On Possibility Space

Posted by The Architect on February - 13 - 2024

I’m pretty certain the concept already exists in the world of game design theory, probably under a different name; all I found on a (not too thorough) search was this (https://www.possibilityspace.org/tutorial-generative-possibility-space/), which comes close, but doesn’t quite fit. Especially not since I wish to address that topic from the specific perspective of a CRPG designer.

So what am I talking about? Well, rather than trying to come up with a definition, I’ll start with a formula:

[Range of Possible Events] x [Gameplay Impact]

Let’s disassemble this by way of a few examples.

We have a narrow corridor. There are no branching passages, no doors, no events of any kind, it’s just straight forward linearity. That’s our minimum possibility corridor, and a series of this would probably not lead to a very exciting game.

Let’s change it to a catacomb hallway. It still goes on and on, but now includes a number of events – maybe a few tombs in side niches. Yep, let’s say there are ten tombs altogether.

Now, if each of these tombs holds the same bone dust in every instance of the game where we walk the hallway, and this bone dust has zero relevance in the game otherwise, we haven’t really expanded the possibility space. (I hope this sentence gives an outline of the picture I’m trying to paint.)

At this point, some ideas are pretty obvious: The tombs could hold loot, for instance. We could give each tomb a random loot table for a wide Range of Possible Events. Then to better illustrate the element of “Gameplay Impact”: What happens if these loot tables only list historic weapons, each of which is surpassed in power and efficiency by the standard weapons in the next weapon store? Obviously, the Gameplay Impact is minimal, unless these historic weapons carry ridiculous value because old. Or something.

Now, cranking the Impact up to 11 (or 1111) also isn’t all that hard, at least in theory; just have the loot table include various items so powerful that each could completely change a character’s build. And of course someone would be buried in these tombs, maybe a lowly skeleton, maybe an insanely dangerous wraith king. Enough space yet?

So, now the first question: why should I, as a designer, even care? And I have indeed seen it argued that it doesn’t much matter as long as you’re only making one playthrough of a game. The possibility space for that run would be equal to exactly the one possibility defined by the designer, but if you’re seeing things only once, who would know the difference?

Well, first, I don’t think it works that way in practice. Somewhere along the run, we get saves and reloads, and those are usually a very clear opportunity for the player to notice how much of a possibility space exists for the content of a chest, or the loot dropped by certain enemies. And of course in our time, you can count upon players quickly gathering and exchanging all information they can find…

Apart from that, though, I think it already shows in the general design. The larger the possibility space, the more the designer has to take into account that the player has certain items, abilities, companions etc. available – or, vice versa, does NOT have access to them. This practically demands a more open approach to the design of the challenges, which should flexibly allow for a variety of ways to deal with them. In other words, expanded possibility space gives the designer less control over the actual gameplay events, which in turn gives more agency to the player. And the designer has to account for that. (Yes, this means work.)

I’m aware that a game with limited possibility space could, in theory, just as well offer a variety of options for dealing with its challenges. However, I think that in a completely predefined environment, designing the challenges such that they can be dealt with in multiple ways only really makes sense if there is a chance that the player misses these alternatives. Otherwise, what’s the point? Hey player, here are three guns with unlimited ammo, choose one of them to kill the sentinel robot.

If you want a more practical example, in particular from the CRPG world, please take a look at the well-known Lands of Lore: The Throne of Chaos. The order in which you visit level blocks is fixed. You have limited availability to wander back and forth within a level block, but do so usually only because the puzzles demand it, which are also always the same. Party members come and go as they please, always at the same places, without input by the player. Loot is fixed. Apart from the “character generation”, which is limited to a selection from four predefined dudes (no gals either), every player, on every playthrough, will experience basically the same journey.

I’m not saying that makes Throne of Chaos a bad game… or am I? In fact, that depends on your definition of “game”. As outlined above, the journey through Lands of Lore is predefined in almost every aspect. Accordingly, the player has very little agency in the whole undertaking. Personally, in such a case, I don’t have the feeling of “playing a game”, but rather following a story. This can be fun, too. It just feels a little shallow, and it certainly doesn’t captivate me, doesn’t have me thinking outside the game about what choices I could make, what alternatives there might be. Put differently, I’m not invited to imagine the possibilities if I have a sense that the actual possibility space is limited. When I think about the games that excited me most, that occupied my mind the most, then I always come back to the games that opened up such possibility space.

Of course, creating possibility space certainly isn’t just a matter of randomness. “A chance that a player can miss” alternatives could also be implemented by creating an environment so detailed, complex and diverse that a player almost certainly will only find a subset of the possible tools; and in all likelihood, different players will find different secrets depending on their play styles, making predictions somewhat difficult. Further, there’s the simple matter of player decision, pretty well exemplified in chess: Every piece is known, the start is always the same, yet every decision has the potential to lead down different branches, and to have tremendous gameplay impact. Mind = blown. For a more basic example – no thief in your party? Well, then you can’t pick that lock. (Our job as designers would be placing a backdoor as well.)

The very first Deus Ex is an excellent example of magnified possibility space resting on player decisions and complex environments, but little randomness. The player won’t have enough skill points to learn every ability; specialization is required. You can only choose one alternative for each available cyberware slot, and all of them have significant Gameplay Impact; again, the choices of the player define the possible paths through the game. Obviously, this approach leads to significant balancing challenges. The Deus Ex developers accordingly made sure that for each problem there would be multiple ways to solve it, usually including a hidden one that would not require any kind of skill or equipment to take. Magnificent.

This is certainly not an exhaustive treatise on the matter. I hope I could give you some food for thoughts, though. Maybe think about the games that impressed you the most – did you have the feeling that everything could happen? And as designers, how do we evoke that feeling best?

Happy crafting!

Categories: Allgemein

Comments are closed.