Friday, March 9, 2012

Maddox - Storage Run Footage

The following is a full run-through of Maddox: The Retail Rebellion. Requires some cleaning up between now and Monday, but otherwise functional.

Monday, February 6, 2012

Project Overview - Maddox: The Retail Rebellion

The following is an account of my visual component's current progress. It is titled "Maddox: The Retail Rebellion" and follows the story of one Dr. Robert Maddox, a mad scientist whom the bank forecloses on, forcing him to seek legitimate employment. He seeks such employment at a retail giant called U-Mart, where he works a series of blue collar jobs, unsuspecting of a diabolical plot that lays beneath its friendly vernier. In this post I will give an overview of the game thus far, the approach used to conceptualize it, how that approach differentiated from my earlier approach with The Brothers Riley, and the decision tree architecture driving it.

Game Overview

The core mechanic driving this project is the Madness Meter; a "health bar," or more accurately, a "pressure gauge" of sorts in the upper left corner measuring Dr. Maddox's control over his wits.

His situation in this game calls for him to restrain his manic impulses as a mad scientist and try to be socially acceptable as he works a blue collar job at U-Mart, but his outrageous personality does not repress itself easily. By performing acts that indulge his idiosyncracies, Maddox can "vent" the Madness Meter, but repressing himself will cause it to build up, potentially to dangerous levels. At 100%, Maddox undergoes a psychotic break and gets shipped off to the mental hospital, earning a game over.

However, Maddox is held accountable for his actions by those around him, and his natural personality is often condescending, abrasive, and generally off-putting and disruptive. As such, he's on a "three strikes and you're out" system with each NPC, all of whom have different reactions to when he finally crosses the line--either sabotaging him at work, foiling his various "schemes" (more on this in a minute), or outright reporting him to the manager. U-Mart has its own "three strikes" system as well, which can lead him to being put in worse and worse departments as he tries to maintain his sanity and earn his pay, or even worse than that--being fired. Furthermore, the worse the department he works in, the worse it can affect his Madness Meter; the more menial and disinteresting the task, the more Maddox will rattle his cage.

Pictured: Schemometer

Maddox does have a few tools that can help him out, though, chief among them the "Scheme" mechanic. Schemes are optional, adventure game-style puzzles that the player is able to solve in each level, taking on the theme of crazy inventions and plots for revenge, and the presence of a Scheme is indicated by the lightbulb icon in the lower-left corner of the screen. For a hint as to how to perform the Scheme, Maddox can consult his pet rat Willard for advice, as illustrated in the video above. When Maddox performs a Scheme, he vents a great deal of Madness--anywhere from 25-50%--and the icon shatters.

Schemes come in two flavors--"Spite" Schemes that are made to take out Maddox's frustrations on those around him, and "Work" Schemes that take on the form of inventions to aid him in making his daily drudgery less problematic. Different U-mart workers have different criteria for scoring a "strike" against them (some include working more efficiently and making them look bad), so either type of Scheme in each different department carries a risk to go with its reward.

Conceptual Approach
The Retail Rebellion brings in an extra component in addition to the decision tree architecture developed in prior projects, that being the overarching systemic principle of the Madness Meter, namely, its application of conflicting rules. Per the "Rule Following" approach to psychology, The Retail Rebellion adopts the idea that ambiguity is formed from conflicting identities, in this case the player character's identity as a mad scientist as well as his identity as a blue collar worker at U-mart. Each identity has a different set of rules, and both are evoked through clear systemic elements in the overall design at all times throughout the game, thrown in direct conflict with one another at all times.

Thus, instead of taking the approach that the personalization elements of choice-driven narrative are purely for the sake of role-playing and interpretation of the narrative, the Retail Rebellion integrates them into a balancing act for the player to follow. While it is entirely up to players where and when and against whom they must indulge Maddox's personality and what rationale they use for doing so, they must indulge him and cannot simply ignore his characterization for an "optimal" run through where everybody likes them.

To put it simply, the themes of this story, that of personal versus professional identity, are integrated not just through the types of decisions built into the decision tree, but into the overall systems and game mechanics driving it in a tangible risk-and-reward relationship. The core experience of this story isn't merely trying to get through day-to-day life in a retail store, but that of a mad scientist stuck in a retail store, and the game puts that upfront and uses it to create a unique challenge in how the player navigates the decision tree.

This differs greatly from The Brothers Riley, which simply sought to create emotional connection through its narrative--that of two brothers trying to save their sick mom--and allow the player to explore that narrative in any way they chose. This meant the game was devoid of a unifying core experience outside of the very plain notion of "two brothers in London." While The Brothers Riley was meant to lead players into the heist of the Koh-I-Noor diamond as the core experience and storyline, the build-up chapter that I'd built allowed the player to choose goals outside of that line of characterization; IE, players could avoid a life of crime entirely and take the side of the law, not even getting involved in the heist. Effectively, players could opt out of the core experience of the game, and thus there was no core experience to drive mechanics, develop a story, or develop dilemmas--hence my reservations about the game not having any stand-out moments or exceptional features. In the absence of risk and reward players had full control, and could easily avoid putting themselves in a dilemma or facing an ambiguous decision.

I would argue that the failings of The Brothers Riley are mainly due to poor craftsmanship on my part--but, at the same time, I feel like this scenario makes a strong case against the notion that "agency" is the goal of choice-driven narrative, as too much agency can leave the player devoid of interesting decisions to navigate. In a world of perfect player agency, the player can optimize rather than merely satisfice--and this is not consistent with the human decision-making process, and thus, not consistent with players' expectations of choice.

Decision Tree Architecture

Pictured above is the master decision tree for Maddox: The Retail Rebellion (Part 1), detailing the introduction and the flow from one job to the next in terms of the decision archetypes I've detailed--personal, structural, long-term, short-term, direct, and indirect. Here I've not only detailed the flow of the game's events by these terms, but also associated specific game mechanics with specific types of choices, creating a greater sense of consistency in how the player interacts with and sees the world.

The flow is as follows:

First Node - Introduction of Maddox, as displayed in the video above. Choices here are personal choices for personal choices' sake, both for learning the character and for learning the dialogue system. This takes us through from the scene where Maddox is menacing the UN to his meeting up with his friend, Jacob Haynes, in the apartment. As of now we are in a consequence-free zone.

Second Node - Job Interview. A string of personal decisions (how Maddox answers questions) leads to the first structural choice of the game, landing him in one of many different departments at U-Mart depending on the impression he makes. Three different departments shall be included in the final game: storeroom, complaints, and electronics.

Third Node - Introduction of Job. Each department at U-Mart is represented with a different mini-game and has a different corresponding NPC to give Maddox a tutorial. In the introduction, Maddox gets a chance to make a first impression on this NPC before being thrust headlong into the job.

Fifth Node - Working the Job. During this segment the player partakes in the minigame, a completely rational segment wherein Maddox must balance gaining additional points in the minigame (represented in this chart with "bucks") with his Madness Meter. Each job involves a "cycle" of interactions designed with the explicit intention of being simple to do, and each cycle pays out some number of points. For instance, a "cycle" in the storeroom minigame might involve moving to an item, grabbing the item, and taking it to someone. However, depending on what job Maddox is performing, individual actions in the cycle have the potential to increase his Madness Meter. The more menial the task, the worse it becomes, thus forcing him to find ways to vent--or to slack off.

Sixth Node - Performance Evaluation. After a set number of rounds at the minigame (currently two--adding a "break room" segment in the middle for reasons that are forthcoming), Maddox undergoes a performance evaluation with the interviewer from before. He takes into account any strikes that have been added against him as well as the number of points he's managed to earn thus far, then determines whether or not Maddox should be put into a different department, employing a mixture of rational and personal criteria to route the player towards the rational goal of moving into a better department--or keeping a good one, as the case may be.

Other Elements - Schemes. The scheme mechanic is listed here as a long-term/rational goal due to its nature as a high-risk, high-reward action as well as a potentially significant piece of additional content.

Other Elements - NPC Reactions. As stated previously, NPCs will react to Maddox in one of several ways, all of which are listed here. These are displayed as structural, indirect decisions, as they occur due to reactions to the player's input and can bear significant changes on either the player's relationship with that NPC, their performance in their job, or their overall U-Mart score.

The overall philosophy which this is built around is that it is not desirable to have multiple, different content streams but rather that the player's decisions should change their perspective, point of view, or relationship with one, singular content stream. As such, alternate content streams in this game takes on shallow cul-de-sacs before then altering the main content stream, changing the way the player must perform their duties at U-Mart and their duties to Maddox's manic mindset.

As this game develops further, this flow chart will be updated with more information.

Monday, October 31, 2011

A... Mid-Mortem? on the Brothers Riley

After having been stuck for a couple of months, I've picked up progress again. This first update concerns my addressing the shortcomings of the prototype as of its previous showing; it's written as a post-mortem, but as I'm continuing to develop the project I guess it's more like a mid-mortem or a self-criticism.

"The Brothers Riley" is a point-and-click adventure game following Alex Riley and his younger brother Jack, two boys in a downtrodden family in 1850 London. With their father recently dead and their mother delivering a baby imminently, the two need to find someone to help deliver their new sibling. They have three options set out for them: gather enough money up to pay Dr. Gossar, their family physician, to perform the delivery; depend on the charity of the Church; or employ the aid of Richard Darrell, a criminal with sinister ulterior motives. The player takes on the role of Alex as he struggles to fill his father's shoes, and his actions directly change the attitude and responses of his younger, more impressionable brother.

The story of Alex and Jack Riley's struggle to find someone to deliver their mother's new child wasn't interesting or remarkable in any way. Alex himself wasn't a very interesting protagonist, and his younger brother's contribution to the story wasn't really worth the effort of trying to get him to respond dynamically to the player's actions, nor did anybody I showed the prototype to seem to feel that it had that much impact on their perception of either their own character or the story at hand.

The decision-making scheme itself, I felt, was inherently flawed, with no relationships being developed between the player and tangible resources. There was money, yes--but it had one use, and that was advancing the plot. There was multiple ways of acquiring it, but not multiple alternatives to spending it--so the element of risk, so heavily emphasized in my research, wasn't really present. There was a sense of personalization, but it was very weak as there was no risk or ambiguity of player identity; the game puts the player firmly and foremost as a boy trying to care for his mother, and interactions as a big brother, as a boyfriend, and as a worker are comparatively minimal as none of THOSE are used to advance the story or made out to be goals.

The only thing I had to fall back on was emotional investment in the characters, which I don't feel I obtained. Because of the limited scope of this project, I overly constrained myself and rushed to meet quotas for introducing characters and different types of decisions simultaneously, with little in the way of build-up or context. Richard Darrell, for instance, is barely established, coming out of nowhere and seeming to just assume himself as an important figure without the player learning anything about him. Other characters for this game were stilted and heavily founded in stereotypes, such that I was barely able to remember anybody's name apart from the two lead characters. For me that's a big problem: I'm known among friends and colleagues for being able to create characters with memorable names with remarkable consistency, so if I'm calling someone "the girlfriend," or "the priest," or "the bully," instead of by their given name, I am definitely doing something wrong.

A big part of what threw me off was relative inexperience developing adventure games. There's a certain way these games have of introducing characters and surprising players, I realized. You don't play an adventure game to deal with people your character is familiar with--in this case Alex's entire neighborhood. You play an adventure game for that odd experience where you're at a subway station and have to ask a stranger for directions, but he turns around and -- oops, he's a platypus! You then embark on what will undoubtedly be a very memorable anecdote. Expositing on existing relationships is difficult in this particular genre since the player and their character do not share a memory of past experiences, but starting a new one is very easy. The mode of interaction lends itself very naturally to the act of meeting and inquiring as opposed to the act of having a reunion or a discussion.

My first pass at this project wasn't much at puzzle or interaction design, either, and in hindsight it seems like each character could be made more memorable by adopting some kind of a puzzle around them. 

Phillip Stone, for instance--the owner of the factory where Alex works--could be more of an interesting character to interact with if he's constantly and stupendously drunk. This could then become a puzzle, a character trait, and a choice all in one: sober him up to get information out of him, try to interpret his rambling as hints for some other objective, or try to make him pass out so that you can safely steal a possession from his person. The player then has tangible experience with one of his core character traits, and getting key objects from his home becomes more interesting than walking in and just picking them up. Interacting with Mr. Stone must be necessary, as should be the case with most of the other characters. As it stands, most of them can be ignored and offer no challenges like this.

A greater distinction must thus be made between "goal characters," "puzzle characters," and "tool characters," and some traits must be identified for them beyond their basic identity. Characters can not exist merely to offer opportunities for interaction--each one must serve a purpose in either offering direction, offering a resource, or offering a challenge--or else, a hint, or an opportunity for a future challenge/goal that the player does not yet recognize. If a character does not serve a purpose with respect to any of these design goals, at least in the context of the point-and-click adventure game genre, it is almost entirely wasted. This is a fundamental difference from traditional fiction writing, which was the perspective from which I was originally writing this.

There were a few things that have worked about this project so far--they were just on a very small scale compared to the fundamental issues holding it back.

I did find an effective way of introducing personal choices. While characters in themselves can't exist JUST to offer choices for no reason, characters that the player interacts with can be helpful in presenting theme-based, personalization-oriented questions, both to help introduce their own core character traits and to help the player's character develop and the player orient their point of view to those character traits. A good example of this is Margaret Darrell, who tells the player to put their trust in God; Alex is then offered the choice of putting their trust in her words, or refuting her and blaming God for the problems his family is facing. As religion is a defining trait of the time period, it is a major character trait and offers one of the most important devices for developing the player's character. That said, this is a pattern that could be taken further, with different, important aspects of the story--the player character's attitude towards crime, or authority, or his family's reputation, for instance--being meted out among the main cast as they are met. This method is rendered ineffective when it is unnecessary to interact with most of the cast--and thus the player wouldn't experience most of these events--but nevertheless proved an interesting way to build context if experienced.

I am working on a revision to the project, trying to develop a stronger, more localized structure for it. IE, I'm going to axe the city and go with someplace smaller. Maybe a portion of the city--a single street instead of two--and Phillip Stone's mansion, as Mr. Stone presents the strongest opportunity both for mischief and revision (I just presented a means of revising him).

I will also be trying to mete out the risk factors among the NPCs as opposed to depending on one partner NPC as the sole risk factor -- IE, getting Mr. Stone drunk may carry certain consequences, using mischief as a means to get what you want may anger people that could help the player later, et cetera. It's very likely that Jack Riley will become the player character in lieu of Alex, as Jack has stronger conflicts and more growth to do and, as a mischievous young urchin, is probably the most fun to follow ("Horse's arse." "What was that?" "I said Holiday Pass.") and should be allowed off the leash of his big brother.

A revised document on this revision will follow shortly.

Wednesday, June 29, 2011

Case Study to-do list

I've been evaluating a lot of games by my Personalization/Discernability/Gameplay Integrity/Narrative Integrity rubric, but you'll probably notice that this dropped off sometime not too long after my initial push with my thesis. This is for a couple of reasons.

First and most obviously, games with narrative choice systems are incredibly time-consuming and I've been a very busy grad student, tending to spread myself thin and overextend my resources volunteering for other projects. I've been focusing a lot on filling out my technical background, which means I've been in workaholic mode for the last quarter.

Second, I've had to refine my thesis and its goals a lot. There's been a good few times I've re-considered narrative choice in general due to my interests as a game designer being on a little bit of a different path, scope issues with the material of the thesis and my presentation to the thesis board (which I expected to fail up until the very last week, when I figured out exactly how to wrap it all up), and, in particular, due to scope issues trying to discern a good visual component. That said, I didn't want to jump the gun and spend hours evaluating these games if my thesis should change dramatically.

Third, as I've stated previously there is and has been a need to clean up my rubric a little bit and make the qualifications for good/bad discernability and personalization a little bit more concrete and streamline the process. Ideally I'd like to judge it on five or more sub-points so as to create a clear -5 to 5 or a -10 to 10 scale. From my previous case studies I should be able to do so easily.

That said, I've gathered together a to-do list of games that I still need to perform full evaluations of.
  • L.A. Noire
  • Dragon Age: Origins (referenced in thesis, playthrough not complete)
  • Alpha Protocol
  • The Witcher 2
  • Fallout: New Vegas
  • Overlord (Playing it anyway, might as well analyze it)
  • StarCraft 2 (Played, not evaluated)
  • Fable 3 (Playthrough not complete)
  • Deus Ex
  • Iji
Too many case studies, too little time to get them all done. I may need to narrow down my choices.

Tuesday, May 31, 2011

On scope and engine research

For the last quarter I've had a bit of a crisis with regards to this thesis and how to go about executing the visual component. The research I've had well in hand since the beginning, and the overall process for developing nonlinear story content, I feel, I've made into something pretty straightforward. The really big issue up to this point has been more a matter of what's in scope for me to reasonably make. For the last few months up to this point, I've been trying to conceptualize a one-act play based in part on interactive cinematic games like Heavy Rain, with full emphasis on the narrative system in lieu of mechanical systems in order to more fully develop the potential that I saw in that sort of project. For a while it looked like I was going to base it on the "Kobiyashi Maru" from Star Trek, which I felt would keep the project self-contained and provide an interesting theme to explore.

Such a project is more appropriate to a cinematic designer or game animator than it is to me, my skillsets being more in game systems, technical design, content design, and writing. The vast majority of the work on my project would come down to heavy animation and cinematic directing, which I have positively no experience doing--much less for a nonlinear story. I'd have to recruit someone to do that for me while I plan the content, and I can't abide having a thesis project where the majority of actual production work is being done by someone else rather than myself. As such I had to re-think some things in order to make this project feel reasonable and underwent a few discussions about it with Professor Cookson. To sum up his advice on the matter: I'm making too big a deal out of making the presentation of it super-professional and cinematic and should just be focused on the content. He suggested it wasn't entirely necessary to make an Unreal game and that I could just as easily put this together in Flash and it would be perfectly acceptable.

In the meantime I've been doing a lot of work with Unrealscript this quarter, developing a scripting library and tutorials that will enable students to more easily develop original content and games. It's been highly successful and as such I've become confident in my ability to generate content and mechanics with it. Professor Cookson and I agreed that the work I've already done in this regard is definitely a feasible foundation for the project. As such, I'll be making this a 2.5D side-scrolling game based on my side-scrolling platformer scripts. I'll have more information on this project very soon as I've got a clear picture of what I'm trying to make. As usual scope control will be an issue, but I at least feel confident in my ability to generate gameplay-oriented content as opposed to cinematic-oriented content.

Revised Production Schedule

·      March-June: Development environment research
·      June 10: Design documentation, thesis paper outline
·      June 5 – June 27: Team recruitment period
·      June 11-June 27: Documentation revisions
·      July 11: Thesis case studies
·      July 25: Alpha build of visual component
·      July 28: Thesis body first draft: research, development, production process outline
·      August 15: Beta build of visual component
·      August 15-August 29: Beta testing of game and narrative systems
·      August 18: Thesis body second draft: evaluation of production process thus far
·      September 1: Beta testing results and revisions based on feedback
·      September 12-Oct 3: Crunch
·      Oct. 3: Full content freeze, final bug testing
·      Oct. 24: Bug testing and fixes completed; user tests for effectiveness of narrative begins
·      Nov. 14: Final draft of thesis paper, first pass – factoring user feedback
·      Nov. 25: Final draft of thesis paper, last pass: full development retrospective