The Art of Steering – Bringing the Player and the Character Back Together

The Art of Steering – Bringing the Player and the Character Back Together

The rhetorics of Nordic larp often imply that role-players play in an intuitive fashion guided by the character, rarely or never contemplating their actions during the game. In reality, however, we are often keenly aware of what we are doing as our characters and why. This paper explores the practice of making in-character decisions based on off-game reasons – also known as steering.

In discussions about role-playing, there is a tendency to treat the character as an entity separate from the player. While we need some kind of separation to understand the contextual difference of killing an orc and adjusting a name-tag, this separation also obscures some important processes of roleplaying. As the participants in a larp enact their characters, the choices they make as characters are not always driven by diegetic (in-game) motivation. The rhetorics of immersion, character and coherence would have us believe that characters in role-playing games, at least when played by “good” role-players, do not let extra-diegetic motivation invade the game world.

In the actual practice of roleplaying however, player motivations seep into the game constantly. The player of a tyrant might choose to play in a more benevolent style when interacting with beginners, or a vampire character might leave an interesting scene because the player needs to find the restroom. These are basic examples of steering, of doing things in a game due to the player’s reasons – rather than the character’s.

While the idea of steering complicates some ideals of what players ‘should’ do, we consider it a critical player skill in most larps. We hope that by naming it, we can provide players with a useful tool to discuss their craft.

We define steering as follows:

Steering is the process in which a player influences the behavior of her character for non-diegetic reasons.

In other words, while the player’s character is an entity within a game world, the behavior of a steering player is motivated by reasons outside the game world. To manage this contradiction, steering players almost always attempt to maintain the semblance of coherence in their character’s behavior.

Specifically, players attempt to ensure that characters maintain the outward appearance of coherence for the character’s actions, from the perspective of other characters first and other players second. In other words, a player who is steering strives to maintain the illusion that the actions of her character make sense as a whole.

Whenever possible, players also attempt to maintain the internal coherence of their understanding of the character. In the above example of the vampire player looking for a restroom, the player undoubtedly fails to preserve internal coherence, but she still seeks to maintain the outward appearance of coherence for other players.

Steering is often subtle and nuanced. As an example, the player of a prison guard might be considering whether her character should pursue a love interest or fulfil her character’s guard duties. In deciding that pursuing the love interest will make for a better game, she subtly decides to heed the pull of the romantic interest more strongly, maintaining her internal coherence while actually influencing her play based on a non-diegetic decision on how to generate better play.

By definition, steering is always intentional. Thus, you can never steer by accident, and it requires conscious choice and effort. The behavior in the above example would not have counted as steering if the player was just deeply focused in the romantic affair and would have never considered the effect on the larger game before deserting her post. Instead, she consciously evaluated the impact of her actions, and then acted towards deepening the romance. This can happen quickly and semi-consciously so that the player can stay in the emotional flow that inspired the choice – but it is still a marked moment the player can identify afterwards. Of course, we do steering decisions so often and so quickly that we often forget about them before the larp is over.

Steering can be used to create good or bad play. Usually such definitions depend on the play culture and the overall dynamics of the game: In a gamist aesthetic, playing to win can be seen as acceptable, while in games focusing on a play to lose aesthetic, the players are expected to steer towards failure. Steering can even be immoral or unethical, for example if a player uses her character as a pretence for stalking another player.

Not all character actions result from steering – only those actions intended to guide the character to a specific effect for reasons that exist outside of the character’s conception of the world. At a minimum, we consider the reflexes and unconscious reactions of the player as external to steering. An example of the difficulty of establishing a line between steering and not-steering is player attraction toward other players: If a character’s choice to pursue a romance is influenced by the desire of the player it could be seen as steering – but only if the player is aware of this.

It is also important to note that steering is something one does to one’s own character. There is by definition no such thing as ‘steering others’. However, through steering her own character, the player can also change the way others are playing and influence the direction of the larp as a whole. Indeed, that is often the goal.

Dual Consciousness

We believe that knowing how to steer properly is one of the most important player skills.

Since steering breaks down the division between the player and the character and exposes the moment-to-moment reality of play, it is a useful tool in taking a brutally honest look at what happens in the practice of larp.

Most of the time during larp runtime, players have the dual consciousness of looking at the event both as diegetic, and as non-diegetic, as play and as non-play. This dual consciousness, or bisociation, informs most of their actions. It is an important part of playing and games; standing with one foot within the border of play and another outside it can not only be powerful, but also instructive.

Viewing something both as play and as non-play not only teaches the viewer about the thing she is looking at, but about the overall structure. This helps in understanding the socially constructed nature of reality as a whole, but specifically it helps in understanding how a game functions. This competence at reading situations on multiple levels is a skill that can be developed in play and applied when steering.

Steering Examples


Physical needs. Food, sleep, warmth, etc.
Looking for someone. Searching for another player to play a scene or to get the car keys.
Documentation. Posing for or avoiding a camera. Filming in characer.
Logistics. Entering hostile territory because that is where the toilet is.
Physical safety. Not running in the pitch-black forest even when your pursuers do.

Smooth Play

Coherence. Preserving the external coherence, even at the expense of your internal coherence.
Legibility. Overplaying emotions to make sure they are conveyed to other players.
Game mastering and fateplay. Pushing the game towards some direction as required by larp design.
Retrospective rationalization. Smoothing over the plot holes of earlier bad steering.
Post-hoc player vetting. Mitigating the perceived damage to the game caused by a ‘bad’ player.
Theme. Accepting that vampires are real in two minutes.

Aesthetic Ideals

Narrativism and dramatism. Making a better story for yourself or others.
Gamism. Winning conflicts, gathering power.
Immersionism. Avoiding heavy game mechanics that might detract from character immersion.
Bleed. Seeking maximally intense emotional impact.
360° illusion. Avoiding the sight of the parking lot in fantasy games.
Play to lose. Sharing secrets loudly for eavesdroppers to hear them.

Personal Experience

Boredom. Looking for stuff to do. Picking up fights.
Staying in game. Not leaving the haunted mansion even when two people are dead.
Relevance. Getting closer to the perceived core of the game, or seeking more agency.
Overcoming disabling design. Deciding that your character wants to become a revolutionary only after you realize that most characters only talk to revolutionaries.
Avoiding the same-old. Not rebelling against the tyrant in two games in a row.
Attraction. Getting to play with skilled or cool players.
Player status. Doing things likely to increase one’s status as a player.
Shame. Not wanting to do or to be seen doing certain things, even as a character.

Ethical and Unethical

Consent. Observing a slow-down safeword such as “yellow” or “brems”.
Trust. Creating a safe situation in which to play demanding scenes.
Inclusiveness. Including characters that have nothing to do at that moment.
Harassment. Using the larp to stalk another player.
Revenge. Killing your character because you killed mine in an earlier game.

There is nothing mysterious about this process. It simply means that a player is able to see at the same time both the cheerful friend who gave her a lift to the larp wearing old army surplus clothes, and the frightful commander of the space station her character could never approach. Both of these things are true at the same time. Recognizing the difference between the diegetic and the non-diegetic is the difference roleplaying is built upon. However, that separation is not actual, but rather one made in interpretation.

The idea that one realm, the non-diegetic, is allowed to influence the other realm, the diegetic, may seem wrong, even immoral. Indeed, the idea of steering may seem like anathema to roleplaying. Is not the key tenet of roleplaying the idea of portraying a fictional being in a fictional setting – without the petty motivations a player may have outside roleplaying? Yet steering is not a bad or an undesirable thing to do. In fact, many players steer almost all the time when they are playing. The diegetic world of fantasy never maps completely on the physical world, nor does the body of the player completely become that of the character. The draw of larp is that it is not-real and that it feels real.

Steering and Immersionism

The concept of steering – and the criticism of motivations originating with the player – emerge from a tradition that values character immersion as an ideal. Immersion is perhaps most frequently defined as moments when player forgets herself – when the dual consciousness of simultaneously being a player and a character fades away and player only focuses on being her character. This experience has been characterized, for example, as the player pretending to believe that she is her character (Pohjola 2004) and as bracketing the everyday self (Fine 1983).

It has been compared to ideas such as flow (Hopeametsä 2008) and wilful suspension of disbelief (Pohjola 2004).

In the Manifesto of the Turku School, Mike Pohjola (2000) argued that character immersion should be seen as the ideal aesthetic of the larp. But with an ideology that forbids dual consciousness comes some baggage – it prohibits steering:

Sometimes it might be fun to do something that is not in strict accordance with the character, but – unless the GM has specifically asked you to do so – THAT IS FORBIDDEN.

Mike Pohjola, Manifesto of the Turku School

The psychological idealism focused on immersion has faded since the turn of the millennium. It is now commonly acknowledged in the Nordic larp discourse that even when player’s focus is on her character she still does not become the character. The idea that someone could use character immersion as a moral justification for punching another player in the face would universally be found ridiculous.

But even as full character immersion has been found impossible, this rhetoric of playing true to the character has persisted. The dogma of character fidelity can be seen whenever players discuss whether it is realistic that the king fell in love with the peasant girl, or whether it was credible that mortal enemies joined forces in order to win the war against orcs.

However, as the player cannot psychologically transform into her character, the problem of Pohjola’s statement is that it is impossible to determine which actions are in “strict accordance with the character”. Even as a player, one can determine several credible courses of action for almost any situation the character can be in.

This uncertainty and ambiguity about what would be fitting for a character is what makes steering possible. If there was always one right choice for a character to make, steering would be meaningless. It is this very uncertainty that is the site for steering – the minute choices a character makes. Steering is rarely about making major life choices and often about pushing a discussion gently in a new direction.

Indeed, the skill in question is not entirely dissimilar to the skills one needs when steering conversation away from difficult topics in an everyday social situation like a polite chat with colleagues over coffee. When you understand that you have a potentially inappropriate joke that is perfect for the situation, you still decide whether or not to tell it. Sometimes that decision may be done very quickly, subtly, or half-unconsciously.

The strict reading of immersionism presented above appears to be incompatible with the idea of steering. However, contemporary immersionists do not argue that character immersion is an overwhelming and persistent state. Rather, it is seen as an aesthetic ideal and a goal to strive for when playing.

From this perspective, we actually argue that some amount of steering is even a requirement for immersionist play. The immersionist player seeks to ignore and forget the fact she is larping while doing so. This wilful suspension of disbelief requires the player to maintain internal coherence of her character: It might be hard to forget yourself and become a medieval queen if you are standing on the balcony with the clear view to the parking lot. Getting a powerful immersionist experience of committing a tragic suicide is more likely if you consciously choose to commit one.

Or, as Pohjola wrote himself years later:

Whenever we see interesting developments that will enhance our story, our experience and our character immersion, we have to jump at the chance to engage with them. Otherwise we’re not doing anyone any favors. In a larp you should be your own game master and help your own character immersion by building a better game for yourself.

Mike Pohjola, You’re in Charge of You

The reason why the idea of steering is sometimes seen to be at odds with The Manifesto of the Turku School is probably historical. It was written at a time when player motivations were seen to be influencing Finnish roleplaying too much.

Although it was a response to Dogma 99 (Fatland & Wingård 1999), it was actually directed against gamism (steering to win), dramatism (steering to create interesting scenes and stories), and bad roleplaying (for example, steering on the expense of coherence).

The idea of steering shows how rare moments of real immersion and flow are. By lifting the dogmatic ways of talking about the play experience tinted with the idea of immersion, it helps account for many of the actions a player takes during runtime. By shifting emphasis from the ideals of playing to the actual practice it illuminates what we really do while roleplaying.

Designing for Steering

The idea that larps contain characters that are there to direct the play is as old as larping itself. This is what the non-player characters and other game master controlled actors have been doing since the beginning of larping (cf. Stenros 2013). However, player characters have done this since the beginning as well – even if it was not always directly discussed.

Explicit steering instructions have been a part of the tradition of Nordic larps at least since the emergence of fateplay (see Fatland 2005), a style of making larps where players are given some instructions on how to behave in certain situations – the character Claudius, for example, was fated to die in the larp Hamlet (2002).

More recently, larps such as The Monitor Celestra (2013) have introduced the idea of having large amounts of characters with pervasive and persistent steering duties. In the Celestra, which featured strict naval and military hierarchies, higher-ranking officers were expected to generate play for their subordinates. For example, the commanding officer of the Colonial Navy was instructed as follows:

As the Major in charge, your foremost duty is to act as a game master for bridge and CIC personnel, generating interesting play and putting flavor into the tasks of running the ship […] Always keep in mind that your job isn’t to be an effective Major, but to be a good player/game master, and enable interesting action for others.

Character material, The Monitor Celestra

While all players had similar duties, the higher the character was in a hierarchy, the stronger the expectation of steering was. This was of course a practical solution: By having the Major to steer hard the game masters could alter the course of the entire larp, as she could use her diegetic authority to impact the game for all her subordinates, shielding them from the need to steer.

This mechanic worked rather well for members of those hierarchies, especially compared to older and more selfish play styles (see Fatland & Montola in this book for a detailed discussion).

Although the top brass was expected to steer the most, Celestra explicitly encouraged following the philosophy of play to lose, which basically expects everyone to steer in the larp. The following play instruction was given under the heading “Rules” in the briefing materials:

You are expected to play to lose, prettily. In a game where experiencing the journey is the whole point, winning is moronic. Losing, on the other hand, is dramatic and cool since it puts a spin on the story and contributes to emotional impact.

The Monitor Celestra Briefing

These games have established a new steering norm along the ones such as gamism, dramatism, immersionism and bleed: In these games, players are expected to steer in order to play to lose. This anti-gamist stance can arguably contribute to many other play aesthetics, as it “puts a spin on the story” for dramatists and “contributes to emotional impact” for immersionists and bleedhunters.

Obligatory and Heavy Steering

Sometimes it is every larper’s obligation to steer. Barring some unusual arrangement, role-players share an almost universal implicit obligation to steer for coherence. Different game styles have different conceptions of what coherence is, yet internal logic of some kind is valued in all larp cultures.

In some roleplaying games, especially larger larps with less-tightly organized plots, what would be seen as a significant coherence conflict in another game may be glossed over by all players concerned as they acknowledge tacitly that a conflict has occurred by choosing not to fix it, as it would require too much work on the part of disparate groups of players.

In other games, often smaller or more tightly plotted, it would be seen as a serious problem for such a breach of coherence to occur to start with, requiring either heavy steering by all parties to fix immediately or possibly (in some play cultures) a break of play so the ‘truth’ of the situation can be decided directly by the players off-game. Usually, when coherence cannot be achieved by steering, the next solution is to ignore the problem; to steer play away from the mess.

Two examples can help clarify this. In long-running campaigns the character arcs can become increasingly improbable. Like in soap operas and superhero comics, certain ancient acts may be de-emphasized by those character’s players.

In larps this works particularly well, as no one can go back in time three years to check and nitpick what actually and specifically happened. In larps that use them, mechanics like experience points can also shift balances between masters and apprentices or parents and children, if players put in different amounts of play time.

Another example comes from the second run of The Monitor Celestra, where at one point the key to the hyperdrive was stolen, and a dozen characters got involved in recovering it from the characters who used it as leverage in a negotiation. Problems arose because the game organizers held that no such key existed, as some player had improvised it up. As the characters raced to solve the issue the gamemasters ignored it; as far as they were concerned, this plot did not exist.

The game masters could still not solve the problem simply by issuing a decree, because too many characters were involved with the key.

In the end the issue was solved twice in the game – once very rapidly due to game master pressure and again by some characters not being aware of the first time – and only then were all the characters able to move on. No equifinal understanding on what actually happened can be produced.

When characters are forced to steer hard, it causes wider ripples in the play. Specifically, one player steering hard may leave another player confused about the steerer’s character’s identity, her relationship with the second player’s character, or the events of the larp. This is sometimes unavoidable, especially when a player is forced to steer in a character-breaking way. This is a specific kind of game incoherence associated with steering that many players, especially heavily immersionist players, may consider unacceptable.

Steering can be characterized as character-breaking steering when the player cannot maintain her internal sense of coherence. For example, if the player is executing a game master directive that is important for the larger plot of a game but finds their character has moved away from the gamemasters’ expectations of who they would be when the instructions were originally specified, they will need to steer their character to ensure they fulfill their obligations to the game, but will do so knowing that this action does not make sense for the character.

Likewise, a player may realize part-way through a game that they have played themselves into a corner, and if they wish to continue playing or return to the main plot of the game, they will have to simply reinvent part of their character. While to be character-breaking, this shift need only be incoherent to the player, when done poorly (or under extreme circumstances) it will often result in the character also appearing incoherent to other players.

In order to repair the disruption created by heavy steering, players sometimes engage in retroactive rationalization, wherein they decide on the thus far unvoiced rationale for choices they have already made to maintain the appearance of coherence. For instance, a player who forgot their character’s sidearm may later steer, deciding that their character was feeling especially secure that morning, and thought they would not need it.

If the player discovers that this will cause a coherence problem with other players expectations, they may engage in retroactive rationalization retconning – if they have not already told the other players, they can changing their prior retroactive rationalization. In this example, the player might decide that instead of being supremely confident, their sidearm was actually stolen, allowing them to integrate with a game mood of suspicion and paranoia.

The roleplay agreement (Sihvonen 1997), the social contract that participants treat the player and the character as separate entities and refrain from making judgement about one based on the other, is a cornerstone of roleplaying. Without it establishing trust amongst players to also engage in anti-social behaviors, like playing a villain, can be hard. The concept of steering does not obliterate the role-play agreement. However, it needs to be modified; the separation need not be between player and character, but diegesis and non-play. Indeed, it is the character that acts as an alibi for steering. The player can choose what she wants to do or what best fits the larp, and as long as it somehow makes sense in relation to the facts of the character thus far established, it is acceptable.


Sensitivity to other players – knowing when and how to steer – is a key player skill. A considerate player can create play for others, pace drama, include others players, support beginners, and avoid hogging plots and secrets. A good larper steers in a nuanced way that is invisible to other players and does not damage the coherence of play. Steering is not a bad thing to do in a game, and most of us steer much of the time while we are playing.

Just like good steering contributes to the game, refusal to steer can detract from it. If one player does not steer, her fellow players may be forced to steer even harder to sustain the game. It is not rare to encounter a selfish player in a larp with a preference to avoid steering who expects other players to accommodate her play style. The other players may end up steering hard to maintain play and allow her to preserve the immersive flow instead of caring about the overall game.

Steering is a skill and not all players are good at it. Steering coherently and reliably requires thinking and performing simultaneously on two, three, or more levels while maintaining an accurate model of both the perceptions of both other characters and their players.

Players holding on to an ideal of playing entirely without dual consciousness may even argue that the expectation of steering ruins their game. Steering is perceived by some players as distancing them from their character. In part, the degree of distance perceived may relate to how quickly players are able to slip between different levels of play.

It is not necessarily the case that more intense emotional experiences require less movement between levels of player consideration, but this appears to be true for some players. Some players and some game contracts may consider steering to be cheating, as in those contexts, only diegetic concerns are considered to be acceptable as motivations for player choices. We believe that such contracts are often self-deceptive, and that acknowledgement of the role of steering in play is critical to designing for character immersion in the context of a coherent, functional game.


We would especially want to thank Juhana Pettersson as well as other organizers and participants of Larpwriters’ Winter Retreat 2014.


  • Hamlet (2002): Martin Ericsson, Christopher Sandberg, Anna Eriksson, Martin Brodén, with a large team. Interaktiva Uppsättningar and riksteatern JAM. Stockholm, Sweden.
  • The Monitor Celestra (2013): Alternaliv AB, with Bardo AB and Berättelsefrämjandet, with a team of 85 people. Gothenburg, Sweden.


Steering is the process in which a player influences the behavior of her character for non-diegetic reasons.

This article was initially published in The Knudepunkt 2015 Companion Book which was edited by Charles Bo Nielsen & Claus Raasted, published by Rollespilsakademiet and released as part of documentation for the Knudepunkt 2015 conference.

Cover photo: Theory jam at the Larpwriter Winter Retreat 2014, by Johannes Axner, is licensed under CC BY-NC-SA 4.0.

Become a patron at Patreon!


Markus Montola (b. 1978) is a Helsinki-based game designer, scholar, and entrepreneur, and a long-time contributor to the Nordic larp theory discussion.
Jaakko Stenros (PhD) is a game scholar working at Tampere University. He has published eleven books and numerous articles, has co-created a few larps, and has curated exhibitions at the Finnish Museum of Games.
Eleanor Saitta (b. 1978) is a hacker, artist, designer, writer, and barbarian making a living and vocation of understanding how complex, transdisciplinary systems and stories fail and redesigning them to fail better.