I've been playing in nomics for years. Several are still going on.
Some words of advice:
1. Either you are going to have to have an actual winner, a cataclysmic rule-trimming now and then, or the game will die off by apathy. In fact, the first two won't even guarantee your game will survive.
2. You'll want some rule to allow new players in to replace old players quitting. In longer-running nomics, the entire player base will be replaced multiple times.
Winnable nomics don't generally need this, although you'll still want it anyway.
3. There are two general kinds of nomics: Procedural, and Substantive. Substantive nomics are all making a game to play, and the nomic is merely part of the process. Procedural nomics are mainly into modifying the nomic itself.
4a. For substantive nomics: Think of the weirdest, feature-burdened, most unbalanced, least playtested and overall ramshackle game you have ever played.
Your Nomic will that game look like Go by comparison.
Every player has their own vision of how the game should be, and the result will be an amalgamation of anything that seemed cool at the time. Expect bizarre, overcomplicated, poorly-fitting, and outright broken rules. There will be no playtesting. Needless to say, it will be a blast.
4b. For pure procedural nomics. Indescribable. Ordinary conceptions of 'nonsensical' rules will no longer be sufficient. The ruleset will be at war with itself, and lose. You have been warned. Have fun.
4c. It's also possible to mix the above. They don't tend to be quite as insane as pure procedurals, but definitely add much more weird.
5. Use the right starting ruleset. The character of your initial rules will determine the character of the rest of the game. There are several levels of verbosity, each with vastly different feels. For example, take the following proposals:
A: "Every turn you get an apple, and you can eat for 10 health."
B: "Add the following rule to the ruleset:
'An Apple is a kind of Item. At the beginning of a player's turn, that player receives an apple. Apples may be eaten to gain 10 health points. To do this, a player removes the apple from their inventory and increases their health by 10.'"
C: "Create a new Kind of Tool with the following characteristics:
{
Name: Apple
Use Action: Eat: Destroy this Apple: The User adds 10 to their current Health Points.
}
Create a new rule with the following text:
{
At the beginning of each player's turn, an Apple is added to that player's Inventory.
}"
All of those are, more or less, how the same proposal would appear in three different nomics. No one way is 'right'.
Also, if you are looking for a substantive game, do not use the Suber ruleset. Don't be fooled by the fact you can change the rules, you'll have to fight with them until you get what you actually want. Specifically, rule 112 will cause you suffering the moment you want something other than a simple, point-based victory.
If you're looking for a procedural game, the Suber ruleset is acceptable as-is, but flawed. The turn-based nature isn't suitable for online play. You might also want to make the judging system more democratic. Fix both of those and you'll be fine.
6a. For substantive nomics: It's much better to have a goal too close and achieved early than a goal too far which is never achieved.
6b. For procedural nomics: Have a good Justice system, because things are going to break massively. Have an emergency system when the Justice system breaks.
6c. For all nomics: Short proposals are better.
7. You WILL need a full-time admin guy, and quite possibly multiple.
8. Have fun. Don't take things too seriously.
----
Leafsnail: I wish I saw games that cut-throat. Most nomics tend to be too nice.
Taking that situation, B would be in a great situation to negotiate with C or D, while meanwhile A has torpedoed his own chances. Exiling just D would have the same difficulty: Everyone is going to see the opportunity for a cheap ally.
D's vote to C, though, seems cunning enough to pass. Of course, you haven't actually gotten rid of any enemy votes, and now C is even more powerful...