Draccan's choice (in random order) for the bread and butter components of good mmo development:
1. Player housing
Player housing is not about house decorating. While fun for some in SWG. It is about creating a persistent world that is worth populating. A persistent world with player housing means yet another reason you want to go by the game, to check up your inventory, to feel you aren't not just playing yet another combaf game. Player housing makes sense.
2. Non-combat "classes" (or focus)Non-combat classes actually adds to the game. They attract players to a game that wouldn't otherwise play. This means fighters will feel more unique and more like heroes, it also opens up the game for more interaction when the fighter visit the crafter.
3. Abolish auction houseIntroduce player vendors (and some system to find goods from far away). It allows characters to interact and socialize more, even bargain directly with the crafter.
4. Great ressource system and focus on crafting
Crafting is key to a persistent world. And combined with that is a good ressource system. Ressources have to be shifting. There need to be balance. Some ressources should be fought over. And there need to be long term reason for ressources to fought over - if there is no reason to have them it destroys the balance, if they are too hard to get or only top guilds can sit on it forever it is bad too.
5. A flexible class system focused on skills
I believe the best system is flexible class system. "Iconic" classes like in Bioware's upcoming mmo is plain boring. Aion is not much better. SWG did it pretty well. On the other hand is Darkfall a great example of the problems of a true 100% skillbased system (where many tend to be nuking mages now). Some sacrifices have to be made, some direction. Again, SWG did this well. It could be even better with more nuances and choices. In fact it can be much better.
6. Death penalty
No death penalty means the constant never-ending zerg. There has to be consequences. Games these days seem to go for either light or no penalty or the extreme ffa loot (Darkfall, Eve, Mortal Online, Dawntide). FFA loot is problematic - scares people away, makes ganking highly profitable, and uber guilds possible.
I propose a mixed system with random factors. Make death unpredictable. Example of what I would like to see:
A) a slight loss to your ability to raise skills for a while. Maybe a COH'esque style where it makes it slower to progress for a while.
B) a random chance about 10-20% to lose ONE random of your top ten most valuable possessions.
C) Item decay when you die (5-10% max). Higher for items you didn't insure.
7. No duo-factions anymore
Minimum three factions pr. game. Preferably five. More factions means more chances to ally and make good politics over ressources. Also there must be a faction system. WAR and WoW are great example of why duo-factional pvp gameplay is stale.
8. Overt / covert pvp system
It is great to have a pvp system like in SWG where you could go overt when you want and covert when you want. This means that you can choose not to pvp for a while but also go back in the swing when you want. Some parts of the could be automatic pvp. Others are consensual (50% would be good). This allows for some areas to hold special rewards and ressources to fight over without being incognito.
9. Player-driven economy
Any mmo developer with respect for itself should hire an economist to figure out the macro and micro economy of the game. It is hugely important. SWG did for a while in the beginning. So does EVE.
A good game will be 80% player driven and the remaining 20% is in there to spice things up and to regulate the economy. Things should be built and crafted. And great loot should half the time or more be recipes. You should even be able to create recipees based on your existing ones. Invent new stuff. Buildings, Guild Houses, Player Cities, garrisons, watch towers. NPC strongholds working for you should all be craftable.
Combat is hard to nail down. But I personally like a mix of twitch and traditional RPGs. SWG did quite well before SOE broke it. AoC is very close to perfection, but have problems with fatalities and combos. These are both premises to work on. SW ToR seems to introduce some goodies in form of a Smuggler cover system. Ambition is important here!
11. Instances and raids
As few instances as possible. And as few raids. I am tired of RAID centric games. RAID and mini-games take away from the game world. Raids might be fun for some. But if 80% of what you guild does is go into a raid to fight some npc's why not make a non-RPG about it. Really!
Players should be kept in a living-breathing world, not playing in smaller solo groups. Find revolutionary new gameplay.
12. Animations and graphical style
I prefer AoC over Aion. Why? I can't stand the (hundreds of youtube videos I have seen of) Aion combat. Why? Sure the graphics have polish, but way to many particle effects and disco lights. I can't stand my mmo to look like an 1980s disco music video. I was a teenager in the 80s, that's enough for me. AoC offered a beautiful mmo, that is unparalleled in beauty for any mmo. Great standard that art director Didrik Trollesen set there. And no more manga-cutie-elfy-elves, thanks!
13. Servers and technology
Most games launch big - WAR and AoC sold easily 1 million boxes and retained about 100K players. Cool. But they way servers were crying to be merged three months later. But most developers don't care about this, so they don't create the suitable server transfers before many months later, sometimes a year before they get around to it. I demand new ways of creating a world. And not solving it by sharding and instances alone. Why not created several versions of the same content, so you can have one server like in Eve?
With one server you could have more of the same content, two or three similar bosses. And implement a system where the less players there are, the more likely your missions or quests take place closer to the others playing. That means the whole wide world is reachable and can be visited, but your questing takes place closer to people a larger portion of the time, so players won't be too spreaded out.
14. Dynamic content
We need genuine dynamic content. Wandering surprises. NPCs fighting and rebelling against each other. 50% of an mmo's development should be about dynamic content. In SWG Vader should all of the sudden attack a local town. In AoC you should spot Conan doing things, not theme park all these events. Also there should be 60% quests that are events or happenings not of your standard bring this there or kill X of these. Guild Wars II is promising something along these lines. Can they deliver?
That's it for now! I will expand, change, ninja-update this whenever I have time!