The Evolution of the Threefold Model

by John H. Kim

         In a prior article, I looked at "The Origin of the Threefold Model", which covered roughly the period from 1994 to 1998 on the UseNet newsgroup (rgfa). Here I want to look at different directions that people have taken the ideas which started there.

         The rough idea of the Threefold Model has taken root in several places, but the ideas have also changed, in some cases drastically. It has not evolved into a single canonical form, but rather (like real evolution) has influenced several branches which are now quite distinct from each other. The major branches I consider would be "GENder theory" (by ScarletJester as posted on Gaming Outpost), "GNS" (by Ron Edwards), the "3D Model" (coined by Mike Holmes and elaborated by myself), and the "Three Way Model" (by Petter Bøckman).

Year (roughly):
     1997    1998    1999    2000    2001    2002    2003    2004    2005

                                                             ___Later GNS___
                                         ___Middle GNS_____/
                                        /                  \
                      ___Early GNS_____/                    \___3D Model____
                     /                 \
                    /                   \___GENder________
RGFA Threefold____/____________________________________
                                            \___Three Way___________________

The Dilemma of the Threefold

         I think the biggest problem which people found about the original Threefold is that it was unmotivated. That is, the three vertices represented patterns of decision-making, but they did not represent any clear goals of play. Fundamentally, the model was developed over the question of how decisions are made, both by players and by the GM. i.e. How do you decide what happens within the fictional reality? When asked "What happens next?", what type of logic do you use to answer?

Each of these are methods for decision-making. However, they do not represent motivated styles for their own sake.

         The later models which were influenced by the Threefold tend to interpret these approaches more as goals -- i.e. as values for what the players get out of gaming, rather than just patterns of what is put in. So, for example, Dramatism is sometimes interpreted as "art for art's sake" while Simulationism is sometimes interpreted as realism for its own sake.

         Another issue with the Threefold Model was that there were many open questions about the scope of the model. Does it apply to both GM decisions and player decisions? Does it apply to decisions made during campaign design, or only to decisions made during dynamic play? The original model had no defined scope -- i.e. the terms could in principle be applied to any area, though how useful they would be varied. This was an issue because it meant that there was not a canonical meaning of a "Dramatist game".

Sorcerer and Early GNS (1998 - 2000)

         Within a year, the ideas of the Threefold model started having influence outside of rgfa. As Ron Edwards describes it, he had heard about the Threefold Model some time in 1997 - 1998. [1] As he put it,

The basic notion of the Threefold impressed me: it was time to talk about goals and priorities independently of everything else, then to see whether everything else flowed to and from them. I started several discussions about this in the Sorcerer mailing list, some of which survive and are archived at the Sorcerer website.

It then became a topic in the mailing list for the game which he had designed, Sorcerer. However, his terminology also included the types of action resolution used by Jonathan Tweet in Everway -- namely Drama, Fortune, and Karma.[2] This became became known as DFK in related discussion. He wanted to develop the Threefold split, but since he already was using the term "Drama" for something else, Ron changed the name from "Dramatism" to "Narrativism". On the basis of these discussions, he wrote his article "System Does Matter" in January of 1999. [3]

         This and another essay ("The Nuked Applecart") were originally slated to appear in a GPA newsletter project. However, that project eventually folded. Both articles were soon published on the Gaming Outpost website. As the theory appears in this form, it is fairly close to the original Threefold Model. The definitions given in that essay are:

  • Gamist. This player is satisfied if the system includes a contest which he or she has a chance to win. Usually this means the character vs. NPC opponents, but Gamists also include the System Breaker and the dominator-type roleplayer. RPGs well suited to Gamists include Rifts and Shadowrun.
  • Narrativist. This player is satisfied if a roleplaying session results in a good story. RPGs for Narrativists include Over the Edge, Prince Valiant, The Whispering Vault, and Everway.
  • Simulationist. This player is satisfied if the system "creates" a little pocket universe without fudging. Simulationists include the well-known subtype of the Realist. Good games for Simulationists include GURPS and Pendragon.

Note the definition of Narrativism which is defined by "story", while the definition of Simulation depends on lack of "fudging". This seems to match with the rgfa distinction, if we can assume that "fudging" means meta-game influence on in-game events. However, as an independently discussed branch, this came to be known as "G/N/S" and eventually simply "GNS" (for Gamism, Narrativism, and Simulationism).

Gaming Outpost and GENder (2000 - 2001)

         This essay was debated extensively on the Gaming Outpost forums. At some time during 2000, the concepts of Stances were being integrated into discussion -- an expansion from the original Narrative Stances proposed by Kevin Hardwick.[4] Around this time, as well, M.J. Young made an attempt at clarifying GNS preferences with his "Gamer Preference Quiz".[5] His quiz adapted three questions suggested by Ron Edwards, and added another three. This lead to a split between M.J. and Ron, however. As M.J. puts it,

At that point in time the argument seemed to stress whether one player (or one game) could be dedicated to more than one "goal" of the model, Ron suggesting that ultimately everyone would land in exactly one category and I insisting that my experience showed considerably more variation in the play of individuals. I think we ultimately agreed that some players stick primarily to one category and others wander by prioritizing different aspects at different times, but I won't swear that was in a specific sense "agreed".[6]

The preference quiz is at least quite concrete in its ratings, with numerical ratings for different preferences. However, many people felt that answers depended too much on what the imagined circumstances were.

         Early in 2001, a variant model was developed by a poster known as Scarlet Jester. In previous discussion, he had been an advocate of diceless play. He was critical of GNS as expressed in the "System Does Matter" essay, and formulated what was considered a competing theory which he called "GENder" theory, substituting the term "Explorativist" for "Simulationist". He also compared with the original Threefold Model, here referred to as "GDS". As he expressed the difference:

GDS is about decisions. The Ron Edward's model is about techniques. GEN is about desires. Absolutely. Why did it evolve this way? Taking guesses based on information available I'd go with the following: I think that GDS got changed to the Ron Edward's model because Ron wanted a model to design games with, focusing on those aspects of games that he finds most interesting: stance and power distribution. The Ron Edward's Model got evolved to GEN because my group realised (and I believe RGFA understand this too) that the techniques that the Ron Edward's model advocates as the points of it's triangle are actually applicable to any player desire. That is, the Ron Edward's points are actually bottom level techniques that support top level desires (or decisions in GDS).

I consider this a vital point to the shift to the approach. The rgfa Threefold was explicitly about how decisions are made, but didn't probe underlying reasons for such tendencies. Moreover, it was not expressed in terms of what decisions were made: i.e. the distribution of power and credibility. In the "System Does Matter" essay and subsequent discussion, Edwards emphasized the relation to game design. Scarlet Jester's GENder model emphasized the points as goals to be pursued rather than just as procedures for decision-making. Interestingly, Ron Edwards rejected the change to the taxonomy, but later adopted the concept of exploration as the central focus of "Simulationism". While it was viewed as a rival theory, Ron adopted much of GENder theory in the next formulation of GNS (see below).

         GNS remained a controversial point on Gaming Outpost, to the point of discussion of it eventually moving off of that forum. As M.J. Young writes,

I suspect that it was tension between his [GO representative Nick Lalone's] annoyance at the dominance GNS and related game theory discussions had on the forums here and what has by others been called the "Cult of Ron" that drove the discussions first into the Sorcerer forum here and then to The Forge.

A number of objectors felt that the GNS discussion tended to look down on the Simulationist and Gamist modes, while elevating Narrativism.

Development on rgfa (1999-2001)

         Meanwhile, rgfa began to slowly dissolve as a community starting in 1999. Although the 1997 gathering in Canada was fun for those who went, plans for future meetings dropped. The causes of this are probably many, but the upshot was that polite conversation ended and there was little change in people's positions and many ad hominem attacks. While the Threefold terminology was firmly established, it had a number of strident critics. Intriguingly, the controversy had the opposite split as Gaming Outpost.

         On rgfa, the Threefold Model was preferred primarily by posters who identified with Simulationism. Those who identified with dramatic or story values -- such as David Berkman and Kevin Hardwick -- tended to object to the model. A common complaint was that it cast story priorities as lacking in believability. The FAQ specifically denies this -- but many felt that the Threefold in usage was different than the FAQ definition. There were also those who felt that Gamism was looked down upon in the theory, notably Brian Gleichman. He felt that it was important to define Gamism in terms of valuing tests of player skill, not as competitiveness or rules-following for its own sake. While some posters welcomed his ideas, he continued to criticize the Threefold, eventually deciding that the model was better off dropped.

         Another common objection was that the Threefold categories inherently pidgeonholed gamers. Essentially this was an argument against any categorization or theory. They argued for no models at all, rather than any changes to the Threefold Model. I think partly this was hampered by a lack of purpose to the rgfa forum. Rgfa arose spontaneously out of debate, and had no constructive purpose as a group.

         There were several discussions of a fourth "Social" axis or mode. This would be decisions made to please another player, with the canonical example being not killing a PC because the player had a bad day and would feel bad. However, the group was unable to reach any consensus on how to integrate this into the model. The problem was that all concerns could potentially be social -- i.e. a GM might decide to choose based on story because she knows that her players value story. That is a social decision but also a Dramatist one. As originally conceived, the Threefold axes indicated the logic used by an individual to determine answers about the game-world. Social concerns seemed to be at a different level, but which was never incorporated.

         With increasing discord and periodic flames, a number of the posters to rgfa began to leave. Around mid-2001, rgfa began to experience lags of several weeks with no posting, followed by threads with subjects like "Hello... Hello...". There were still periodic debates by a limited set of posters, but in my opinion the spirit of earlier debate had died.

The Forge (2001-2005)

         Ron Edwards and Ed Healy created the original site for "Hephaestus' Forge" in 1999 as a site devoted to independent role-playing publishing. However, Gaming Outpost and the Sorcerer mailing list remained the forums for discussion. In early 2001, Clinton R. Nixon and Ron redesigned the site as "The Forge", including a set of web-based discussion boards. From the start, the boards reflected GNS as an central issue with its own dedicated forum. During the first year, Ron Edwards and Hunter Logan began to head a efforts towards an FAQ document ("GNS 101") on the GNS model. However, there was a falling-out between Hunter and Ron, and the FAQ efforts were dropped. Hunter eventually decided that GNS and the Threefold were not useful as models, and instead came to advocate a longer list of 15 player goals.[7]

         In October 2001, Ron posted a more lengthy essay entitled "GNS and Other Matters of Role-Playing Theory".[8] The new essay reflected a changed and expanded view of the GNS definitions. Three vital new concepts were:

  1. The concept of Exploration adapted from Scarlet Jester's GENder theory.
  2. The concept of Dramatic Premise adapted from dramatic theory author Lajos Egri.
  3. The concept of stances adapted from Kevin Hardwick's narrative stance model originally posted on rgfa.
In this essay, "Exploration" is now defined as a core activity of role-playing, and more specifically exploration of the set of Character, System, Setting, Situation, and Color. "Premise" is defined as the primary goal of the participants -- i.e. whatever a participant finds among the elements to sustain a continued interest in what might happen in a role-playing session. In light of these two concepts, the three GNS categories are redefined as:

  • Gamism is expressed by competition among participants (the real people); it includes victory and loss conditions for characters, both short-term and long-term, that reflect on the people's actual play strategies. The listed elements provide an arena for the competition.
  • Simulationism is expressed by enhancing one or more of the listed elements in Set 1 above; in other words, Simulationism heightens and focuses Exploration as the priority of play. The players may be greatly concerned with the internal logic and experiential consistency of that Exploration.
  • Narrativism is expressed by the creation, via role-playing, of a story with a recognizable theme. The characters are formal protagonists in the classic Lit 101 sense, and the players are often considered co-authors. The listed elements provide the material for narrative conflict (again, in the specialized sense of literary analysis).

Note that this is an essential change. Simulationism still has a note that players may be concerned with internal logic, but it is not part of the definition.

Scandanavian LARP

         In a completely different venue, the ideas of the Threefold Model had been taken up in the community of live-action role-players in Scandanavia (Norway, Sweden, and Finland). Petter Bøckman wrote an adaptation of the Threefold Model which was published in the book, "As Larp Grows Up", published for the Knudepunkt 2003 convention.[9] However, at that point it was already referred to as a "classic". In commentary on Petter's adaptation, Morten Gade said the following:

     This theory is probably the most classic of them all. Heck, it's hardly a theory anymore, as the concepts of gamist, dramatist and immersionist have gone into our daily larp vocabulary.

     In the foreword to this anthology, we there is nothing as practical as a good And if any theory can prove this to players and organisers, it is the three-way model, which most players can actually relate to. Oh, I could go on about the beauty of this theory.

     However, we should be aware of the bias in the theory. Or perhaps not in the theory itself, but in the readings of the theory. Because more often than not, this theory is used to legitimise an attitude about 'good' and 'bad' players. Some places, dramatists are seen as the best players of all. Other places, its the immersionists. Very seldom, it's even the gamists. But we should be cautious about this.

         In his adaptation for Scandanavian LARP, Petter Bøckman substituted the term "Immersionist" for "Simulationist". The definition he gives is:

"Immersionist" is the style which values living the roles life, feeling what the role would feel. Immersionists insist on resolving in-game events based solely on game-world considerations. Thus, a fully immersionist player will not fudge rules to save its role's neck for the plot, or even change details of background story irrelevant in the setting to suite the play. An immersionist organiser will try to make the plots and setting such that they are believable to the players.



  1. Ron Edwards, Re: Origin of the Threefold (August 3, 2003)
  2. Jonathan Tweet, Everway, Wizards of the Coast (1995)
  3. Ron Edwards, System Does Matter (1999)
  4. Kevin Hardwick, "Narrative and Style" (July 11, 1995)
  5. M.J. Young, Gamer Preference Quiz
  6. M.J. Young, Re: Looking for GENder theory (November 8, 2002)
  7. Hunter Logan, "Player Goals: The Impossible Dream #3" (2003).
  8. Ron Edwards, "GNS and Other Matters of Role-playing Theory" (2001)
  9. Morten Gade, Line Thorup, Mikkel Sander (editors), "As Larp Grows Up: Theory and Methods in Larp" (2003)

John H. Kim <jhkim-at-darkshire-dot-net>
Last modified: Wed Nov 16 13:21:39 2005