Tuesday, January 8, 2019

A NOTE: Illusionism

What’s Illusionism, and why should you care while running a game?

Illusionism is a method of setting building or adventure design that raises the question of player choice or agency. It’s the practice of Game Masters or designers of changing encounters or events in game that will follow a specific path or create a specific scenario. The term itself is sometimes used pejoratively - but like most other things in a hobby or fandom that get people angry it’s a nuanced issue with multiple perspectives. Obviously GM actions that force specific results or events can be a problem for players when the players feel that their choices will lead inevitably to the same results. This is somewhat like the tendency of players in of computer RPGs to skip through cutscenes and dialogue because they expect that decisions about plot developments or the motivations and plans of their characters are pre-scripted or will inevitably lead to the same result.

In a tabletop game roleplaying the plans and motivations of both player characters and NPCs or factions in the setting are an area that tabletop can manage extremely well with a GM there to make the setting react reasonably to unexpected plans, shifts in allegiance or changes in player goals. In a video game the next adventure or scene has already been scripted and designed and cannot be modified should the player decide they want to do something else or approach things in an unexpected way, while a tabletop Game Master can easily change things to adapt a scenario to unexpected player decisions. While such improvisational adaptations might share elements of illusionism, they generally result in players feeling more connected to the setting and events then if the GM had simply run events as initially expected.
Ogre art from the 1st Edition Monster Manual

So essentially the danger with illusionism is that players will feel their actions and desires are meaningless, advancing a plot that the GM has already designed even when they want to take the game in another direction, and the advantage of it is that it can create more dramatic or responsive reactions to player decisions and character plans, goals and personalities. In dungeon crawls the advantages of illusionism rarely outweigh the risks - because the nature of the setting already contains a great many logical, diegetic (that is resulting from the story itself) restrictions on player choice. Many classic types of Illusionism are also more dangerous for the GM to use in a dungeon crawl because the sorts of decisions, especially decisions about character movement and encounters, that it tends to effect are ones that are already a focus of player attention in a dungeon crawl. For players to accept the GM violating the assumed nature of the setting: time, physics and other constants (for example putting the same mundane creature in multiple places at the same time to force a player encounter) it's best that these interventions are secret, unknown and relatively unimportant.

An alternative way of viewing illusionism is as a sort of inverse of ‘simulationism’ - the impulse to make your game as realistic as possible - entirely controlled by realistically modeled rules and chance. In the perfect (and entirely impossible) simulationist game the GM wouldn’t make any decisions, only consult rules and tables of chance and likelihood for even the most mundane events. As much as illusionism’s bad reputation is earned when GM’s use it excessively or clumsily to negate player decisions, simulationism deserves an equally bad reputation when ‘realism’ and the inescapable tyranny of chance are used in ways that make game boring or lead to frustratingly pointless character deaths (such as a GM who claims that characters have a 1 in 100 chance of slipping on tavern stairs or choking to death on a beer).


Illusionism in Dungeons & Dragons has a long history, and while the 1st edition of the Dungeon Master’s Guide focuses more on simulationist play style (often to its benefit), even in 1979 it includes advice on when (rarely) and how to ‘give players an edge’ on some die rolls for the benefit of story, overruling unwanted die results and refraining from letting characters die to bad rolls when the players took reasonable precautions (See page 110). Gygax was aware of these common methods of illusionism, and accepted their usefulness within certain contexts and with a number of cautions: illusionism should “never seriously harm the party or a non-player character” and “ALWAYS GIVE MONSTERS A FAIR SHAKE”. On the subject of character death he was sanguine, writing that even deaths caused by pure bad luck should only be avoided with some significant cost to the character and that “[w]hen they [the characters] have done something stupid or not taken precautions, let the dice fall where they may!” Gygax’s illusionism is reactive, with a primary concern for fairness to both the GM’s creations (NPCs and monsters) and players, as well as the purpose of avoiding player demoralization.

The sorts of illusionism that are mostly frowned upon in classic play circles are largely preemptive or punitive and seek to corral or direct play in the manner the GM deems appropriate. The risk is that in pushing the story forward a GM will trample players’ perceptions that they are making meaningful decisions while also producing a dull game where adventures follow predictable paths to trite conclusions. The abrogation of player choice is the most harmful as it eliminates one of the biggest advantages of tabletop games - the freedom to explore and elaborate on decisions that depart from genre or perceptions of how a scenario is designed to proceed.

An example of how and why illusionism can harm the game is what I like to call Schrodinger's Cult. Often critiques of illusionism focus on relatively discrete elements of an adventure - moving a single encounter from one place to another. The Cult is a large scale failure - rigid design intended to create plot, tension and excitement but which can trap the GM in a terrible situation. The adventure is of the urban sort - a mystery about kidnappings and corruption that will lead to the secret lair of of a necromantic cult beneath the ruined districts of the city. The designer decides that the discovery of the Anciene Cultes Des Ghoules will be far more exciting and dramatic if it includes the betrayal of the party by their allies and proceeds to build up an investigative adventure where there are several factions who seek the party’s aid with various problems, all ultimately leading to the Cultes Des Ghoules. Illusionism comes to dominate because whichever faction the players choose to ally with or befriend will ultimately be revealed as the cult - but until that point it’s just one of several factions - it’s evil only a potentiality.

There’s a good chance that this sort of illusionism will work well, especially with players that are somewhat incautious - the players may trundle along investigating ghoulish crimes, encountering a few cult foes and slowly unraveling the nature of the cult and their hideout location until they bring these facts to their allies seeking aid and the cult leaders reveal themselves - fleeing the characters after a quick tussle and rushing off to their underground lair. What happens when prior to this perfidy the players decide to investigate the factions? Instead of simply doing ‘quests’ or investigating crimes the party stakes out various faction’s headquarters, trails their agents and otherwise seeks to understand their secrets. If the GM wants to maintain the betray scene he needs to keep the party confused about who the cultists are: sowing red hearings for all the investigated factions that make them equally suspicious or concealing evidence of the cults involvement with each faction. The GM of course has ways out of this bind - but they will cost the betrayal scene, which also potentially denies the players knowledge of the Cultes Des Ghoules labyrinth, and they will change the dungeon crawl in that hideout dramatically because they party will have proof of the cult’s identity to present to the other factions - possibly gaining swarms of allies that the dungeon scenario doesn’t envision. Given the potential downsides many GMs will double down on the illusionism, doing whatever they can to keep the adventure running as designed, and the players are suddenly very likely to notice that they have been deceived and their clever plans crushed beneath the needs of story and GM fiat. This is a disheartening result for both players and GM, and the possibility of similar failures should raise the question for GMs and designers of what are the risks and rewards to using illusionism? In an investigation or exploration scenario the risks that the players will choose to investigate or explore the nature of factions and potential allies is high - the danger of Schrodinger Cult becoming a problem for the GM is high.

Lesser or smaller bit of illusionism also hold dangers - small dangers for small transgressions. Players seeing the GM recants on lethal rolls may feel invincible and then resentful when the GM decides to reverse the practice. Alternatively players who see the GM modify rolls to mitigate bad results might suspect a GM is doing the same to target and punish them when a string of lucky rolls makes an enemy exceptionally potent. Players who fear that their efforts to ferret out rumors and avoid dangers are being bypassed with unavoidable ogres in the forest might simply stop caring about rumors and scouting, thinking it entirely pointless. None of these bad outcomes will ruin your game - but they will close of avenues of play and potentially deprive the GM and players of the wonder and fun that variety, the unexpected and the triumph of a cunning plan can bring to table top games like Dungeons and Dragons.


The classic example of a poor use of illusionism is a GM who moves a predesigned set piece encounter to whatever location the characters go to next, to keep a story structure in place, or simply because it’s easier than having more then one encounter prepared. In some systems or play styles a linear string of (likely combat) encounters is expected because complex, tactically thrilling combat encounters are where the majority of play takes place. In such a system everyone expects the GM to offer minimal choice to players - it’s an accepted part of the Ethos, because the excitement is unraveling how to beat the carefully designed encounters with mechanically complex characters and environments.

This is not how a classic dungeon crawl works however. In the maze of the dungeon, player attention is focused on the spatial, faction/role playing and literal puzzles of the location so GM efforts to manipulate these aspects of play are more likely to be noticed, and this is what tends to make illusionism unpopular in this sort of play. Moving about encounters with the goal of creating a more cinematic or climactic story not only makes players feel like their decisions to scout, investigate and plan for encounters are being ignored, but is easily noticed and so denigrates the idea that there’s a comprehensible setting logic the players can grasp and turn to their advantage or analyze their setbacks.


The idea that Dungeons are spatial puzzles is something my last post discussed, but even accepting the essentially geographic nature of exploration, this doesn’t necessarily dispense with the idea that a well designed dungeon might not also include a more narrative plot, scenes and a scripted climax. Why wouldn’t one?

For example, one might have the players invading the lair of the Anciene Cultes Des Ghoules beneath
Ogre Magi from the 1st Edition Monster Manual
the crumbled ruins of a tormented city. The maze of sewers and cellars is very much a dungeon that the party will explore, seeking to stop the fleeing cultists and put a stop to some dastardly plan of human sacrifice. Within the dungeon one could set up series of scenes: spots the fleeing enemy dashes through, the party arriving in to interrupt a foul ritual at its climax or the cult leaders fleeing deeper into the forgotten sanctuaries of the dead. The issue with this sort of plot within the dungeon is that it will either depend on a very linear map or the heavy use of illusionism. Scene will need to be scattered about (usually with triggers in multiple locations for the same scene) so that they proceed in a coherent way or the players don’t miss important clues/moments. Alternatively (or in combination) the map can be made linear, random encounters limited or removed to avoid distractions, and shortcuts, secret bypasses and multiple entrances removed. Even when a designer manages to balance the spatial puzzle of the map with the narrative puzzle it’s a fragile creation - disrupted by crafty players who figure out an unexpected spatial way to bypass important or dangerous scenes. The risk of pick axes, wall breaking spells or even windows, grates or splitting the party allowing players to circumvent scenes and force the GM to reveal the illusionism in likely absurd ways or engage in heavy handed fait to prevent the players clever interventions is high. These dangers of circumvention and disruption aren’t dangers when the dungeon crawl is a purely spatial structure, only when the spatial map and the narrative map overlap. That’s one of the stresses for illusionism, where it starts turning bad and stifling player choice and creativity - not because it’s inherently a terrible GM choice, but because it’s part of a larger friction between two “maps”.

The dungeon map is obvious - it’s something you can look down and see: walls, rooms, stairs, doors and such. As modern humans, used to cities and buildings a dungeon map makes sense to us intuitively - players and GMs alike agree on the way doors and corridors work, and when they don’t the language of architecture allows a common set of shared understanding that allows for comprehensible result. Take the 1989 movie “Die Hard” as an example - it’s largely the story about how a lone hero uses spatial trickery and architectural knowledge to defeat a dangerous enemy. In dungeon crawl terms John McClane’s solutions to the spatial challenges of the Nakatomi Plaza dungeon are excellent, and the shared understanding of John’s player and the GM (that elevators have hatches in them or that there’s maintenance access ladders in air duct and elevator shafts) keep everything reasonable and comprehensible.

A narrative is a sort of map as well, but a bit less clear with fewer agreed upon meanings and structures. Every scene is perhaps a room - certainly they contain the same sort of keys and elements (encounters, puzzles, role playing challenges) as a dungeon room and lead more or less logically to one or more other scenes. Of course the conventions of moving from scene to scene are less obvious then the doors and corridors of a dungeon map, requiring a great deal of flexibility on the part of the GM, a greater comfort with illusionism or metafictional and genre enforcing rules. Still, it should be possible to draw a map (or perhaps flowchart) for a scene based adventure and perhaps a worthwhile exercise to see if it was well stocked with alternate routes and choices through the flowchart and contemplate how much illusionism is really necessary to properly explore the story.

The difficulty of managing, responding to and evaluating player choices as they interact with both a spatial and narrative map increases and it becomes more likely that constraining and limiting player choice will seem very attractive or even necessary to the GM. This si why in the dungeon crawl illusionism that wouldn’t be especially difficult to use in a scene based adventure becomes far more stultifying. Exploration focuses player attention on the imagined environment while illusionism bends the environment in service of plot, a practice that works best when players don’t notice it and perceive it as the GM overriding their decisions.



You can’t totally escape Illusionism, or if you can the price can be high. Rules aren’t perfect, and absurd, game ruining situations (not Total Party Kills [TPKs] - which can be fun in a variety of ways) happen. Illusionism can prevent these situations, or mitigate them, and even back in 1979 people playing Dungeons & Dragons realized this … but … illusionism comes at a cost, and should never be something to fall back on out of ease and habit. Whatever your game style, the sort of brute force GMing techniques represented by illusionism, even if secret from the players, should only be used intentionally. This blog is largely about how to run a very specific type of game - the Dungeon Crawl - based on my experiences doing so, but in that it’s also about using intentionality, rules modification and understanding both the goals and limitations of: mechanics, genre and play ethics to get the play experience or design principles that the GM and players want. Illusionism is a GM tool, but one with the potential to effect the play experience. Sure, one can use illusionism to better shepard players through the hidden doorways of a scene based narrative - but it can also clumsily negate their careful plans or clever solutions to problems. This potential risk means that illusionism isn’t something to be applied lightly, and it’s more useful in some types of play then others.

I’d like to thank John of the Retired Adventurer blog, Ian B., Brendan S. of Necropraxis blog, Paolo Greco of Lost Pages Press, Nick of Papers and Pencils Blog, Eric B. the other folks who discussed illusionism with me on G+.


  1. Mmm. Probably should have read this before adding my comment to your Descent into Avernus post.

    I guess I just don't get this:

    "You can't totally escape illusionism, or if you can the price can be high."

    What is the high price?

    I mean...look, just some cursory internet research shows the number of Americans that died by choking in 2017 (whether on their own beer or otherwise) was about 1.5 in 100,000; the number that died by falling (whether by slipping on stairs or otherwise) was 11 in 100,000. Those kind of odds are so infinitesimal that it doesn't deem modeling with a dice roll, even in game that wants to champion "simulation" over illusionism. Simply put, not everything requires modeling in the game.

    So what is the high price for forgoing (or trying to steer clear of) illusionism? That the monsters won't get "a fair shake?" That my carefully crafted scene/moment gets blown up by unexpected player interference? That a player becomes discouraged by the loss of a character and decides the game is "too hard" and not worth playing?

    I just don't think any of those things are a particularly high price to pay. And apologies if I seem combative on the subject. I'm just confused and in need of explanation. I get the pitfalls of using illusionism. What I don't grok is the costs of going the other way.

    1. Like the impulse towards illusionism the impulse towards simulationism is additive. Simulation breeds more simulation, it breeds rules lawyering - not from lack of mutual trust as illusionism does - because adherence to "the real" becomes more and more important and the GM or designer's fidelity to it is always up for critique. Sure you or I may know to draw the line at rolling to see if the PCs choke on a beer or slip and fall - but the player can always present that possibility, the insistence that maybe the assassin slipped on the roof while approaching the PC's inn widow, or maybe the ogre choked on the head it bit off becomes possibile. It's a 1 in 10,000 chance, but that should be modeled, because it could save the PCs. This is obviously an exaggeration - but what is the threshold of chance that needs a mechanic?

      Sword blades break a lot - should we have a mechanic for that? What damage die does a dagger do - because it seems like it would depend on who was using it and how? Must we model exhaustion in combat? How can a combat round take a minute and allow only one arrow shot - even trained musketeers can fire 3x a minute or something?

      This doesn't only lead to conflict at the table, but mostly to excessive rules, overwhelming mechanics.

      Simulationism breaks down mechanics and gameifcation because suggests that the goal of the game is not mutual storytelling or fun, but simulation. It subverts play to simulation. In the same way that illusionism breaks down trust because it subverts player autonomy and mechanics to predetermined narrative.

      That's my take at least - middle path and all that.

    2. I see what you’re saying. Here’s the thing: from my perspective, both sides require a strong-minded DM (not that D&D play in general doesn’t or doesn’t benefit from such). However, on the side of simulation, the DM has to be an editor, knowing where to draw the line and allow play to happen. On the illusionism side the DM becomes a director, curtailing player agency (through various methods).

      Again, I don’t disagree that there are players that prefer the latter brand of game play, but it’s not one I prefer to promote.

    3. I can't say I prefer excessive simulationism to excessive illusionism, but I think the later is the problem of the present. Not that the time consuming mechanics and encounter design of excess simulationism doesn't lead to excess illusionism - nothing's all one thing.

      Sure both problems can be prevented by good GMing, but good GMing, while not hard, is a bit of a skill (it helps to be good at facilitating meetings and telling campfire stories?) and understanding the risks and pitfalls of illusionism is I think helpful. Especially if one is used to seeing it used, because the play style one is familiar with makes it more necessary. Illusionism is a huge problem for open world/sandbox/location based games but it's not so much trouble in scene based/adventure path ones. A GM coming from the second to the first needs to be more careful about it, and it's baked into a lot of the 5E advice as well as the streaming play ethos.

      For me trying to point out its risks is useful - because thinking about it can help make one a better GM.


Old Games

Let’s talk about old tabletop roleplaying games - specifically the kind of games played in the 1980’s and recently depicted in the nostalgia...