Let's talk about goals in storytelling games. Not really the far-reaching "save the princess" or "kill the evil dude" kind of goals, but the local and moment-to-moment goals that face a player throughout the experience. I have sort of touched upon this in the scene-approach to high level story telling story design post, but want to discuss it a bit further. I think this is another major reason why there still a need for either violence or puzzles to drive the story forward. The reason being that the player does not know what they should be doing otherwise.
In a non-interactive story the characters can behave in certain way because it works for the narrative. They think about the things that are relevant to the story being told, and perform actions that have interesting outcomes. In an interactive work, it just does not work like this. In order to control outcomes, the player would basically have to know the future of any action, something that is neither possible or desirable. Thus, the game will have to guide the player into making the actions that the story requires in order to give an engaging output. This is where violence and puzzles come into play.
Games based around violence teach player the following: You need to kill anything in sight and when you run out of things to slaughter you need to progress until you find more stuff to fight. Once the player accepts this the problem with having is non-issue and the role of a designer is to weave a story into this mode of progression.
Puzzles have a different set up. Here the whole idea is to constantly bring up riddles for the player to solve, and then create a story out of that. What gives the player goals and directions are the requirements of the currently encountered and unsolved puzzles.
A more abstract and direct version of this is simply to have a certain goal artifact and then evolve the entire game about retrieving this. Portal, Braid, World of Goo, etc are all examples of this approach.
Puzzles and combat are the most popular ways of settings up goals, but others exist as well.
Most platform games, like Super Mario, just have the player walking in a single direction. While these games tend not have much story content, there are more experimental games that do. Passage, One Chance and Everyday the Same Dream use this approach quite directly and are sort of very simplified platformers. I think Dear Esther and Journey can also be said to use this kind of approach, as the player does not really have any other goals than moving in a certain direction.
Another approach is to not give the player any explicit goals at all, but let them just interact until something interesting happens. There are not really any longer games that uses this method to implement player progression. The Path and some of Vector Park's games do it to a certain extent, but then only for a single specific scene. Adventure games use in limited sequences, like when being thrown into a new location and forced to explore, but never for any longer stretches. The problem with this approach is that the player cannot really make any plans, which brings down the sense of agency and engagement quite a bit. This makes this only work for short bursts, often when a sense of disorientation is appropriate.
Finally, I have to mention the on-rails method, which is essentially what Walking Dead and Heavy Rain do. It is sort of similar to the "walk this way" approach of platformers, but just removes the required interaction for forward movement. These games drags you along whether you want it to or not, only letting you interact in very small and specific situations. An interesting aspect is that the "interact until something happens"-approach can work quite nicely here, partly because there are often relatively massive amounts of exposition before each interactive moment. This combined with a closed of scene makes it possible to set up a goal using purely plot means.
This pretty much sums up how any storytelling games goes about creating low-level goals.
When starting out the new Super Secret Project I was very much into "interact until something happens"-approach, but it did not really work. The lack of gaminess did not make the player more into the story, but created frustration and made them spend most mental energy pondering "what the hell am I supposed to do?". Our current approach is instead to use a combination of puzzles and the "walk this way"-approach.
Puzzles tend to always give certain feel to the environments; machines to boot up, broken bridges to cross and that sort of thing. This limits the range of goals quite a bit and is often quite evident in games. For instance, in Amnesia: The Dark Descent there is always the need to open some form of door. The challenge here is to be creative of course, but it is a very hard problem. To make puzzles out of various situations is one of the biggest challenges we face.
I think it is also very important to recognize that a big part of puzzles is to provide goals. Starting with Amnesia, we stopped seeing puzzles as challenges and instead view them like interesting activities. Focus is put on making them engaging and fitting to the narrative, instead of (as was the case the before) making them challenging. One cannot remove the challenge entirely though, because then a certain immersive quality of the puzzle is lost. There needs to still be a certain amount of "revelation" taking place in order to feel as if you are really making a connection with the game's world.
The "walk this way"-approach is very interesting as it gives much more freedom in the kind of environments that can be used. Now you can place the game in just a about any situation without any need to figure out ways to use it gameplay wise. The main issue is that you need to make your environments very linear. For the approach to work the goal must always be very clear, else it turns into a puzzle. In order to keep players engaged, it is also important that there is reason for continue going in a certain direction. This can either be the promise of some reward when getting there, or a steady flow of interesting things happening along the way.
I am unsure how long a game can be and what kind of stories can be sustained by only using the "walk this way"-approach. All the current ones (that I know of) are quite short. Interestingly, the more complex and direct the story (like One Chance and ImmorTall) the shorter, and the more abstract and vague (Dear Esther and Journey) the longer. This might just be by accident, but might also be a sign of some kind of limit to the approach. Worth nothing is that compared to other approaching, there is little inherent engagement in this one. Simply moving forward simply stop being interesting after a while and something else is needed.
I think the question of various ways to set up goals is a really important issue but I do not see it addressed very often, or really at all. For some reason any design articles I come across are based on a type of design and then just take that as dogma. Perhaps I am just missing all the nice papers/articles out there?
Also interested in hearing if I missed out on any ways to create the low-level goals in a storytelling game!