I would say learn the basics of logic bricks, then “graduate” to learning python. It’s not as difficult as you think. You can then move on to better your graphic abilities.
As you learn, you can make “learning projects”, which are little demos in which you incorporate a certain feature you learned. For example, a mini FPS demo in which you have to shoot the zombie cubes trying to kill you, or a ball game in which you test out keyboard controls and such.
Here’s the thing: It’s better to have a polished and very highly functional game (through python knowledge) and have simple graphics than have awesome graphics with little functionality. A clear example of that is minecraft. You could probably make all mincraft models in a week, maybe even less. But it took a few years to get the gameplay down and it was arguably the most successful indie game ever made.
In terms of the actual development process, you should always start with a prototype, which means that you’ll be dealing with little graphics and more programming (so using spheres and cubes as players and enemies), just to see if you can actually get the game to work. The prototype should include all significant gameplay portions you want to have in the game. This works as proof that you can actually make the game, so that you don’t start making the game and end up quitting after a month’s work because you learned that you don’t have the necessary skills to make it. Then you can actually work on graphics and gameplay without fearing that you can’t complete the game.
If you want to learn some of the basics of the blender game engine, you can check out my youtube channel in my description. I cover topics such as adding controls to a character using logic bricks, basic level design, and creating a complete simple game that functions properly.
I’m also planning to start a python programming series, though I’m not sure when that’s coming.
best of luck!