In part two of this serial, we explored three examples of the aesthetic flaws of videogames, and showed how each could be understood as a special case of cognitive dissonance in games, and also as a clash between player practices. This week considers how we can design games to avoid such dissonance.

When we think as game designers – in terms of rules, game mechanics, or systems – game dissonance is solely a negative aspect of the development process, something to be eliminated wherever possible. This may seem depressing; after all, the most effective way to avoid causing dissonance is to never vary from what players have previously experienced. This is, in fact, the typical situation in AAA games, which seldom innovate and prefer instead to take proven player practices and iterate upon them very slightly, counting upon volume of content and degree of polish to wow their players. Commercially, it's the winning move, creatively, however, it's a fairly underwhelming strategy, and one reason that AAA studios don't put everyone else out of business is that more agile, more inventive, smaller developers can attract players via their creativity.

However, when we think as narrative designers, the ways game dissonance can manifest are so varied, and the circumstances so disconnected from anything that we could avoid by simply playing it safe, that there is more of an opportunity for creativity. Again, AAA games don't like to take risks - but in this case, it is to their cost, for the default way of telling stories in these titles is to intersperse an animated movie with the gameplay, and this is not a way to avoid game dissonance but rather a certain way of causing it in at least some proportion of players - indeed, this is what Hocking's concept of 'ludonarrative dissonance' expressly drew attention to.

How do you create game stories that avoid game dissonance? The key is that all of the components of the videogame must align, and the player's expectations must never be disrupted by the situations presented. Thus a player who is expecting that their character's story is going to be developed in cut scenes beyond their control is less concerned by this limitation. An ironic consequence of this is that the more that the developer throws players a bone in terms of creating an illusion of agency (for example, by offering a choice), the more constrained certain players will feel when the developer once again takes full control of the narrative and gives the player no role except that as observer.

Key to avoiding game dissonance in narrative design, therefore, is managing expectations. For instance, if you let the player choose conversation topics, you need to ensure that choosing topics remains an option throughout. If your story requires the player to have a specific conversation at a certain time, you must at the very least ensure that it is the player's choice of topic that kicks it off. This is much easier to do in a game that is, at its heart, investigative, than in a straightforward quest adventure, although even this can be made to work with care. If the player never chooses topics, on the other hand, there is no problem that the dialogue triggered is set in advance: the player will quickly learn to expect this.

I suggested back in the first part of this serial that a common cause of game dissonance (and one integral in Hocking's complaints) is when the story techniques of the screenwriter are injected into videogames. Part of this problem is the screenwriter's fallacy: the mistaken idea that what makes videogames a powerful narrative medium is that the player can make choices in the story. But from a screenwriter's perspective, those choices are only pre-prescribed options in an already planned-out plot. The screenwriter can never give the player agency using their methods, although they can (and do) bring NPCs to life by using a toolkit that is much better suited to characters that are not under the player's control.

Rather than simply inserting choices into a screenplay-style plot, narrative designers can offer players far greater agency by respecting the player's freedom within the game world. Where they go – and when! – is often more than enough to create opportunities for storytelling, and although developers are cautious about timed elements in game stories (with good reason!) even this can be used when it is done with care. For instance, if the player is told by an NPC that someone is being stalked by the secret police on the other side of town, we can easily tell (as developers) whether they went straight there to rescue them or nipped off to the shop first to buy some more ammunition (or engaged in some other activity that was clearly not appropriate). Having two variant encounters (one if the player goes directly and one if they were dilly-dallying), you can surprise players with a greater sense of presence in the game world than they expect! You can even cover the two cases cheaply with two variant lines ("What took you so long?!"). If you choreograph encounters in a way that the player realises they are being hoodwinked, that will cause game dissonance, but if you honour their agency by reflecting their choices, they will be drawn deeper into the experience.

However, this preceding example is still, at heart, a choice that we have planned and anticipated. To get to a more fulfilling place in terms of player agency we need to have game systems that can, in themselves, allow players to express their agency. A screenwriter approaching a scene of peril might foreshadow a plot device that saves the protagonist from danger. When this same trick put into a game, the player is told to fetch the plot device, and then that saves them - which works, but is flat and dull as a player experience, being effectively a fetch quest. A narrative designer would do far better to design an encounter that the player can prepare for within the game systems. Knowing that a Fire Elemental is attacking the wizard's tower gives the player a chance to equip items granting fire resistance; giving the player spells or magic items that detect fire in the tower, in order to tip them off, heightens the sense of agency even further.

Remember that dissonance will manifest whenever any part of the game fails to align with the player's experience and expectations; this means the game systems themselves need to avoid clashing with each other, and also with the story materials. For instance, if you have a fantasy game in which an ancient sword of great power is a key plot device in the story, players will experience dissonance (or at least grumpiness) when acquiring the sword does not give them a new weapon! Avoid this, where necessary, either by making the plot device something the player can carry but not use (e.g. an orb only a sorcerer can use, but the player character is a warrior), or by adding a limitation to the weapon such that despite its power, the player can use it only sparingly (for instance, because it drains their life force while they are wielding it).

Elegant narrative design, therefore, is not about writing a screenplay-style plot and trying to shuffle it in between playable sequences, but about having a set of game systems that you can use to advance and mediate the player's narrative experience. This is why International Hobo has, for some twenty years now, urged developers to perform narrative design in pre-development, when everything is up for grabs, rather than hiring a writer later in the process to merely write dialogue. You are asking for game dissonance among your players when your story is the last thing you try and put in place. It ought to be the among the first things you put into place - and it ought to be constructed from the elements of the game design or not at all, if your goal is to minimise dissonance among your players.

Clever narrative design can also help in small but subtle ways to head off game dissonance between conflicting game systems. Games that let the player recruit an NPC retinue in a camp or base constantly face the risk of rupture when those characters can do nothing to help the player in the challenges they are encountering. That conceptual gap is not Hocking's 'ludonarrative dissonance', it is rather (quite ridiculously) 'ludoludic dissonance', which is why 'game dissonance' may be a better term all around. When two game systems work well conceptually in the game design but risk contradictions in the logic of the game world, always try to ensure there are reasons for things to be this way. When your military base is preparing to defend against an attack, its soldiers have a reason not to come with you on a mission. If the base is a mobile military hospital, you have even stronger reasons why you are not building a private army. These kind of decisions are simply not part of a screenwriter's toolkit (at least until they become an authentic game writer, which is always a possibility!), but they are central to the narrative designer's craft.

What these examples have hopefully made clear is that when we are conducting narrative design, avoiding game dissonance is also creating an elegant narrative design, which is to say, creating stories from the raw materials of the game systems themselves, giving player a sense of agency within those systems, and creating the fictional glue that holds together the logic of the game world. It is agency, not mere choice, that games bring to the artistic table, and whenever choice is substituted for agency we are not going as far as we could. Those planned choices can be valuable to a game story, indeed, they can be essential to avoiding dissonance and giving the player a sense of an active role in the game world. But they cannot substitute for honouring the player's freedom within that world. The screenwriter's fallacy flows from taking too great a control over the player's story, such that we know all the ways it can possibly turn out. There's a place for that, but it's not the best we can do, and it comes with high costs - it destroys the inevitability that makes conventional drama (especially tragedy) so emotionally intense, and it undermines the player's role in the process, thus betraying both the screenwriter's craft and the game designer's.

Finally, it is worth considering the question that Hocking brought to the table in analysing Bioshock: how do we make themes work in videogames? Here, the screenwriter's toolkit has more to offer, but is still slightly inadequate to the task. The way we make themes work in literature, theatre, and film is to have a topic manifest within and through the structure of the story, or via symbolic plot devices - the solid-gold Maltese Falcon is the perfect symbol for the greed its story revolves around. To evoke theme elegantly, the participant in the story ought not to see it coming a mile away, or if they do, they should feel the weight of tragedy (Whitehead's "remorseless working of things") precisely in its inevitability. Yet players, unlike readers and viewers, are resistant to inevitability; it is a cause of game dissonance that they can do nothing, unless enormous care is taken to create the inevitability (something few if any games have managed).

Therefore, rather than relying solely on conventional narrative techniques for expressing themes we ought to try and bridge between the game systems and the story systems, making both link up via theme. If your game involves murderising an ethnic minority like goblins to gain XP, you have an opportunity for themes that reflect upon this violence, or that paint the player's greed for personal advancement as the cause of suffering in the game world. If your game entails managing a farm, the logistics of that process give you opportunities to explore themes directly connected to what the player is doing - and perhaps reflect real tragedies in fictional form. Indian farmers committed suicide after scurrilous multinational corporations sold them 'super seeds' that functioned as a form of extortion since the crop yielded no further seeds, requiring it to be bought again every year at crippling expense. There are themes here ideal for inclusion in a game, although probably not in a commercial product.

The danger in relying on the existing narrative toolkits when creating videogames is both that the player will fall prey of game dissonance, and also that we will fail to explore the more unique possibilities in the media of games. Hocking was right - we still don't have exemplars for our medium as a storytelling form, even though we have had them for a long time as 'merely' a game. Tetris is one such exemplar. I hope it's clear why we might think we could do better if our goal is to push our creativity further, and take games into novel and unexpected places. But standing between where we are and where both Hocking and I hope we might be able to reach is game dissonance. Understanding what causes it, and appreciating how to avoid it, is thus the epicentre of the struggle to take games further as an artform, a battle which places narrative designers squarely into the front line.

Interested in game writing and narrative design? The second edition of Game Writing: Narrative Skills for Videogames is coming soon!

In part one of this serial, the concept of cognitive dissonance was introduced, along with Clint Hocking's ludonarrative dissonance, concluding that it was only a special case of a more general phenomena: game dissonance. This week considers how game dissonance functions.

When we talk about videogames, we tend to focus upon the aspects of the experience that we are enjoying. As a result, when our enjoyment is disrupted, we usually conceptualise this as a flaw in the game, although there are times when we will admit that a game is simply "not for us" and step aside. This creates a selection bias whereby we are misled not only about "what games are" (we will tend to think that what they 'are' is what we happen to like), but also about how our experiences in games are functioning in terms of our cognitive faculties. As a result, enormous volumes of text has been written about the qualities of great games and how to make good games, yet almost nothing has been written about the aesthetic flaws of videogames.

In February 2015, I published an article that addressed this deficit. It came at an interesting point in my life as a blogger, as encapsulated in the three comments it received. One was from Spry Fox's Dan Cook, who up until that point had been a regular correspondent with me on game topics - and indeed discussions with him over the wrapping paper fallacy had led directly to my writing the piece in question. Another was from game design legend Raph Koster, another of my long-standing game blog correspondents, and someone who I had perpetually only just missed encountering in person during my many years at GDC. The third was from a new face, Chris Billows, who would go on to be the most important of my blog correspondents over the next five years, both encouraging me to keep writing, and sharpening through opposition my understanding of a great many concepts that would become key to my thinking.

I eventually buffed up The Aesthetic Flaws of Games into a paper for the Digital Games Research Association (DiGRA), entitled No-one Plays Alone, and presented it in Dundee in 2016. It went on to be republished (in a slightly altered form) in the organisation's transactions, about which I was rather pleased - although as usual for my dabbling in academic circles, it had zero impact upon the research community, which remained steadfastly committed to their prior discourses (as inevitably they must!). This paper was the first time I had cohesively elaborated my understanding of player practices, which is to say, the habits we pick up when we are playing games that go on to assert constraints upon us when we make games. To understand games in terms of player practices is to appreciate that anyone who thinks their game concept was entirely their own invention is fooling themselves; we don't invent new games from whole cloth, we iterate on the player practices we already learned and crossbreed them with new influences, often from other kinds of media.

What do I mean by a player practice? A clear example can be found via recurring control schemes. Dungeon Master (1986) provided the prototype for the FPS by being a novel new form of computer RPG played in real time, and allowing the player to move square-by-square through a dungeon using cursor keys. The same control scheme was inherited by early id shooters like Catacomb 3-D (1991) and Wolfenstein 3D (1992), but as id moved away from Dungeon Master's square tile grid into free movement with Quake (1996), they added 'mouselook' as an option. This was to set the agenda for every FPS afterwards, and by Valve's Half-Life (1998) cursor keys were no longer the standard, but rather ASWD and 'mouselook'. Nowadays, any PC player who plays an FPS inherits this player practice, even though they have no conception that it descends from an older practice. A similar (yet more extensive) example is the conservation of game inventory elements.

Because player practices are the habits of players, they are more than capable of triggering cognitive dissonance when they come into conflict with other mental states. In the 2015 article, for instance, I gave the example of Metroid Prime (2002). This completely tore up the rule book when it comes to the player practices of FPS games, which for consoles had been set with GoldenEye 007 (1997) in response to Quake. The console adaptation of 'mouselook' was twin stick controls - but this practice had never taken off in Japan, quite possibly because of the smaller community of PC players in that country. Nintendo handed development of Metroid Prime to a US developer, we can deduce, to attempt to come up with a new version of the FPS controls, one that Japanese players could operate. While they broadly succeeded on this account, they also created a serious issue with many US players.

I have called this problem perplexity, meaning not mere confusion (which is a positive part of the play experience of, for instance, adventure games) but rather cognitive dissonance caused by the player having learned one player practice and then encountering a game that it is impossible to play by that existing practice. Thus, some players of Metroid Prime (especially in the US) complained in vociferous terms about how dreadful the control scheme was. But the scheme itself is fine, taken in isolation; indeed, it works rather well on its own terms, providing all the expressiveness of a twin stick control but never requiring simultaneous operation of both analogue sticks.

The problem was not that the control scheme was flawed, as such, it is that it was different to what players had already acquired as a practice, and this disparity caused game dissonance. It should be clear that in this case there was literally no aspect of that dissonance that can be attributed to a disconnect between story and game. Rather, two different player practices - the one already learned, and the one Retro Studios invented - could not be reconciled in a proportion of the players of the game. This is what I'd consider to be the textbook example of the how clashes between player practices cause game dissonance, and in cases of perplexity that dissonance is resolved either by 'converting' to the new practice in its context (that is, loving Metroid Prime) or by 'rejecting' it (and therefore heaping scorn upon the control scheme).

The second of the three aesthetic flaws of games I previously identified is inelegance, which occurs when game systems become so bloated or disconnected that players become aware of something troublesome in the design. Here, the problem is not so much a disconnect between the previously learned player practices and those of the game being played, so much as it is a disconnect between one set of player practices the game is teaching and another. The mechanical systems of the game simply do not fit together, and the player becomes acutely aware of it. Once again, this is an example of game dissonance; two (or more) cognitive models are failing to align with one another. It's unusual, though, as those models are of the same game (at least in the sense we usually talk of a single 'game').

It is interesting to look at examples of older games that switch their control schemes around. In the 1980s, this was very common: Horace Goes Skiing (1982), Aztec Challenge (1983), Beach Head (1983), and Raid Over Moscow (1984) are all examples of games consisting at heart of at least two and as many as half a dozen entirely different sequential game sections, each requiring players to learn different practices. Players in the 1980s accepted this willingly, in part because it was the usual practice in the 1980s to play lots of different games in the space of an hour. Whether in the arcade, or working through an audio tape of pirated games, players routinely played videogames as a buffet at this point in time.

Today, however, we view this kind of wanton collision of disparate play activities as decidedly inelegant, and typically criticise them unless they have some clever conceit to bind it all together, such as the Warioware series' (2003 onwards) microgame concept, where a single command ('Enter!', 'Rub!' or 'Don't Move!') introduces a fragmentary game snapshot. Our player practices have shifted: we now rarely play multiple games within the space of an hour, and are far more likely to immerse ourselves in a single more substantial game for hours at a time. Thus while Paradroid (1985) was considered one of the greatest games ever in the 80s, the use of a frequently recurring mini-game within a game today is largely frowned upon, except in open world games whose gigantic sprawling worlds never manage to have enough to do without either a built-in level editor ('virtual Lego') as in Minecraft (2009) or a pile of content that compensates for this excess of space. Such examples avoid accusations of inelegance at the moment because there is an awareness of a core game world with consistent (elegant) practices, with mini-games and the like being merely 'extra content'. I leave open the possibility that in forty years time this approach will seem as inelegant to players as the 80s games mentioned above tend to seem now.

Finally, the aesthetic flaw of rupture occurs when the imaginary world a game builds in the player's head is catastrophically disrupted by an intrusion of something that cannot be reconciled with it. The player is broken out of the experience, and becomes aware of the elements of their play instead of being immersed in the world evoked by those elements. Once again, this is clearly game dissonance, but here the incident tends to be more acute, as our experience of immersion is interrupted ('immersion' being nothing more than a term for our intense mental engagement inflated in importance by our love of videogames). As with all forms of cognitive dissonance there is a clash between mental states, but here it is because the game is making its representations play by different rules. For example, the world implied by the game mechanics and the world implied by the story might pull in different directions - this is the case of dissonance that Hocking was drawing attention to under the name 'ludonarrative dissonance' (see last week). But rupture also happens when a player's engagement is focused upon the game systems and they are forced to confront the story systems instead - those who are not fans of cut-scene heavy games such as Final Fantasy VII (1997) and its successors are acutely aware of this problem!

We generally fail to recognise that our engagement with most game systems is in itself a story-generating activity. All game systems are representative i.e. they ask that we imagine some specific arrangement. It is precisely because games are inherently representative that we make the mistake of thinking there is an unavoidable clash between stories and games - but what we mean by 'story' here is 'a story in the style of a movie or TV show' i.e. a screenplay. The problem is not and never has been an insuperable gap between games and stories, it is that the stories created by screenplays diverge dramatically from the stories that game systems produce on their own. Sometimes this tension is felt as rupture (the imagined experience collapses), sometimes as inelegance (Hocking's complaint about Bioshock is more of this kind), but in all cases it is game dissonance.

The aesthetic flaws of videogames as videogames are therefore examples of Festinger's psychological model of cognitive dissonance manifesting within our play experiences. These can be further understood as clashes in our habits, our player practices - either between those practices we have learned from earlier games, or within the practices a single game is expecting us to learn and execute. In this way, game dissonance is a more precise concept than cognitive dissonance: it zeroes in on the ways our learned habits condition our experience of the imagined worlds of videogames. Recognising this allows us to think about how we can avoid game dissonance through clever narrative design.

Next week: Elegant Narrative Design