In my humble opinion there are different approaches to a game.
However the first thing should be to write down your idea and think the game through.
- Who is the player?
- What's the main objective?
- What can the player do?
- etc.
This can be useful to have a list of features that are essential for your gameplay to work and to know what requirements your game has.
However you should think about each feature twice "is it really needed?", "how hard is it to make it work? (animations, scripting etc.)"
When all the writing is done and you have a feature list and maybe a design document for your game together you should stick to it.
Adding features during the development process might kill the project in the end, because there is too much.
For a first game: Keep it simple!
Now its time to do something. However what you do with all the written documents depends on what you can do and what you cannot do.
If you're an artist, it might be a good idea to draw some concept art, build a few first models and sceneries which express how the game should look like etc.
Then you could try to get a scripter/programmer as a team mate.
If you're a programmer/scripter, you probably should start with a scripting
Which means: Either look for some free models which fit to your game, or build quick prototype models (a simple box is sufficient for a lot of things) and script the key gameplay features to see if they work together and to have something to show when you're looking for an artist.
If the game is already fun with your simple dummy models, you have a winner and it should be easy to find somebody doing the graphic part.
Thats my personal view on the subject.
However there are more experienced users around here, who are able to give a way better advice.