Without Readability - The Decline of Adventure Games
At long last, we have arrived at something tangible. Dispensing with hand-waving theory, I'm going to discuss how the murky readability of adventure games contributed to their falling from prominence. In retrospect, I should have presented this sooner. If this series is compiled into a set of links, I'll likely pull a C.S. Lewis and renumber this as part two and follow it with the posts about empathy and data.
Before I launch in, I need to reiterate that I have a great fondness for adventure games. Some of my fondest gaming moments came when playing adventures. I want to be able to share those experiences with folks who haven't, rather than pine for halcyon times long past.
It's clear adventure games, of both the text and graphical variety, have lost their once prominent position. There are a number of reasons for this, but one that seems largely ignored is the poor readability of the gameplay systems of adventure games. Sometimes this will be addressed by people saying, "they were too hard." But that claim is actually quite misleading, especially given the notions most have about difficulty in video games.
Adventure games suffered from poor readability. Full stop. The mechanics of adventure games provided a truly colossal number of possible actions, but with no real relationship to each other. There was no systematic way to understand or improve at playing adventure games, beyond learning that you should try to pick up every single thing in the game that isn't nailed down (and those you'll probably just have to pry up).
As a quick terminology note, Jesse Schell mentioned this very issue in his book when discussing player actions as a game mechanic. He identifies "operative" actions as distinct from "resultant" actions. The former would be fundamental actions the player can execute at any distinct moment, while the latter are broader strategies or behaviours that emerge from understanding how the game's systems inter-operate. This is remarkably similar to the mechanics vs. dynamics distinct in the MDA framework, and it seems both are describing the same phenomena. I'll be using the MDA terminology here.
Back to the readability problems in adventure games. Take the quintessential LucasArts/Sierra adventure game. The main interactions in the game consist of using a number of verbs (e.g. push, pull, pick up, talk to, use, etc.) with various objects in the environment. Additionally, your character likely has an inventory that allows you to move some objects from one place to another and use them on other objects.
Unfortunately, they way these verbs operate is based on a carnival mirror version of the real world. Pushing object X does nothing, but pushing object Y will get you closer to a solution. Trying to pick up one object results in the player's character saying "I don't want that," but the character is perfectly happy to pick up a different object.
This provides the player with a tremendous number of possible actions to perform, but no way to understanding how they work in conjunction (because they almost never do). There are a lot of ways to use the game's mechanics, but no enjoyable dynamics arise from them. If the player isn't able to infer the designer's intentions, their only option is trial and error. But the permutations of the various mechanics result in a truly colossal number of actions, only a tiny fraction of which have meaningful outcomes.
As a hypothetical, say there are six areas a player can go during a certain stage of an adventure game. On each screen there are an average of eight interactive objects. The player has five possible verbs they can use on any object. They also have seven items in their inventory, all of which can be used on any object. This means that at any given time, the player has 768 potential actions to perform.
Text adventures (or interactive fiction, whatever your preference on terminology) exacerbate this problem, as the limit to possible verbs is literally the player's vocabulary. When stuck, the player becomes a thesaurus, struggling to find the right combination of words to execute what they think makes sense. But as it's impossible to know when an action is incorrect due to logic or just due to the parse, the player can never be certain their solution was fundamentally incorrect, instead of just linguistically incorrect.
Compound this with the truly absurd logic of some adventure games (see Old Man Murray's discussion of Gabriel Knight 3) and it's really no surprise that adventure games fell out of favour. The dynamics of adventure games are not enjoyable the moment the player falls out of sync with the designer's intentions. Coming full circle, adventure games have a readability problem. With all that being said, this definition of readability should make sense:
Readability is a measure of how easily a player can understand a game's dynamics.
To provide contrast, Braid is a good example of a game with a small number of mechanics giving rise to rich, understandable dynamics. Some criticized Braid for requiring the player to guess the designer's intent. While this is true, the mechanics in Braid number in single digits. Personally, I was never particularly frustrated with Braid, nor did I ever feel like a puzzle was obtuse or unfair. Upon solving, my reaction was universally, "Oh yeah, that makes sense." I cannot say the same for needing a cat hair and syrup mustache to impersonate a man that doesn't even have a mustache.
This post has become much longer than intended. In the interest I've brevity, I've broken it in two. Wednesday's post will discuss some ways in which adventure games can be made more readable.
Before I launch in, I need to reiterate that I have a great fondness for adventure games. Some of my fondest gaming moments came when playing adventures. I want to be able to share those experiences with folks who haven't, rather than pine for halcyon times long past.
It's clear adventure games, of both the text and graphical variety, have lost their once prominent position. There are a number of reasons for this, but one that seems largely ignored is the poor readability of the gameplay systems of adventure games. Sometimes this will be addressed by people saying, "they were too hard." But that claim is actually quite misleading, especially given the notions most have about difficulty in video games.
Adventure games suffered from poor readability. Full stop. The mechanics of adventure games provided a truly colossal number of possible actions, but with no real relationship to each other. There was no systematic way to understand or improve at playing adventure games, beyond learning that you should try to pick up every single thing in the game that isn't nailed down (and those you'll probably just have to pry up).
As a quick terminology note, Jesse Schell mentioned this very issue in his book when discussing player actions as a game mechanic. He identifies "operative" actions as distinct from "resultant" actions. The former would be fundamental actions the player can execute at any distinct moment, while the latter are broader strategies or behaviours that emerge from understanding how the game's systems inter-operate. This is remarkably similar to the mechanics vs. dynamics distinct in the MDA framework, and it seems both are describing the same phenomena. I'll be using the MDA terminology here.
Back to the readability problems in adventure games. Take the quintessential LucasArts/Sierra adventure game. The main interactions in the game consist of using a number of verbs (e.g. push, pull, pick up, talk to, use, etc.) with various objects in the environment. Additionally, your character likely has an inventory that allows you to move some objects from one place to another and use them on other objects.
Unfortunately, they way these verbs operate is based on a carnival mirror version of the real world. Pushing object X does nothing, but pushing object Y will get you closer to a solution. Trying to pick up one object results in the player's character saying "I don't want that," but the character is perfectly happy to pick up a different object.
This provides the player with a tremendous number of possible actions to perform, but no way to understanding how they work in conjunction (because they almost never do). There are a lot of ways to use the game's mechanics, but no enjoyable dynamics arise from them. If the player isn't able to infer the designer's intentions, their only option is trial and error. But the permutations of the various mechanics result in a truly colossal number of actions, only a tiny fraction of which have meaningful outcomes.
As a hypothetical, say there are six areas a player can go during a certain stage of an adventure game. On each screen there are an average of eight interactive objects. The player has five possible verbs they can use on any object. They also have seven items in their inventory, all of which can be used on any object. This means that at any given time, the player has 768 potential actions to perform.
Text adventures (or interactive fiction, whatever your preference on terminology) exacerbate this problem, as the limit to possible verbs is literally the player's vocabulary. When stuck, the player becomes a thesaurus, struggling to find the right combination of words to execute what they think makes sense. But as it's impossible to know when an action is incorrect due to logic or just due to the parse, the player can never be certain their solution was fundamentally incorrect, instead of just linguistically incorrect.
Compound this with the truly absurd logic of some adventure games (see Old Man Murray's discussion of Gabriel Knight 3) and it's really no surprise that adventure games fell out of favour. The dynamics of adventure games are not enjoyable the moment the player falls out of sync with the designer's intentions. Coming full circle, adventure games have a readability problem. With all that being said, this definition of readability should make sense:
Readability is a measure of how easily a player can understand a game's dynamics.
To provide contrast, Braid is a good example of a game with a small number of mechanics giving rise to rich, understandable dynamics. Some criticized Braid for requiring the player to guess the designer's intent. While this is true, the mechanics in Braid number in single digits. Personally, I was never particularly frustrated with Braid, nor did I ever feel like a puzzle was obtuse or unfair. Upon solving, my reaction was universally, "Oh yeah, that makes sense." I cannot say the same for needing a cat hair and syrup mustache to impersonate a man that doesn't even have a mustache.
This post has become much longer than intended. In the interest I've brevity, I've broken it in two. Wednesday's post will discuss some ways in which adventure games can be made more readable.
Labels: adventure games, design, readability
posted by Nels Anderson at
5:00 AM
17 Comments:
In a way the "problems" you describe are exactly what defines the Adventure Game. Problems are there to be solved and the way to solve your readability problems is to play the game.
I would argue that 'guessing the designers intentions' is the main mechanic behind most inventory-based puzzles. If everything was obvious from the outset then they wouldn't be seen as puzzles at all.
Take the Times Crossword - Notoriously impossible for the beginner. But they are solvable. Anyone who plays the Times Crossword will know that the way to solve it relies less on your own general knowlegde and more on your ability to get inside the compiler's head.
That is the puzzle.
From my perspective, adventure games are making a big comeback in the casual space. Hidden object games are slowly becoming more and more like adventure games, with some straight up adventure games selling well on the portals.
On my part, I think the criticisms you level are mainly issues with bad adventure and IF games. A good adventure game will mitigate them significantly. It'll give good, intelligible responses to "incorrect" actions, and in the case of IF, it will provide sufficient hinting and synonyms to minimize "hunt the verb" and telepathy issues.
@Stu There's certainly space for extremely hardcore puzzle games, don't get me wrong. But that's a choice that should be made deliberately, rather than by default. I can't imagine the Times Crossword has a growing audience. If people who enjoy the satisfaction that arises from adventure games, like myself, want to bring this to a larger audience, the Times Crossword version isn't going to do it.
When the game consists of "guess the designer's illogical intentions," there's no way to move from failure state to success without rote trial & error. Some people like that "ah ha" moment when they finally figure out a confounding puzzle.
But the truth is, most players do not. It makes them feel stupid. But the thing is, puzzles that are too easy also make players feel stupid. There are ways to create puzzles that do not frustrate, but old adventure games are not the text to learn that from.
My next post will try and address this more concretely (it's mostly written, but I have to remember nobody else has read it yet).
@JB How big is big though? I imagine they're still a tiny, tiny percentage compared to the match-3 du jour. Do you see them return to a significant presence via these channels?
@Gregory Certainly. I don't think adventure games are fundamentally flawed. The satisfaction they offer seems to be something many hunger for.
I think we need to acknowledge that the classic adventure games succeeded despite their gameplay, not because of it. Watching new players try Grim Fandango makes it pretty obvious these games are not accessible. Many folks creating adventures these day still emulate the structure of the old adventures, maybe toning it down a little.
I think we need to focus more on what and why of old adventures, rather than the how. Far too many seek to recreate the how, which is severely limiting.
Your Bars of Black & White is actually a very good example of a readable, substantive puzzle game.
That being said, I'm less convinced about IF. I think its readability issues are pretty fundamental. When the possible actions consist of every verb in the English language, I'm not sure solving that can be done without solving a lot of hard problems that have plagued computer science research for decades.
I could be wrong, but every single time I try to engage with IF, I end up incredibly frustrated. Lauded games that come very highly recommended, I still can't do it. And I'm pretty patient and tolerant. Given how many folks are far less willing, I'm not sure there's a way spread IF beyond its current niche audience.
"When the possible actions consist of every verb in the English language, I'm not sure solving that can be done without solving a lot of hard problems that have plagued computer science research for decades."
Or, recognising that fact, it could just be a case of building the game around a more focused group of verbs and being up front about them.
I really think that the strength of text input for IF games lies in the variety of nouns rather than verbs (and those nouns should be in the text anyway). But then, perhaps that means I'm not really talking about 'adventure games' in the way you mean it.
@Pacian You're absolutely right. I get the impression that some IF creators/fans find a fixed and display set of verbs to be a antithetical to the form, but I'm not sure there's an reasonable alternative for reaching more players.
Hi. I've written text IF, so I have some bias in this question, but I still want to address it...
You say "every single time I try to engage with IF, I end up incredibly frustrated". I believe you! I expect that the vast majority of computer-game players, when confronted with my text games, will fail to connect.
*But*, it's also the case that I can play these games. There are other fans of the games, and they can play them effectively -- even the badly-crafted ones. This tells me that there *is* some way to understand the games and improve one's ability to engage with them! It's just that we're doing a terrible job of *teaching* those skills.
(Many web sites and essays are floating around on "playing your first text adventure game". That immediately calls the question: why isn't *playing the game* itself the necessary teaching experience? As it is in most modern games -- and all casual games. That, I think, is the specific readability critique that you want to make.)
(My game _Dreamhold_ is an attempt to answer this: it's intended to be the "tutorial level" of text IF. It doesn't seem to have worked; at least, I don't get email from people saying "I played Dreamhold and then I was able to play Shade!")
As to graphical games... the thing is, I *agree* with you about the Lucas/Sierra lineage of inventory-scenery-verb games. I think they're poorly constructed, illogical, and offputting to all but their hardcore fans. (Yes, even to _Grim Fandago_, which had brilliant setting and storyline but frustrating gameplay.)
What was the archetypical easy-to-engage graphical game? _Myst_. Okay, classic example of the easy-to-intuit interface.
But... not so simple as that. _Myst_ provides you with hundreds of choices at any given moment! Your only verb is "click", but you can click on *anything you see*, in a richly detailed visual environment. If the gameplay were a shallow "click on the right thing now" series of "puzzles", _Myst_ would have been unplayable.
(Actually, there turn out to be several verbs, but I won't get into that here.)
The shorthand term for a bad graphical game is "hunt-the-pixel" -- just as the shorthand for a bad text game is "guess-the-verb". In both cases, a skilled player engaging with a well-crafted game *isn't doing that*; the gameplay comes from *figuring out* the important object (not pixel), from figuring out the important action (not verb word). The game has ways to convey that importance. Doing wrong things gives you informative failures. You are never in the situation of trying 1000 verb-noun combinations (or 1024x768 pixel clicks).
But, again obviously, _Myst_ has a much lower learning curve; the cues are more intuitive to pick up on. Text IF's cues are rocky, if not downright metallic.
I remain interested in solving that problem.
My IF game "Blue Lacuna" also has a tutorial at the beginning to try to get the player up to speed, and more clearly elucidate the available verbs. On the other side, the game prints important nouns in a different color to draw attention to them. The only downside of these experiments, I feel, is they reduce a bit of the pleasure of exploration, much like graphical games that give every important object an obnoxious glow. (Which means you always know where to click, but don't get the sense of discovery of finding the right spot to do so.)
One of the best things about Myst IV was that clicking on something that did nothing still produced a tapping sound, pulled from a wide variety of materials (tap on felt, tap on glass, etc.). It was the tiniest bit of feedback, but it was still *something*, and it made your "wasted click" still have a few grams of value. ("Oh, that's harder than it looks. Oh, what a pretty noise." Or even just "Ok, the game *did* actually register a click on that spot.")
Just yesterday I was on Amazon and noticed that Myst IV (now five years old) is still in the top 10 list of bestselling adventure games, above Myst V which came out a year later. It's possible that adding more feedback to the array of possible actions is a factor in this.
This has been an interesting read since the things you have talked about overlap very well with the subject of my masters thesis for EA DICE.
Something that I found very helpful dealing with these issues was the article "Interaction and Narrative" by Mateas and Stern. The authors talk about the concept of Agency as being crucial to player engagement in a game experience.
Agency is the player's sense of being able to act in, and therefor be part of, the game world. To achieve a high sense of agency there has to be a good overlap between the material and formal affordances of the game.
Material affordances are the things in the game world you can interact with and any effects that you as a player can achieve.
Formal affordances is what motivates the player and help him/her choose between and prioritize between all the available actions.
Classic adventure games tend to have a great unbalance between the formal and material affordances - in that the material (stuff you can/could do) outweighs the formal (stuff you should do). FPS games and modern action games often have a better balance - making it easier for players to disregard the right things and never having to second guess themselves.
@Andrew Thanks for taking the time to comment. Given my limited experience with IF (despite efforts), I very much would like to hear the thoughts of experienced IF designers (authors? developers? is there a common term?) such as yourself.
I agree that needing a supplemental essay/article about "how to play" means the game itself has failed to be intuitive. Again, there is certainly a place for expert games, but they aren't going to bring anyone new into the fold. And if the only games to emulate are intended for experts, creators may be exclusive without even intending it.
I don't mean to say that adventure games are inaccessible to everyone. Clearly there are folks that engage with it and enjoy it. But what I'm interested in is everyone that might want to engage with adventure games but are unable to with current and past offerings. Discovering what things are holding them back and how those can be resolved is a question I'd like to see more thought put into.
@Aaron Thank you for your thoughts as well. I think balancing that tradeoff you describe is hard problem indeed. Balancing the satisfaction that comes from exploration and problem solving with making things too obvious and rote is very difficult. Especially considering what's insultingly obvious to one player is satisfyingly discernible to another. Some thoughts on this are coming tomorrow, but they are a first foray only.
I read about Blue Lacuna and was intrigued, but was intimidated by the length. The conversation here is encouraging me to have another go with IF and I'll try my best with Blue Lacuna.
Did you think about issues of audience/accessibility when creating Blue Lacuna? Did you worry that the length may be intimidating to some or were you targeting it towards an audience who would find that level of breadth appealing? Or did it just evolve that way naturally?
@Erik The Runner's Vision in Mirrors Edge is a very good recent example of addressing readability gracefully. Do you know how that feature came about?
I'd like to read your thesis, if it's available online and in English (despite my heritage, my knowledge of Svenska minimal at best). Feel free to drop the link here or to above49.ca@gmail.com. I appreciate the pointer to Mateas and Stern's article as well. I hadn't read it before, but plan on correcting that now.
@Nels: Reaching a larger audience was definitely on my mind with Blue Lacuna. That mission was compromised to a certain extent by conflicting desires to make a "classic" adventure game (which I think I've now gotten out of my system). Regarding the length, it mostly evolved into its final size organically; though if anything, I think the length is a plus for my target audience-- or put another way, it's probably not the kind of thing that people with short attention spans would enjoy.
I think BL is more accessible than the average piece of IF, but not enough to break out of the problems you describe in your post; I'm interested in pushing my next (smaller) project much farther.
@Aaron Very interesting. To clarify, my intimidation wasn't length vis a vis attention span. Rather my concern was that I'd invest some portion of time, become stuck and unable to finish without GameFAQs (or something similar), which taints the entire experience for me.
I've added BL to my "to play" list and I'll do my best to give it a fair look.
Nels, as always you're right on the money. This is an interesting way of looking at it, though, and it'll take me a wee while to digest it.
Here's my 2 cents on adventure games:
As Warren Spector puts it "Games are at their best when they're saying something about the player, not the designer or the game itself".
It kinda ties into readability of games - that, if something looks possible, seems possible, it should really BE possible - but adventure games were/are always doomed to be a fringe entertainment form because they don't allow a player to express themselves. The closest they can achieve is having some kind of branching path, or allowing players to solve linear puzzles in their own order. Compare that with the kind of emergent gameplay seen in Splinter Cell Chaos Theory and you'll see that adventure games as we know them just can't stay afloat.
I apologise to the many IF fans commenting here, but you just have to accept that your genre of games is always going to be an outdated fringe area of development.
Alan Jack: I think you're moving the goalposts. Emergent gameplay is a cool concept, but we're just beginning to explore it. It didn't pop up in 1990 and destroy Infocom in fire and thunder.
Games with linear stories and simple branching structures are as popular as ever. The shelves are full of them. Nels's original point, and the question we've been discussing, are about the *playability* of games -- particularly for genre newcomers. Is the gameplay opaque? Is the first move obvious, and does it lead you naturally to the second move? Is the first move *too* obvious?
(A choose-your-own-adventure book presents every game choice as a short list of explicit choices. This maximizes readability at the expense of engagement and gameplay.)
Text IF has done *lots* of experimentation in nonlinear storytelling, emergent plot structures, emergent game mechanics. I'm not saying that to brag; it's a direct consequence of a form in which one person can write a small game in a month or two, with no budget. It's a huge area of interest within the community; the problem is how isolated the community is.
As for being a fringe area: writing SF novels is a fringe area of the entertainment industry. Low volume, low profit, small audience. But I don't hold it in contempt. I can even argue that it's a cutting edge of the industry; SF TV and movies (and games) tend to draw on ideas that hit the book world ten or twenty years ago. So where do you want to be?
Erik: I tend to see game design occurring on *lots* of levels, not the just material and formal you describe. (There's UI on the bottom, for a start. The "formal" layer may actually be divided into short-term goals and overarching plot shape. Etc.)
It's not really a question of balance; the layers are never of equal prominence. We take for granted (perhaps falsely, but as a matter of practicality) that some layers will be shallow and transparent, while others are deep and engaging. I figure that if *one* layer of interaction grabs the player, then the game will be successful. The challenge with game design (as with books!) is to create something that works on multiple levels.
@Alan See my post today where I discuss these things, similar thoughts there.
Andrew is correct though, there is certainly an audience for text/graphical adventures as they are. But this audience is certainly smaller than it once was and isn't growing much.
While looking at how to bridge the gap between other games and "expert" adventures is interesting, what I'm really curious about is how structural changes can make games with these components engaging.
No type of game exists in a vacuum either, or at least it shouldn't. Good ideas from action-oriented shooters could be appropriated for non-violent adventures and vice versa. Saying one genre or style is objectively better than another isn't particularly valuable, IMHO.
I still go have concerns about the "type anything you want without guidance" interaction of IF, but I think there is important things to learn from IF (even if it's not to everyone's tastes).
@Andrew I meant to make this point explicit, but the low production costs (both in terms of time and money) for IF is one of the form's greatest strengths. Iteration is essential for the evolution of any design and a single creator being able to do so with such celerity is quite a boon indeed.
@Nels I totally agree that game design is more complex than just matching a set of affordances. But in working with Mirror's Edge and analyzing tones of other contemporary FPS:s I have found that a huge source of player frustration/confusion/brakes stems from the fact that the beautifully rendered game worlds of today falsely communicate more interactions than they actually afford. Among the classics are corridors of doors that can not be opened, very fragile looking stuff that is impervious to weapon damage etc. In most fast paced action heavy games the player is however quite unlikely to run into too much of this (partly due to a balanced set of affordances - CoD is a good exampla of such a game). In games of a more explorative nature, with a more open structure, these problems become more salient - now there will emerge conflicts between what the player feel they should be able to do and what they actually can do given the rule set of the game.
I'm hoping to have my thesis done by mid June and I'll mail you a copy then. =)
@Erik Fantastic, it sounds like we've very much on the same page. Good luck finish up the thesis and I look forward to reading it when it's complete.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home