For a long time this blog had the motto "MMORPGs should be as credible and consistent as possible and as little as necessary". In my opinion, developers did not try hard enough to make MMORPGs consistent with the simulation. I changed this motto a few weeks ago, but the reason was that I wanted the freedom to blog about a bit more than just MMORPGs.
I still think that "MMORPGs should be as credible and consistent as possible and as little as necessary". But recently I found myself on the other side of that battle. I tell people that what they want to do is a part of the "[..] and as little as necessary".
Before I go on, I'd like to talk a bit about "(abstract) gameplay" and "simulation". This distinction is at the heart of this debate. Every game has abstract gameplay. It is what is left after you removed all the "fluff". From the gameplay PoV it doesn't matter whether I shoot my rifle and hit you for 3 dmg, or whether I shoot a fireball and hit you for 3 dmg. It's exactly the same. I could just as well throw leaves at you for 3 dmg.
Most older games are very abstract. Take Chess or Poker. The simulation is there, but it is very weak. In Chess the simulation is war, in Poker the cards just kept their names (queen, king, ..) after hundreds of years of iterations. There are reasons why the Ace is a powerful card generally and not the 9. And these reasons are found in the simulation, because from the gameplay point of a view it really doesn't matter.
As time passes, players reduce games to their abstract gameplay; it is often called the "meta game". Players try to ignore everything that is "fluff" and concentrate on how to achieve their goals. And that's why the abstract gameplay is important if you want to keep players playing. Because if the gameplay is bad, the game does not last long. Even if you had the most wonderful graphics.
From a publisher's point of view, you can say that the simulation is the more important the less a player knows about the game. And the gameplay is the more important the more a player knows about your game. Unfortunately the two not always agree. E.g. from a simulation point of view wooden houses should burn down when you throw fireballs at them. And it's quite possible nowadays to technically make it so. The reason it is not done is usually for gameplay reasons.
What I have been arguing repeatedly on the blog is that when facing a problem like above, one should try to find a solution that combines good simulation and good gameplay. Current AAA companies don't try hard enough, in my opinion. And they are especially venturous when it comes to harming the simulation with the business model and for no gain in gameplay, Psychochild.
But let me be very clear: If you can't find a solution that combines good gameplay and good simulation, gameplay first!
It is important to understand that the pure simulation often isn't all that hard to code. If Blizzard wanted to not have you buy water after you killed the god of death, they could. If they wanted a combat AI that is more consistent with the simulation, they could. They wouldn't need a sophisticated combat AI to do that. E.g. they could just remove threat modifiers for tanks.
When I tried to figure out how to make MMORPGs more consistent with their simulation (of a fantasy world), I found very early that I would first have to understand why a game like WoW can become the most successful MMORPG of all time. After all WoW is quite terrible at the simulation. They ignore it all the time. And still even I played that game for years. Can't have been that bad, can it!?
You can call that "understanding your enemy". Before you try to beat somebody at something, you need to understand what makes him strong in the first place. Yesterday, on Azuriel's blog, I told him that if he doesn't understand why millions of non-raiding players played the lvl70 endgame of TBC for literally years, he shouldn't try to argue that the changes in WotLK were for the better. Especially since the statistics don't really support that assertion. If you don't understand why your car works, don't dabble with the engine.
A lot of players think that it would be very cool to have player homes and castles to fight over and actual trade and many more things. Count me among them! But if you want to make a game that introduces these features, you better make sure you understand why current games don't have them! Because, Blizzard isn't stupid. They, too, see what players would love. There are reasons why they don't always do what players want (Thankfully. - They cave in much too often since WotLK).
That doesn't mean that you should clone WoW when you make a new MMORPG. Please no! But it does mean that you need to understand Blizzard's reasons for not introducing open PvP and battles over player-owned castles, before you implement it yourself. If you think that Blizzard is just too stupid to see this opportunity, then you are the fool.
I am going to blog a bit more on storybricks in the future, I guess. I wanted to keep this post abstract.
If you want to improve NPCs' consistency with the simulation in games like WoW, you first need to understand why current AAA companies didn't already do it. Because NPCs could be made more life-like very easily with a little scripting. There are reasons the companies don't do this and rather overhaul every single class mechanic (no, the team is not lazy. But the management should try to invest that $100 million monthly profit instead of buying stocks back, in my humble opinion).
If you assume that the only reason they didn't improve NPCs already is that it's too expensive (=too risky), I disagree. This plays a role, yes. But it is not the only reason they are careful. They also have no idea about how to get good gameplay out of it. If you can solve that problem count me among the enthused.