on a new, much smaller project as well as the monstrosity.
If you're completely new, inspiration is where I suggest you begin — not from theme or mechanics, but by making your own version of a game you like. This provides a solid scaffold you can venture out from. Modifying something that already exists is far easier than creating something entirely new, as changes away from the original can be evaluated in isolation. If the game stops being fun, it's the specific change you just made to it, that caused it.
Make sure you take tiny steps to begin with. You don't know anything yet, and seemingly-reasonable changes to the source game can unwittingly corrupt it. Your first version of your game should basically just be the original game.
Even as an experienced designer, I don't stray too far from existing game types. It's very easy to spend a long time wandering around the possibility space, looking for a plausible game. Stick to making things that are similar to existing genres.
With enough time and work, a game can drift so far from the game that inspired it, that it will be a game in its own right, with no discernable genes remaining from the source game.
This kind of design is an excellent learning experience, as you will begin to understand why the original game was great. Even as an experienced designer, I do this kind of design, and I'm going to do more of it. However, once I've created my new game, I fully understand why the original was great. I can then create entirely novel games within the genre, without referring to the original, or to my previous game.