These days I start with a design document, recently this is more for legal reasons to get my games under ownership of my company, but it helps with deciding if a game is a good idea or not - that is the first step "should I make this game?".
Warren Spector of Deus Ex fame has a good checklist to meet before making a game:
- What's the core idea?
- Why do this game?
- What are the dev challenges?
- Well-suited to games?
- What's the fantasy?
- What are the verbs?
After that comes the sanity check:
- Has anyone done this before?
- What's the one thing?
And then the high-level "what will this add to the world", games are art after-all:
- Do you have something to say?
There are some points that many RPG Maker projects fail to respect, this is mostly because the vast majority of people using RPG Maker are not professional game developers with experience. The first 4 points are the ones commonly failed by RPG Maker projects.
I'll discuss why they fail on these points.
First Six
The core idea is failed mostly because people build their RPG Maker projects around "story", which is a valid core-idea, however story itself is made invalid by point #4 (read further down). The core idea should be something like "stat building" or "item acquisition" or "interactive battle system" or "online component"; traits that are common to RPGs that you are focusing down on.
Why do this game questions "what's the point?". Perhaps the market is right, perhaps you've seen a gap in the market, perhaps you've seen other games fail to capitalise on a really good idea. Going back to core idea, if your core idea is "story" (which it shouldn't be) then the question here is "Why do this game when everyone else on the planet can use this story and make the same game?".
Dev challenges should be self-explanatory. Be strict and harsh about finding where your weaknesses are in game development and understand them. Figure out how to make up for them/get around them. Limitation breeds innovation.
Well-suited to games is the one that's massively failed by the majority of the RPG Maker community. This is because a lot of people focus down on a very narrow story and end-up writing a film script (so their game is better suited as a movie). The tell-tale signs of this is a 10 minute introduction that has the player mashing the continue-button over and over again to get through a back-story they don't care about or a cut-scene about characters they don't care about. If it's all about the story, then write a book. The reality is as blunt as that. Always leave room for your game to exist.
The fantasy isn't your story, it's the role the player goes through. It's sometimes a subtle thing that you can't see yourself, so you need to inspect and analyse your ideas. For one of my projects, the fantasy is having a character who depends on you, but you are just their guardian, you can only help them, can't influence their decisions. For Final Fantasy games, the fantasy tends to be having a party of friends adventure together, each building on their individual personality traits and bonding on the journey.
Verbs are what the player does in the game world. Manage inventory, select battle commands, explore, sneak, hold conversations. It's the basic game-play that you need to focus down on. Working out your verbs is how you get your scope defined.
Second Two
Has anyone done this before? Figure out your competition. You always have competition, never say "no-one is competing with me" because you're just kidding yourself in that situation. If someone HAS done this before, figure out what you're going to improve on. Investigate on how it was done before and why something has or hasn't worked. "Done this before" can also apply to stories. This is also a good check as to if you the idea is too similar to an existing project to the point where there's no reason to release it. Many RPG Maker games are so similar to each other that there's almost zero reason for them to exist. I've played dozens of RPG Maker games and forgotten their titles or their stories or characters purely because they've all blended into one big mush of generic-RPG in my memories.
What's the one thing? This is very abstract but you should know what your one thing is once you've reached this point of the project analysis. Some games don't have an obvious "one thing", but if you ask the developers of the game in post-mortem they will tell you what the "one thing" was for them. For Final Fantasy V the "one thing" for one of the developers was actually the expressiveness of the character sprites. They were able to effectively convey a range of emotions to the player through sprites and choreography; that was different to other RPGs of the time (1992).
Last One
After all the other points, you should have something to say. This could be the theme of your story, this could be the lack of your story, this could be how the player interacts. It's a very abstract and high-level thing to think about. Is your game making a statement about current social injustices? Is it making a statement about the games industry itself? Is it a love-letter to games you enjoyed in the past and hope other players can enjoy too through your game? Is it a statement that RPGs have focused too much on story due to an invalid audience feedback-loop and have left gameplay at the wayside, resulting in mediocre RPGs that we have today? Are you making an RPG to say "This game is what RPGs should be like"? You don't need to be a philosopher, you just need to know what your game is saying; it is art and it should be adding something to the world.
Some of these points I can argue to have changed, in some cases I think more points should be added and I may disagree about some of these points, but it is a nice beginner cheat-sheet to figure out what your game is and if it is a good idea to pursue it at this time.
On the subject of writing your plot; a good way is to begin with the start and the end of the game, leaving the middle blank to insert all the cool scenarios they want the player to go through. When I say start and end, I literally mean when the player first gains control and when they last have control (the scenes before the credits). Then it's think of cool scenarios to put in-between and then write your story to connect these together. Half-Life and Tomb Raider are examples of games written this way.
This is also how games can be built effectively; prototype the start and end of the game, prototype the cool things that happen in the middle, then start linking these together (writing the story, re-arranging the bits in the middle). Prototyping means; get the map, cutscene and enemy battles (if any) sorted out.
Iterating over these during testing is what sorts out the balance, decides the equipment stats, enemy stats, stat growth, etc.