Warning: This post contains mechanical spoilers for Gravity Bone. If you don’t believe that’s possible and have yet to play the game I assume no responsibility…
In the comments to my previous article on the difference between literature, film and games, Chade questioned my use of the term explorative to describe games. He made two points and I’d like to address them both here:
1. Exploration typically implies that there is some sort of euclidean property in the game space such that strategies can be labelled “near” and “far”, and that “near” strategies are easier to see and reach than “far” ones. I don’t think this is a very accurate statement.
During play there are certain mechanics, certain abilities, that you are aware of and can use immediately and other mechanics that exist in the game space but of which you are unaware or unable to use at the current moment. Consider BioShock, during play you might have the Electrobolt Plasmid and be aware of the way it interacts with pools of water. The ability to electrocute enemies standing in pools is a visible strategy but if the conditions are not right (No Adam, no enemies, no water pools) you still cannot use that ability. It is visible but inaccessible, it is a mechanic you are aware of but one that is far from your current state. As well as Plasmids you might also have a pistol, and if you have enough ammunition you can use it immediately. It is a strategy that is both visible and accessible, it is near to your current state. As you navigate, explore the possibility space of BioShock your distance, in terms of resources and time investment required from particular strategies changes.
Chade’s second point seemed concerned with the differences between fixed and dynamic elements of a game:
2. Exploration typically implies that the player is in control of his movement around the possibility space. This completely understates the importance of the interactions between the game’s various components. In reality, the player is not able to grasp all the possible strategies available to him, and he is not able accurately predict the consequences of his.
If you visit a new city and choose to explore it, you might gain as much pleasure from the local colour, the people and interactions you witness, as you do from the architecture and layout of the city itself. Exploring the workings of dynamic systems is just as meaningful as exploring a fixed location. If that city is in a foreign country where you are new to the culture and social mores you could well find that certain behaviour has an unexpected impact on people you meet. Exploring how this might be the case, how different actions lead to different consequences is part of the appeal of exploration. Watching how people go about their lives in a culture you are new to is a form of exploration even if you have no control over how events play out.
There are numerous reasons why game are pleasurable, spectacle, or the simple joy of taking an action and seeing a response are two of the more easily identifiable ones. In mechanical terms a lot of that pleasure is derived from the self-guided exploration of the possibility space, we play with the system, the simulation, to see where the boundaries are, see how it works. The mental image we build up of the rules of the system and what they mean is unique to each of us. It is shaped, guided, influenced by the designed rules of the game but it cannot be directly controlled by such. Because our experience of game mechanics is a mental process of exploration and discovery it is entirely possible for that mental process, our own mechanical fabula, to be changed by external knowledge of the system. It is possible to have mechanics spoilt.
Mechanical exploration is a process of learning, we can be told the answer and the process of learning will still hold some of its appeal. However the reason for and experience of that learning process has been dramatically altered.
Consider Gravity Bone, I’ve heard a lot of people encouraged others to play this game and I second those recommendations, with one caveat. I would never tell anybody how long the game is. The reason for this is that doing so alters the explorative experience of the game mechanics; which I am about to do now.
During the game you are given tools to use which map to the number keys much like many other games. The number keys 1, 2 and 4 are used but 3 is not. When you play the game you see this and logically assume that you will get a new item that is assigned to the number 3; this never happens. Knowing that there is nothing new after we get the the item assigned to number 4 is not a story spoiler as it tells us nothing about the plot of Gravity Bone. What it is is a mechanical – or more specifically an interface – spoiler. If you do not know you have all the items you will expect more, you will believe there is still some area in the logical possibility space of the game that remains unexplored. Knowing that you never get further items changes the possibility space of the game as you are no longer expecting something else. You are unable to find this out on your own and the nature of your exploration has been radically altered, maybe without you even realising.
Let us return to that foreign city. Imagine you have a guide book with you, you can reference it to find out exactly what you’ll see around a particular corner. Does that change what is around the corner? No. Does it change the physical sensations of walking around the corner and witnessing it with your own senses? No. Does it change the nature of your exploration of the city? Undoubtedly; you have not found something new and formed a new connection between it and other parts of the city, you have located something you already knew about. That is a significant change in experience but it is one that happens so frequently that I wonder if any of us realise what we have missed.
17 replies on “Mechanical Exploration.”
I think the manner in which we experience stories and games is so similar as to be almost identical.
I’m taking an approach of examining them separately for the moment in order to hopefully make the similarities all the more apparent when I do directly compare them.
I’m wondering if this great discussion has actually moved well beyond a concept we can fruitfully call “the spoiler.” When we talk about spoilers, I think we’re almost always talking about taking away from the reader a single thing–the moment of surprise, whether it relates to a plot point or a mechanic (two things which I believe are actually one, but that’s a different discussion). Because people like surprises, they tend not to like to have them taken away.
But what we now seem to be discussing, still under a rubric “spoilers” is actually much deeper and more deeply implicated in the way games tell stories. For me, it amounts to the fundamental question–that aforementioned different discussion–of whether game-mechanics and game-story are in fact two versions of the same thing.
I actually just played Gravity Bone about 5 minutes ago, before I even read this post. While I knew going in that the game was short, I didn’t know how short. So I didn’t think anything of not having that inventory slot filled. I really don’t think anything was spoiled, though, because I knew nothing of what was going on. That’s a particular case, however, and if the same was told to me for a different game, my reaction might be different.
By the way, Gravity Bone is a very interesting game. I won’t say anything else in case someone who hasn’t played it reads this, but I enjoyed my time with it. If anyone hasn’t played it, do so.
Indeed, it is certainly possible to describe the process of learning about a game’s systems as an exploration. My question is: is it a particularly good word to use? I am not sure that it is, although admittedly my reasons are pedantic.
“The ability to electrocute enemies standing in pools is a visible strategy but if the conditions are not right … it is a mechanic you are aware of but one that is far from your current state… you might also have a pistol, and if you have enough ammunition you can use it immediately … it is near to your current state. As you navigate, explore the possibility space of BioShock your distance … from particular strategies changes.”
Let’s look at this in more detail. At one moment in time, you possess both a pistol and an electrobolt and are near a monster. Various strategies are available to you. As you have a pistol, you can attempt to strafe in and out of cover while shooting the monster. You can attempt to ambush him with your pistol from up close. You can attempt to circle strafe the monster while shooting him. These strategies are always available in close proximity of one another. They are not just “close” right now. They are always “close” to one another, and if they player is near one, then he is near the others as well. The player can drop his pistol and move away from these strategies. He can pick one up and move back towards them. But the position of these stragies with respect to one another stays constant. This is genuinely analogous to a possibility space which the player can explore, while the underlying structure of the space stays constant.
As it happens, the player also has an electrobolt. He can easily switch from the electrobolt to the pistol. So in that sense, these strategies are “near” one another – or at least, they are at this point in time. But if, at a later date, the player installs a different plasmid, then the electrobolt and pistol strategies will be very far from one another indeed. Now we have broken the analogy to physical exploration. We cannot explain what has happened by imagining the player moving around some structured space. Not only has the player moved away from the electrobolt strategy, but the electrobolt strategy has moved away from the pistol! Apparantly, the player is exploring in the midst of an earthquake.
“In mechanical terms a lot of that pleasure is derived from the self-guided exploration of the possibility space, we play with the system, the simulation, to see where the boundaries are, see how it works.”
This is obviously true, but is exploration the best term to describe this? For instance, what sort of exploration is it when the player is not able to backtrack? Let us imagine, now, that the player has taken advantage of some complex string of interactions to defeat this monster. He has frozen a big daaddy in the middle of an oil spill at just the right position so that a hacked turret on one side of the room fires at the monster, it also hits a splicer on the opposite side of the room, who then, while firing back at the hacked turret, hits an explosive barrel which sets the oil spill on fire.
When will the player get the oppurtunity to explore this particular nook of the possibility space again? He needs two big daddies at just the right position, an oil spill, a hacked turret, and a great deal of luck. Despite being in this position in the possibility space a mere fraction of a second ago, he is now an extremely long way away from it, and will likely never get to go back. If he is moving around a possibility space, then he must be travelling very fast. And yet, many other things about the players position in the game have not changed at all.
Of course, you can make your concept of “possibility space” more abstract. Perhaps each point now indicates some abstract “way of playing the game” that can be applied in every situation within the game. But I think it is very naive to say that the player is thinking and “exploring” at that abstract level during play. In reality, players are caught up in the specifics of the decisions they are making on a moment to moment basis.
The word exploration brings to my mind a picture of the player navigating over a stable ground of possible strategies, briefly viewing each patch of ground from far away as he mulls the possibility, before closely examining each patch of ground as he tries various strategies out. The player moves around this space with intent, and is roughly free to choose which pieces of ground to explore. The structure of the possibility space, meanwhile, stays obligingly stable as the player analyses each patch of it.*
While this is true on a local basis, it breaks down very quickly. In reality, the structure of the possibility space is anything but stable. The player has some ability to move with intent, but his intent is commonly foiled by the dynamics of the game. The player moves quickly away from some parts of the possibility space, while simultaneously staying completely still with respect to other parts. This idea of play as exploration may be broadly true, but the specifics can be completely wrong.
~~~~~~~~~~~~~~~~~~~~~
* Of course the rules stay stable while the player analyses them. This is true, but if we are imagining the possibility space as something that the player navigates, then the rules cannot be the medium in which he moves. Every rule applies to the player at every point in the game. The player does not move at all with respect to the rules of the game.
Woah… I appear to have been beaten.
I could counter by saying that I’d consider the possibility space as a dynamic n-dimensional construct but as well as sounding even more pretentious than usual I don’t really think it helps. Suffice to say I never considered it to be in any way analogous to a physical landscape, which is why I brought up the concept of a city teeming with people. In such a location specific circumstances can arise entirely outside of your control yet you are still relatively free to explore.
I accept the term explorative has problems, I originally preferred experimental but that seemed to have more adverse connotations. Still I think either term is more useful, in terms of encouraging discussion and analysis, than the somewhat fuzzy term “interactive”.
Exploration certainly does a good job of describing some aspects of playing a game. To make my points, I am having to be very pedantic … perhaps more pedantic then is usefull. Also, I am taking my own interpretation of the term as gospel. Perhaps that is wrong.
And as you say, we would not be having this debate had you used the term “interactive”.
I must admit I am actually struggling to come up with a word that I really like. Perhaps “investigative”?
Agency driven?
@Moromete: In response to your comment in the previous thread I’ve tried to explain why, at least for the moment, I’ve avoided using the term agency.
Though it is flawed I’m going to stick with explorative for the next few posts as I believe it opens up interesting possibilities for analysis that I hope to develop further over the coming week.
I plan on expanding on the specifics of the possibility space in a post I am currently working on. At the moment my current method for visualising the possibility space is as a directed graph, with objects as nodes and interactions as arcs.
I was being accurate when I described it as a dynamic n-dimensional construct. Dynamic being that it is able to change over time, often beyond the direct control of the player. In terms of its dimensions I was originally thinking in terms of each object being a separate dimension (including the player), with axis represent ting interactions, however that quickly becomes complex and requires some degree of folding and self intersection to work properly.
After a bit more thought, I don’t think I really have a problem with the term exploration … it is a natural word to use, once you talk about a “possibility space”.
And that is my real problem. The words “possibility space” get bandied around so often, I have certainly used them many times, and yet I don’t think I have ever seen anyone precisely define what they mean.
Can we define the possibility space using rigorous and specific concepts?
For starters, I have basically been arguing that the possibility space is not a space at all – at least, not in the traditional sense.
What is the possibility space made of? What does it mean to say that the player is in a particular position in the possibility space? Is it a strategy the player is using? Is it the actions occuring in the world? Is the the subset of rules that the player is focussing on at this moment?
What properties does the possibility space possess? Is it really a space? Perhaps it is a different type of set?
You suggested that it might be a dynamic high dimensional space. Is that a specific suggestion worth following up in a rigorous way, or is it a hand-wavy argument that stymies further debate? What are the dimensions? In what way is the space dynamic? I don’t think that the properties of high dimensional spaces will invalidate my critisicms, but it sounds worth following up. At least it would be more interesting than saying “possibility space” and leaving it at that.
My current working version, which should appear in next post, is starting off as the simplified directed graph I described. Even in that form it can be examined and analysed in interesting ways.
I do plan to built on that base concept but at the moment working out a representation of the possibility space in a form that I can actually visualise and work with has been challenging enough for my sleep deprived brain.
The form you seem to be thinking of is that of MinMax Trees, the prime example of their use being in analysing Chess moves. They are very useful but get prohibitively large very quickly, making them difficult to represent and analyse in any concise fashion.
A directed graph sounds like a much more promising concept then the euclidean space I have been envisioning.
But why is each object a node? Wouldn’t a node represent the entire set of “interesting” game state (possibly including the history of the game and the players own motivations and biases)?
Given a specific game: what is the optimal set of information to store in each node, so that the graph is as “informative” as possible, in terms of watching the player navigate the graph and making “meaningfull” comments about his progress.
(Whatever “informative” and “meaningfull” mean … so much for me being rigorous …)
Yeah, you can’t reasonably include the entire game state in each node … you to store less information, so that the player can visit a single node multiple times.
That is the problem I was getting at when I wrote “what is the optimal set of information to store in each node, so that the graph is as “informative” as possible”.
Luckily, most computer games have some core gameplay which is repeated again and again, with the previous stage having a minimal impact on the next stage. This contrasts pretty strongly with a game like chess or go.
Anyway, I am interested in what you will propose …
Thanks Chade for setting a precedent. Now there’s no need to ever worry if my comments are too long ;)
I sometimes find it helpful to summarise these discussions, [and please note that I’m free with my paraphrasing and inserting own interpretations, so please excuse]:
– In the previous two posts CrashT looked at the way games communicate meaning, as compared with other media, and pointed out a role for what he called “exploration”.
– In this post two aspects of “exploration” are explained [to be referred to as E1 and E2].
_____(E1) Exploration involves the player investigating the gradient of potential actions (i.e. certain pre-actions make other actions more or less possible)
_____(E2) Exploration involves the player investigating the “possibility space” of the game. This term is briefly explained as being the mechanics of the game.
– Chade then comments [please excuse me for crudeness]:
_____The spatial analogy of E1 is flawed, for although particular actions may be near/far, this measure need not stand between actions. I.e. because my relationship to each action may be independent, there is no “space” to speak of that may be coherently explored [in the E1 sense].
_____E2 contradicts the role of narrative (in various games) of restricting our option, or otherwise making unlikely, of exploring a particular event/action. The dynamics of gameplay (similar to what was described for E1) make the analogy of space incoherent.
– The discussion then continues with CrashT suggesting a more definite definition of “possibility space” and also suggests that perhaps the term “interactivity” may be more appropriate.
I’m given to understand that the term “possibility space” comes from its use in math, where, per WolframMathworld, it means “the set of all possible values the events may assume.” In other words, its all the different things that could happen during a game.
[I henceforth abbreviate “possibility space” as PS]
– But a starting definition is barely a start, and Chade notes that it remains to be made clear how it relates to ideas like “position within PS” or “strategy within a PS” or “focus within a PS” etc. And also, “how does it contribute to the conversation about games?”
– CrashT suggests analogising PS as a dynamic n-dimensional construct, and apparently its “TO BE CONTINUED” in the next post.
Now I want to make a few points of my own:
1. Don’t loose focus. Look at the beginning of this comment. Your original aim was to look at the “form” of games. Will considering PS as ‘a dynamic multi-relational array of potential game-actions’ assist you in demonstrating the difference between games and other media? I’d be wary of producing a term [viz. “PS”] that while coherent, is too broad to be useful.
2. I wouldn’t recommend using “interactivity” here. It’s a bit of a loaded term, being used in critical theory to describe what the audience brings to an experience/interpretation of an artwork.
3. @ Chade: Obviously you didn’t elaborate, but regarding your last post, are you sure the “minimal impact on the next stage” clause doesn’t contradict games like Civilization?
4. Above Chade noted that___”Despite being in this position in the possibility space a mere fraction of a second ago, he is now an extremely long way away from it, and will likely never get to go back.”___i.e. that gameplay often restricts the possibility of actually exploring all the possibilities. I won’t elaborate, but compare the way a casual gamer and a “hardcore” gamer [e.g. the Quake fanatic] may play the same game. Maybe this phenomena, call it “intricate exploration of PS” may be useful for discussion there.
Yep.
Two more things actually:
5. I have no idea why my comment went all italics half-way through.
6. Again: don’t loose focus.
I think a definition of a PS like the one you’re attempting is useful for elaborating on the “meaning” of a game like Tetris. E.g. the PS of Tetris is as follows…. …and this PS invites the player to enjoy a familiarity and mastary of physics and pattern generation.
But what about a game like follows [that I just made up]:
– Cuscene showing princess being kidnapped.
– Action sequence requiring avatar to travel from one side of the screen to the other.
– Cuscene showing princess being rescued.
This game still allows exploration, e.g. the different speeds at which the avatar crosses the screen, but it is essentially a movie interupted by a “player-achieve-X-to-continue” screen.
Would you deny that this is a game and call it a movie with a break in the middle? I’d suggest that what makes this game “work” [in my fictional world] is the fact that the player feels “responsible” for saving the princess (a responsibility, note, which a movie watched could not enjoy).
I.e. PS may often not be helpful in describing how meaning is communicated in games.
I’ve fixed the italics issue.
I don’t want to present some false hope for my coming post; it does not present some grand unified theory of visualising the possibility space. What I do attempt however is to use a simplified version of a relational graph to examine the inherent meaning in the mechanics of different games. It is a first step and one I plan to continue.
My aim is not so much to look at how games differ from other media so much as how they are similar, in that they all attempt to communicate a concept, idea or emotion.
Just a quick note: Pala, you asked “are you sure the “minimal impact on the next stage” clause doesn’t contradict games like Civilization?”
Sure it does … and plenty of other games besides. But something you mentioned in your own post, and CrashT mentioned in his, is that there’s no point introducing a theory that is too broad and abstract to be usefull.
Gameplay is so complicated and varied. When a notable set of games come along, with some nice properties that make it easier to analyse them, then I think you want to hang on to that “niceness” with all your strength, and damn the games that don’t happen to fit into the paradigm.
As it happens, the paradigm of some repeatable core gameplay, where each stage has little impact on the next, is really very common in the computer game industry in particular. Whether that is enough “niceness” to let you analyse a modern computer game is another question …