I've always been very much interested in game design. Most of the time I experience game design from the point of view of the consumer, the player. But some of the important game design questions from the developer's point of view do pop up when I am the dungeon master (DM) in a pen & paper roleplaying campaign. And maybe the most fundamental one of them is the question of whether to be gamist or worldist in your design.

Gamist design considers the game to be first and foremost a game. It doesn't matter what the numbers represent, but it is important that the game achieves a state of "flow", where every single encounter is perfectly balanced to be neither too easy, nor too hard.

Worldist design considers that everything in the game represents something in the game world, and that it is most important that this world is believable. There need to be interesting decisions which result in real consequences.

Most games are a mix of the two. Some people deplore that the gamist design philosophy is on the rise in video game design, leading to games in which you always win, always progress, and never experience setbacks. But there is also a backlash of extreme worldist game design, with unforgiving sandbox gameplay with very little flow. Neither gamist nor worldist game design is wrong, the eternal quest is to find the good balance between the two of them.

In terms of a DM of a pen & paper roleplaying game, some of the choices are inherent in the system you play. For example 4th edition Dungeons & Dragons is very gamist, and a character or monster can be at the same time frozen and on fire, poison gas affects some monsters who obviously don't breathe, and you can strangle a gelatinous cube. But there are still gamist vs. worldist design decisions to make in the way that you run an adventure.

The gamist view of a D&D adventure is that the story is a thinly veiled excuse for players to go to a dungeon or similar place and have a series of perfectly balanced tactical combat encounters. Some DMs go so far as to fudge dice rolls when things go wrong, so as never to kill player characters. The worldist view is that the DM represents a fantasy world which has to be believable and logical, and there have to be consequences to the decisions of the players.

On the one side Sid Meier once said that a good game is a series of interesting decisions. On the other side in order for interesting decisions to exist, there needs to be visible difference in outcome between the consequences of each decision. Which means that if the players make a "bad" decision, logically the outcome has to be bad as well.

A typical problem here is if the players know about a location where a large number of monsters reside. They could decide to approach that location carefully and try to find smaller groups of monsters to attack, or they could go with what one of my old groups called "Plan A" and just do a frontal attack. It is inherent to the system design of D&D that fighting a number of smaller groups is easier than fighting all of those monsters at once. So in a worldist design the players have a choice between several easy fights or one big and hard fight. A gamist design would want to avoid that situation, because the philosophy is to only have balanced fights. To some degree dungeons are inherently gamists, because outside a few rare occasions the monsters in a dungeon never cooperate and are usually encountered small group by small group, with nobody ever raising an alarm. But once you play an adventure outside a dungeon setting, more worldist consideration tend to appear.

Personally I am principally opposed to fudging dice and guaranteeing a good outcome of any encounter regardless of player decisions. For me decisions need to matter, or they aren't fun. But I would try to keep the encounters in a certain range: A good decision might lead to an easy encounter, but which still isn't trivial; the bad decision would lead to a hard encounter, without that being an automatic total party kill or frustrating.

The main problem is how to signal that to players who are gamist in their personal expectations. The people going always for Plan A, frontal assault, are those who had the experience of gamist design in which Plan A still leads to a balanced encounter which the players would win without too much trouble. So why come up with a complicated plan if you expect equally balanced encounters regardless of what you do? On the other hand if every decision has too serious consequences, the game stalls into an endless discussion because players don't dare to make a move in fear of that being the wrong move. How does one keep the game going without making decisions either terrifying nor inconsequential?