Tips from TK
Co-creator of IOT, GM or co-GM of Imperium Offtopicum V, Ab Antiquo, Imperium Offtopicum VI, Tiberian Sky, Iron and Blood, Iron and Blood II, Iron and Blood IV
Rule No.1: There Is Never Enough Time
Let me echo Mosher's first rule here.
You design a game and you thought, hmm, this is a good ruleset, I have ten, twenty players, a managable number, and then you jump into the game, and you found out that the amount of work you need to keep the game running to a high standard requires time that you simply do not have.
You thought updating will take you an hour, two at most. But this then becomes four, five, six hours. A typical Iron and Blood update took between 8-12 hours to put together, and that wasn't even a very complicated game. Iron and Blood 2, had it continued, would have eventually taken 24+ hours to update. Because the longer a game goes on, the more things you have to keep track of. Especially if, like Iron and Blood, the game world is meant to be dynamic, with lots of technological advancement, societal and cultural developments, economic booms and busts, empires rising and falling, and lots of unit types. In fact, you will need other people to help you.
Rule No.2: Delegate Tasks
You might need a co-GM, though personally I am uncomfortable with the idea, myself. If you have co-GMs, communication is key. Make sure everyone shares the same vision for the game, otherwise you'll be pulling it apart in different directions (this does not mean imposing your own vision on everyone else; synthesise ideas between different people, but once the game starts, make sure to stick to that group vision, with some leg room for rule adjustments). Set up a schedule for updates and things, depending on how ambitious your game is. If it's a story-driven game, it might help if different GMs look after different groups of players or regions of the world.
If not a co-GM, then have at least someone who keeps track of diplomatic agreements, roleplaying, player stories, and the like for you. Or people who can give advise on where the rules need to be balanced, if any. Though ideally this should be done before the game starts, which brings me to my next point.
Rule No.3: Start a Complete Game
As Lighthearter said, start the game knowing you know how to calculate everything, and the mechanics are finely tuned and balanced. I've been guilty of launching a game before it is ready, myself, and it is NOT fun and guaranteed to end the game prematurely. Unless it's a Calvinball-type game, you should not be making major changes to the rules once the game starts. Actually, even if it IS a Calvinball-type game, you need to at least have some idea about how the game will develop, if it is to last more than a couple of days.
That's not to say you can't make minor tweaks to the rules; the keyword here is "minor". If you really need to, if there's something that needs to be fixed, then fix it.
If possible, do a test run of the rules before starting the actual game.
Rule No.4: Know What You Want
Do you want a war game? Or do you want a nation-building game, a trade-focused one, or is it about explorers and settlers? What sort of timeframe are you looking at? How realistic? Is it going to be story-driven? How much influence will the GM have in the game (eg will you make events that the players will have to respond to, or will there be any NPCs). If a player leaves (and players WILL leave) what will happen?
Rule No.5: Find Elegance in Simplicity
The majority of IOTs die not because of really bad rules or bad GMs, but simply because either the GM or the players lost interest after a few turns. There are many reasons for this but one of them is that the game is too complex.
A streamlined, easy-to-understand ruleset makes it easier for players to follow the game and for you to manage it. It's easy to run wild and add new features and then you end up with 10 unit types and 20 government types and a massive tech tree and hyperrealistic economic system and then your brain explodes. Think: how can I achieve what I want in this game or model something while making it as simple and accessible as possible.
Rule No.6: A Dynamic Game is a Fun Game
But not too simple. You don't want a static game. Because then it's not a game anymore. It just exists, being boring.
It's all a matter of balance. You want to keep things moving, frontiers shifting, cultures evolving, economies developing, relations between countries changing, but at the same time you have to take into account the contraints of time and player interests. It's a delicate act.
Rule No.7: Communicate With the Players
Keep everyone informed, make the rules and important links and information easy to access. Make it as less of a chore as possible for people to play the game. Social groups is one way you can make information organised, but it means players (and you) have to check both the thread and the social group, so it's really a double-edged sword.
Think: you are their servant.
Rule No.8: Make Things Look Nice
As IOTs are mainly text-based, this mainly concerns formatting of posts. Make it look nice and make important things stand-out or easy to find.
The map needs to look nice as well; use pleasing colours and realistic-looking borders. If possible, tailor the map to each game's needs. Some games call for epic maps with lots of elements (provinces, icons, units, counters, etc). Others are just Risk-style maps with relatively few elements. Personally I prefer simple maps; see Rule No.5.
Rule No.9: Make It Special
What is special about this game? How can you make it standout? Is it a fantasy world or perhaps a region or historical era not previously explored? Or is it a new take on an existing idea. It should just be its sheer epicness.
Rule No.10: Focus
It seems such a strong word to use for what is essentially a playful mapping exercise with online buddies, but yeah. Focus. Resist the temptation to GM two games (Rule No.10.5: You Are Not Taniciusfox), or start working on another IOT or NES. One IOT at a time, and give it your best.
Of course, RL work/study is important. If running an IOT means you are neglecting real life, then don't.
Next Time: Tips on Making Maps for IOT