I have so many unfinished and cancelled projects that I would feel much more comforable writing how to *not* make games, but I will try to give some output too

1. CHOOSE AN IDEA WITH POTENTIAL
Games I want to make - Among my games, the ones I really liked turned out the best. Don't get me wrong, you can still make a game you would not enjoy playing and it will be very good and players will praise it. It's just... these I liked seems a bit better. And the design phase was faster. Not enough games to make a theory out of it, but I lean to "making games you want".
Games I want to have made - This is stupid. I dasagree. It's like you make games that The Internet needs? Because there is some moral obligation or imperative for a certain game to exist? This way of thinking is for me a sure sign of a game that will never be finished. Also, I find the circles lacking "Games players want to play". Come one, games are to be played by players, you have to at least acknowledge that part if not making it a priority

Otherwise it will be just a design feast to satisfy your creative needs only, not a game. Replace the "want have made" circle with "players want it" circle and I will agree.
2. ACTUALLY START THE DAMN GAME
Register, login, logout system. You can start coding thes right away, even if you have no clue what kind of game you will make yet, right?

8. USE AWARDS, COMPETITIONS, AND OTHER EVENTS AS REAL DEADLINES
Confirmed. When I was making Lords, I had my holidays trip as a deadline. I made the game exactly in 2 weeks

It was playable and was put online. I simply had no choice, I had to, absolutely had to finish it before the deadline

10. TAKE CARE OF YOUR MENTAL AND PHYSICAL HEALTH
* exercise - confirmed, exercising is a miracle cure for a coder/designer. This is a thing I have no doubts about. Do not forget that it is about providing oxygen to the brain, not to produce muscle mass, so do it outdoors.
* food - yes, this can affect it, I can see some changes in productivity depending on food availability
* sleep - not sure, I can code when half conscious without making any bugs when I'm in a state when I'm unable to talk properly with people do to sleepines anymore

Althrough I would not dare to attempt inventing new solutions to coding problems then or rather refrain from designing. But just regular coding of typical stuff would be ok.
12. SAVE IT FOR THE NEXT GAME & 13. CUT. IT. OUT
It was written for retail games, we are talking about browser games. This is a different world. We can simply add the features later

Actually you should not include all features you plan in your first version, it would be a waste to do it before testing the other features first or gathering feedback from players.
14. IF YOU DO QUIT, SCALE DOWN, NOT UP
I'm a fan of scaling down. Every moron can design a playable game with thousands of monsters, locations, quests and items. But to properly scale down it take a true game designer. Personal opinion.
15. THE LAST 10 PERCENT
I made Lords fully playable in 2 weeks, then I used 6 months to polish & fix it. These numbers should be self explanatory

Not included in the article:
WHERE YOU ARE NOW?
Everything changes after you made your first finished and playable game. Plenty of problems above will instantly disappear. Your first goal should be crossing the rubicon of first released game at all cost. It's worth more than reading 1000 best tutorials.
TIME
People get it wrong, they either think they have infinite time and start too late or they think that thay don't have at all and start too soon. First you should learn coding. If find it a bit weird when someone with 1 yera coding experience start maikng a game. Why don't you enjoy just coding small thing? Exploring the language? It's like these people do not enjoy some process of coding, as if coding was making sense only if it lead to a game being made. It's almost impossible to make a game if you don't love the very process of coding. Or at least it will be a sad experience (you don't love coding, do something else, don't torture yourself by making computer games). On another spectrum are those with Blizzard's mentality, the game will be made when it is made. No, it won't be made, you will die of an old age before that could happen

Do yourself a favour, make some deadline. But do not make it too soon, first give yourself the luxury of wild coding (let yourself be a coding kid before you start being coding pro!).
KNOW WHERE ARE YOU GOING
Everything above is a minor annoyance. The real thing is that most of us do not know what kind of game they intend to make. Above all reasons listed or not listed, this one gathered the biggest tool of dead projects in my case. You have some blurry vision... can't formulate it with words... start making it... got stuck... don't know where to proceed now... you ask other people... they give you some incompatible ideas... you are unable to explain them what you want to achieve... you try to push forward... you get exchausted... you find a new idea which is more clear... you think it might be better... you discard the old one... you proceed with the new idea... cycle repeats.
My most serious case of this disease was during our One Month Game Coding Competition 2010. People on IRC started to welcome me by "Chris, what kind of game are you making *today*?"

In the end I failed to make an original game and just made some standard basic game in one day, just before deadline. The unability to decide in details what kind of game I was going to make made me outright waste 97% of the time assigned for that project!
I don't know how that one could be solved. Maybe we should accept the fact that our abilities are limited? That is is humanly impossible to design a whole game just by ourselves? That we have to build upon other games, invented by thousands other disigners through the ages? That we can invent only a smal portion of our game and copy most of it from other games?
But to end this via optimistic statement, from personal experience: you don't learn much from success, you learn from failures
