Let me begin by saying that game creation isn't an overnight process, 99% of the time. Making a game takes patience and time, and in alot of cases it will take experience, unless ofcourse you're one of those people who aren't willing to learn, and simply copy and paste from tutorials. If you can't understand, there are plenty of helpful tutorials and examples to help you, and i'm sure alot of the click community would oblige.

Now, to the tutorial!

Idea:
Once you have to idea, you're half-way there. If you know exactly how you want the game to play, what attacks/manuevers the player will have, movement styles...etc...then the game will be easier to make.

Genre:
The genre affects every aspect of the game. This takes from you're idea. For instance, if you wanted to make a spiderman game, your most probable first choice would be a platformer, however for other gaming ideas you would need to consider this carefully.

Originality:
The most important part of a game. Originality is something which allows games to thrive. No one is going to want to play a game which bears striking resemblance to pac-man, or any other game for that matter.
Saying that, however fangames are another issue. They must have similar attributes to the original game...but with a difference. No one will want to play the same game, if it's remade exactly like the original.

I've found that unique attacks, and a good story will help make your game original...which is what i will address next.

Storyline:
A good, intriguing storyline will keep the player interested in your game to the end. What makes a good storyline? Clever planning and proper layout. No one will want to play a game which has a 'mixed up' story line; Can you imagine playing the 'Home and Away' game? =P

Playability:
Once you have decided what attacks or maneuvers will be unique in your game, you have to think about control, and difficulty. "Will i be able to program this effectively? Will people playing my game be able to control these attacks? Are the attacks necessary?", are all questions you must ask yourself... Another point i'd like to make: Location, Positioning and Access, of the keys that is. No one is saying you can't have 53 attacks in a platformer game, but the shortage of keys may be a problem. The reason shift and control are the default buttons in TGF is because they are located at the edge of the keyboard (location), they are positioned so they can be easily reached by your left hand when controlling with your right (positioning), and each button can be pressed within a fraction of second between each other (access). Can you imagine 'Q' being jump, 'U' being fire and the enter key being punch, for example?

The movement engine must be relatively smooth. I'm sure you've played a game in which the movement was jittery, the screen shook and was unbearable. If not, then surely you can imagine it. The wall collision and other small things must be bang on, otherwise the core game engine will be faulty.

Graphics:
The graphics in a game define how you remember it. The image you have in your mind comes from the graphics. Even if you're not Picasso, your effort it always rewarded. You won't get anywhere if you have a 'It'll do' attitude. Give graphics your best, as they are often the hardest thing to create well in a game. I'd rather see simple, well animated graphics, than poorly coloured, ill-proportioned ones. My opinion purely.

Conclusion:
I've given my opinion on making a good game, however, i'm sure alot of other people have other views, which i'd like to hear. If you've just read this, and you think i've made an error somewhere, don't hesitate to correct me, i'm only human. =P