3 Designing for Role Play

Dungeons & Dragons as first published did not employ the construction role-playing game or even role playing at all: the closest it comes is an offhand mention that “players must decide what role they will play in the campaign, human or otherwise, fighter, cleric, or magic-user.”1 But that usage would have broken no new ground in 1974, as the exact term role-playing game had been used to describe political wargames conducted by the military since at least a decade earlier.2 As a cluster of games exhibiting qualities similar to D&D entered the market, reviewers and fans inevitably began informally negotiating a name for this new genre that was “not strictly a ‘war’ game” anymore, the legend on the D&D box notwithstanding.

In 1975, George Phillies wrote an article that described D&D as adding a “fourth dimension to the wargaming scene,” beyond the existing ones of board wargames, miniature wargames, and Diplomacy, which he linked to a recognized wargaming phenomenon: “The popularity of D&D arises from its ability to appeal to the ‘Rommel Syndrome’—the feeling that one actually is the character represented in the game” (AW 2 (8–9)). Gary Gygax replied in July 1975 that “Phillies finds that the appeal of D&D might rest in its fulfillment of role playing, i.e., allowing participants to imagine themselves as some super-powerful (or just plain extraordinary) character in a fantasy world” (EU 9). Gygax did not share Phillies’s view, arguing instead that the game’s appeal lay in its “constant challenge,” furnishing a “never-ending exercise in problem solving.”3 But after Gygax’s casual use of role playing, it began to creep into reviews of titles such as Empire of the Petal Throne and En Garde in the fall of 1975 as a shorthand way of expressing their similarity to D&D. The review of En Garde in Strategy & Tactics 52, for example, called it three things: a “boardless, role-playing, free-form system.”

So, what did people think that role playing meant at the time? Its everyday connotation of assuming a character during play would not clearly delineate it from existing ways the two cultures had gamed or, indeed, as people would soon point out, from games such as Monopoly. Gygax seemed to dismiss it in 1975 as an identification with character bordering on wish fulfillment. But there was no standing consensus on the meaning or implications of the term: it was born adrift and towed around by a discordant pluralism of voices spanning the two cultures. Because even the genre’s founding game remained silent on this matter, no authority could summarily settle this dispute. Reviewers could recklessly accuse any work of being a role-playing game, and designers could similarly slap that label onto their products with little fear of contradiction. To learn what early adopters meant by role playing, we must survey the furious period of design innovation that immediately followed the release of D&D, listen to how commentators handled the term, and examine how the community tried to encourage players to adopt roles.

As TSR faced increasing competition in the games market and became more protective of its trademarks, rival firms embraced the term role-playing game as a euphemistic way of claiming kinship to D&D without running afoul of any legal concerns. The publisher of Tunnels & Trolls (1975) first paved that path in its advertisements in 1976, and the follow-up title Monsters! Monsters! explicitly referenced role playing in its text. This may have initially prejudiced TSR against the term, but it quickly swept the community; before the year’s end, TSR itself put role-playing game on the cover of one of its new games, Metamorphosis Alpha, and in 1977 the Holmes Basic version of D&D would now identify itself as “Rules for Fantastic Medieval Role Playing Adventure Game Campaigns,” with no wargames in sight.

Once the first commercial products began to identify themselves as role-playing games, the community struggled to isolate the role-playing element in this family of games. But the market did not settle quickly or exclusively on the label role-playing game, and alternative terms in use shed light on the way people positioned this new phenomenon. For example, Mark Swanson floated the term Ego Involvement for the genre early in 1976, though it saw little uptake (WH 2). Later, the descriptor adventure game nearly overtook role-playing game as a designation for this commercial category in the hobby industry, as it encompassed board games distributed by many of the same companies and implied nothing about playing roles. Some games distanced themselves from frivolity by expelling the play from role playing to yield the longer and more academic term role assumption game.4 An obstinate few would even defend wargame as the proper name for these games in the face of widespread belief in a transition to a new category. But, most importantly, through the end of the 1970s, the three-letter acronym FRP for fantasy role playing remained more common in fan literature than RPG for role-playing game. As play increasingly encompassed system elements not published in D&D, FRP became a descriptor for the entire hobby, a set of common and essential practices that transcended any particular published design—albeit the acronym was often found in the mouths of designers promoting some upstart commercial offering. That it was a “game” was something to downplay, emphasizing instead the roots of these practices in fantasy literature; only when grammatically necessary would anyone call something an “FRP game.” This choice of terminology became one more front in the battle to champion a proper orientation of D&D toward games, stories, and roles.

So perhaps a better question is, what did it mean to early adopters to play a role? For practitioners who came into the hobby from wargaming, it had to imply a different relationship with a character than the one they would ordinarily establish with a given unit counter or miniature figure representing forces on a battlefield. Many early commentators argued that the connection of a player to a character in D&D was fundamental to its experience and popularity. For example, Steve McIntosh observed in A&E 18, “Most D&D players identify with some of their characters so much that the character becomes an extension of themselves into a fantasy world, and the statement ‘my character killed a dragon’ becomes ‘I killed a Nasty Ferocious Dragon!!’” The character’s adventures and advancement became the player’s personal accomplishment. This visceral feeling of protagonism cut both ways, however: “In the same manner,” McIntosh continued, “defeat can really hit where it hurts.”

No doubt that identification was strengthened by one particular form of statement of intention: speaking in the voice of a character. A player’s intention that a character say some specific phrase to a fellow party member or nonplayer character often took the form of a first-person statement, as at the end of the sample transcript in D&D the caller addresses his compatriots with the aside, “Onward, friends, to more and bigger loot!” The resulting bond between player and character could become so intense that some referees found this sort of statement an impediment to play: Lee Gold wrote in A&E 14, “I’ve found it helpful to forbid people to speak in their own persona while playing D&D. All remarks must be made about their character—in the third person. This keeps them focused on the character—and prevents them from identifying too heavily with the character.” This restriction would puzzle Mark Chilenskas, however: “How does one play a character in third person? This seems impossible to me, and would interfere with the role playing for the players even if it is possible, so it must be a bad idea” (WH 9). For Chilenskas, playing a role entailed that sort of first-person protagonism, understanding the nature of characters well enough to speak in their voice.

To play a role meant playing a specific character, one with distinguishing qualities. In order to model characters in system terms, the earliest role-playing games assigned values to various attributes that defined not just characters’ abilities but also their nature. This quantification of character triggered a new dimension of play that began to manifest as wargames shifted into role-playing games: weighing a character’s nature and incentives before proposing statements of intention in game. Players would let these attributes influence what their characters would do, much as early players of Fight in the Skies might base game decisions on some qualitative sketch of a pilot’s proclivities. The way players relied on those attributes as a guide to character conduct was perhaps the most obvious way that players participated in the execution of the system of role-playing games: they had to know these statistics to abide by them. It was thus not an aspect of the system that the referee should or could shield from the player.

Efforts to steer decisions based on a character’s nature—or, as Simbalist would soon put it, to let the character play itself—are one of the clearer markers of a shift from traditional wargames to something new. Acting in accordance with character stood at odds with the long-standing wargaming approach to the role of a commander: wargamers traditionally did not revisit Napoleon’s historical battles to pose the question “What would Napoleon have done in this situation?” because Napoleon’s choices are a matter of public record. The wargamer instead would explore the question, as McIntosh put it, “How would the battle have gone if Napoleon had fought it my way?” (AE 18). Acting as a commander, the wargamer attempts to prevail in the strategic or tactical situation by exercising personal ingenuity. Some would approach D&D in this traditional wargaming fashion: when taking on the persona of an adventurer in a dungeon, a player could exercise personal ingenuity to succeed in the goal of defeating adversaries and gaining in power—a goal that the player and the character could usually be said to share. McIntosh explained that this is one of the reasons why “this identification is not a bad thing” because “it can get people to think and use their imagination to save a favored character.” But this philosophy competed in early D&D play with a more character-centric one: for some players, the goal shifted away from securing success toward faithfully portraying a rich character. Their behavior might seem baffling to wargamers who pursued success in adventure without a second thought to their characters’ natures. The degree to which a system might encourage one or the other of these approaches became the focal point of a heated dispute among designers and practitioners alike.

The community immediately recognized this division in its ranks and explored the consequences of these opposing philosophies. Game designs hinted at restricting character behavior based on properties like abilities or alignment—but above all else it was perhaps the goal of becoming more powerful that dictated how characters behaved. Design decisions heavily depended, unsurprisingly, on the degree to which players understood and participated in the execution of the system and the degree of control players had over how those key character statistics were determined. Access to this model of human nature ultimately enabled interested players to adopt characters defined by their shortcomings rather than by their triumphs, which became central to understanding what responsibilities and opportunities role playing created for players.

Self-Determination

When you generate a character, “your dice throws create Blanks, not people—same thing the Primordial Dungeonmaster did when he was messing around with the clay,” Dick Eney explained in 1975, at the dawn of the hobby (AE 5). “They are not Characters until you Characterize them with the Breath of Life.” Every Dungeons & Dragons character begins with throwing dice for abilities such as Intelligence or Wisdom, but it is up to the player to formulate statements of intention for that character, to decide what the character says and does, and through that process potentially to turn those characteristics into some semblance of a person. But how “blank” is a character, really, once the dice have been cast? Less than a year later, Nicolai Shapero wrote, “If I have a character with an intelligence of 6, and a wisdom of eight, I refuse to run him the same as an 18 intelligence 18 wisdom character” (AE 13). He noted that this could be a career-limiting decision: “This has cost me characters . . . it hurts, every now and then.” However, he insisted that “it is a far more honest way of playing.”

This question of the degree to which the system, as opposed to the personal ingenuity of the player, determines characters’ actions became a key early battleground in the struggle to define and explain role-playing games. In A&E 17 at the end of 1976, we find Richard J. Schwall setting the problem as follows. “Consider two extreme ways of playing D&D: the first is a puzzle-solving game where each player must always have his wits about him in order to find the treasure and to work his way out of potentially lethal situations. . . . The second way is the role-playing game, where each player endeavors to run his character with a personality consistent with his rolled abilities.” In the first, the player’s insight and experience decides the actions of characters; in the second, players attempt to constrain their characters’ actions to their likely ideas and capabilities. Crucially, only the second is what role playing meant to Schwall. We might suggest that in the first case the player strives for success in the game’s endeavors, but in the second the player instead strives to portray a faithful version of the character, warts and all. Schwall argued that these two styles are incommensurable and, indeed, concluded, “I have seen no evidence that there is a good middle ground between the two.”

Glenn Blacow rebutted Schwall by confirming that, “yes, there are two extremes of D&D playing,” but he flatly avowed that “they are not ‘skilled’ and ‘role-playing,’ however” (AE 19). He conceded that in some games, such as Kevin Slimak’s dungeon in Boston, “rolled intelligence must be ignored to survive,” and players must instead use their own knowledge and experience to select successful actions for their characters. But he took that as one pole and contrasted it with his own “other extreme,” which is “the role-playing game where no real danger exists; to wit, where characters either never get killed, or have a vast supply of wishes, etc., so that they can butcher cardboard monsters, bully the far less impressive non-player characters, and collect vast amounts of riches, magic, etc.” Ever advocating for balance, Blacow furthermore objected that “there are—contrary to your assertion—vast numbers of variations between the two extremes.”

In this 1976 dispute, we can discern slightly different early connotations of the newborn term role-playing game in use by the two sides. Blacow seemed to correlate the term with the overgenerous games, where self-indulgent characters face little prospect of death, as opposed to games of significant lethality, where players must think and act competitively to prevail. Schwall, in contrast, did not link role playing to generosity per se but instead to a game where, as he stated, a player will “run his character with a personality consistent with his rolled abilities.” Slimak drew a similar distinction at the beginning of 1977 in Wild Hunt 12, wherein there are “2 styles of playing a character: you can put a bit of yourself into the character,” which Slimak understood as a situation where “the player is the supplier of all the non-physical attributes” such as Intelligence, “or you can put yourself into his place, submerging yourself to the extent which you are able” into that role.5 By “put yourself in his place,” perhaps Slimak meant something similar to what Sandy Eisen wanted at the beginning of 1975, when through willing suspension of disbelief he “lived the part” of his character.

Designers and practitioners began to use the phrase in character very early on to describe this way of “submerging yourself,” where a player directs a character in accordance with the character’s nature rather than with the player’s better judgment. A passage in Bunnies & Burrows instructs players: “Once your rabbit acquires some traits, you should try to keep his behavior in character during future play, even when it is not in your best interest to do so! Believe it or not, this makes the game more fun in the long run.”6 The phrase in character had already entered D&D fan vocabulary the previous year: Sherna Burley praised another referee, “Bravo also on keeping Characters in character” in A&E 6. We can also find early uses of the term applied specifically to playing to Intelligence levels: in A&E 18, Bill Paley explained, “I kill characters of IQ 15+ for stupidity. On the other hand, characters who (in character) have 7− IQs often survive.” That is to say, as a referee, Paley exercised clemency when stupid characters acted stupidly but had none for high-intelligence characters played like dolts.

The notion that players should defer to some preordained personality model for their character when deciding actions predates D&D: people gamed as characters in Fight in the Skies or Western Gunfight, and we see the Western Gunfight and Midgard designers asking players to act “in character” before D&D hit the shelves. But those were based more on informal sketches of character, ones written into the background of a game rather than explicitly quantified in terms of intelligence or wisdom or other traits. To find detailed statistics for characters in wargames, we need to look to Tony Bath. His book Setting Up a Wargames Campaign specifies a character-quantification system that includes abilities such as Intelligence and Martial Aptitude as well as aspects of integrity such as Loyalty, all of which vary from 1 to 6. As the “Controller,” or referee, of his campaigns, Bath used these characteristics to determine the behavior of a vast network of nonplayer characters during the course of events. One can readily imagine how players could apply such quantified factors when deciding on their own characters’ actions, as apparently Charles Grant did.

Restrictions on how personalities may behave come through only in subtle ways in the D&D rules: for example, the original system stipulates that character attributes such as “Intelligence will also affect referees’ decisions as to whether or not certain action would be taken” and that “Wisdom rating will act much as does that for Intelligence.”7 As usual, we have only a vague guideline to parse, but the rules suggest that the referee holds a veto power over statements of intention grounded in the game’s statistics. This hint was not lost on early players. In 1975, Lee Gold included among the circumstances in which “a dungeonmaster should legitimately overrule a person’s call for his character” the situation where “the character’s proposed action is far too rash/dumb for his supposed wisdom/intelligence” (APL 520). This rule effectively directs the referee to police players’ statements of intention to gauge their appropriateness for characters’ Intelligence level—and by extension it implies that players should police themselves accordingly as they play their roles.

Yet with only such brief and opaque language in the original D&D rules, the impact of abilities on play became a matter of interpretation in local play groups: referees and players embraced the rules as guidelines, as Gygax instructed them to, and developed their own practices. Some players no doubt modeled their behavior on characters in fantasy literature such as Holger Carlsen in Poul Anderson’s Three Hearts and Three Lions (1961), which imagines a twentieth-century protagonist mysteriously transposed into the story of a fantastic hero, whom he then “plays” with twentieth-century shrewdness. These players injected into their characters all of the wits and reasoning they could muster. But others found a different sort of enjoyment in “submerging” themselves into characters native to the fantastic setting, true to the simulated characteristics that defined them in the system, for good or ill.

The sprawling discourse in game fanzines soon entertained, as a defining characteristic of role-playing games, this idea of acting within the quantified constraints of characters. Early in 1977, Kevin Slimak began teasing out this distinction by leveling the following accusation at his interlocutor Pieter Roos: “You, like many others, tend to discuss D&D as a role-playing game, rather than the more encompassing fantasy game” (WH 15). For Slimak, considering D&D as a role-playing game encouraged players to make poor decisions when, in their estimation, the characterization system required it: “I’ve just seen too many people doing stupid things BECAUSE of the character they’ve rolled and are role-playing.”

Roos responded by proposing a definition of a role-playing game with exactly that property: “I define a role-playing game as one in which the participants assume a character and act within that role” (WH 16). He underlined the word role to emphasize the constraints that acting within it would imply. Roos however, reinterpreted character abilities to avoid the excesses that frustrated Slimak. “I do not believe that the Intelligence and Wisdom characteristics are actually intelligence or wisdom,” he argued. “Magic Ability and Piety, or Talent and Godliness would be more appropriate” as names for the capacities of a character that D&D modeled with its Intelligence and Wisdom stats. Roos also downplayed the impact of role playing an idiot on the grounds of its associated risks: “If someone wishes to be a stupid character, fine. Like as not either the monsters or his own party will end the joke quickly enough.”

Not everyone would see the humor in playing to lose. Toward the end of 1977, in Lords of Chaos 3, Slimak further bemoaned this consequence of role playing: “I don’t object to role-players as much as I object to the use of role-playing as an excuse for not thinking, or worse, thinking of ways to do the wrong thing. Hells bells, some of the folks writing in this zine roll dice to see if their characters will say what they think up next . . . and that in some pretty dangerous situations.” Given the purported lethality of Slimak’s dungeon, one can readily appreciate his consternation on this point. But, more significantly, he implicitly lamented how players of D&D had begun to subvert the old maxim of wargaming that “anything can be attempted” in favor of a more constrained freedom of agency. Adherents to this new creed could direct characters to attempt things in the game or even say things in game only when they believed the character, as modeled by the system, would decide on those actions or utterances—even when that meant the player must defer to a die roll to direct the character, according to some interpretations of the game system.

By 1977, a few published designs explicitly encouraged players to accept just such constraints on their agency as part of the responsibility of role playing: most notably, Chivalry & Sorcery, published that year, recommends, “If a character is stupid, role-play and have him act stupidly. If he is a fumble-fingered boob who has the dexterity of a hobbled camel, have all the fun you can with him (these make good comic Thieves).”8 This restates Mike Carr’s promise that Fight in the Skies will be more “fun” if you play to the personality of pilot, even if it is not the optimal strategy. Chivalry & Sorcery summarizes this philosophy with the catchphrase “let the characters play themselves.” In extreme cases, its system even calls for wresting control over certain decisions from the player: for example, characters with low Wisdom “will be directed by random determinations whenever faced with difficult decisions,” a situation where the referee rolls the dice and the system will dictate how the character is played instead of the player.9

Text encouraging players to “let the characters play themselves” became practically boilerplate in role-playing games by the late 1970s. The Starships & Spacemen (1978) rules plead:

It cannot be overstressed that the player of the game should play a role as he determines his own character’s actions. That is, the player should form a concept in his own mind as to what type of person his character is and then act in accordance with these ideas. The character’s type should obviously be based on his abilities, as determined by the roll of the dice. Players should always try to act in character; while this may not prove advantageous in one particular set of circumstances, it leads to a much better game overall.10

But if the player must abide by the character’s nature even when that seems disadvantageous, then it matters how a character comes to be unwise or clumsy and how involved the player was in shaping that character. D&D followed Bath’s precedent by having players generate such characteristics randomly at the start of play. The original rules stipulate that “prior to the character selection by players it is necessary for the referee to roll three six-sided dice in order to rate each as to various abilities, and thus aid them in selecting a role.”11 The determination of the famous six characteristics of D&D—Strength, Intelligence, Wisdom, Constitution, Dexterity, and Charisma—thus was entirely outside the control of the player, who was not even intended to cast the fateful dice, as the rules assign that privilege to the referee.

It is perhaps some consolation that the player at least gets to see the results: the player selects a character’s class in consultation with the results of these rolls, so the player must be aware of roll outcomes. The choice of character class, effectively the career of a D&D adventurer, hinges on those dice.12 Characters incur significant penalties to experience-point accrual if their abilities are misaligned with their chosen class: a weakling Fighting-man might earn one-fifth less experience points than an average rival, whereas a burly Fighting-man will amass experience points 10 percent faster than normal. During character generation, characters can reallocate points into their class’s “prime requisite,” but only at a punishing two- or three-to-one exchange rate, and only from certain other characteristics, and even then only if it will not reduce those other characteristics below 9. This steers players to choose classes based on the dictates of the dice rather than playing to any preferences of their own. A player set on being a Cleric might, after witnessing a Wisdom roll, have a change of heart. After all, not everyone wants to play a “fumble-fingered boob,” at least not all the time.

Many early play reports suggest that referees soon permitted alternative character-generation systems with more generous average outcomes—and more significantly, systems that granted to players greater flexibility in the assignment of rolls to abilities. Already in A&E 6 in 1975, Sherna Burley complained, “I also prefer not to spend the time and effort I put into characterizing and making miniatures on someone unplayable,” and thus, instead of rolling three 6-sided dice per ability, she preferred a character-generation method in use in Los Angeles: “rolling for four and eliminating the worst one.” She considered it “a happy medium between shopping for unusually high characteristics, and being forced to play duds.”

Unsurprisingly, these variant systems invited controversy, especially when characters generated with these enhanced techniques tried to relocate to the game of a less-accommodating referee. Glenn Blacow singled out in A&E 9 in 1976 a certain campaign at MIT whose “characters were not allowed to transfer to Edwyr,” in large part because their characteristics were rolled “with 3d6 + 1d4.” Any transplanted character with conspicuously high abilities might face allegations of cheating. One commentator in A&E 18, Chris Pettus, complained, “What is most embarrassing is rolling up a character that you would think was cheated on if it was brought into your dungeon.” Pettus explained how one night he rolled a character with three characteristics at 18, including a Strength of 18/00, and no characteristics lower than 12. “I don’t blame people who look at me strangely when I present his character sheet. I would,” Pettus confessed, yet he insisted, “He is honest and was rolled in front of a witness, even.”

What counted as honest character generation necessarily depended on the discretion of the referee. In A&E 14, Wesley Ives explained how he implemented the following system in his local dungeon: “A player is allowed six rolls for the four Prime Requisites: he rolls six times, takes the best four and distributes them as he wishes between Strength, Intelligence, Wisdom and Dexterity.” Allowing the player to choose the placement of high rolls among the abilities effectively let players choose their character classes, a first step toward self-determination. Similarly, when the Quick Quincey Gazette surveyed its readership late in 1976 on the question “What system do you use for rolling up characters?” Peter Cerrato responded in issue 3 that in his own campaign he would “roll as many characteristics as are needed, and then the player puts them where he wants.” In the same issue, the Gazette related that Edi Birsan in New York would allow “the player to place each roll after it’s made in any of the six requisites that’s still open,” which granted the player significant leeway in assigning high values to desired abilities, but without the hindsight of knowing what the highest roll of the six would turn out to be.

Quick Quincey Gazette 3 also summarized responses on the subject of mulligans, or the cutoff for rerolling a character. It reported that “the most common method of deciding on whether you have a reroll is whether the total of the six requisites (before transfers) is below a certain figure.” A roll of three 6-sided dice will on average yield a result of 10.5, so any character with less than 63 total points of statistics is by definition below average. In Howard Mahler’s dungeon, 63 was the threshold for a reroll; Jim Servey permitted rerolls at the more generous figure of 65. Another documented approach from 1976 required that the character’s abilities total exactly 63. Richard Schwall, to prevent cases where a player “rolls a truly inferior character, below average in all abilities,” had his players “rolling only five of the abilities and choosing the last so that the total of all 6 is 63” (AE 14). If this would lead to a situation where the last ability “is less than 3 or greater than 18, then points must be transferred to or from one other ability so that all six abilities are between 3 and 18.” This typically leads to results where characters are “average in an overall sense but still vary widely in any one ability,” and it could lead to some very dramatic peaks and valleys in character competence.

The incentive to defy the will of the dice grew as new character classes placed onerous constraints on the characteristics required to play them. According to Greyhawk, for example, Fighting-men could become Paladins only if they had a Charisma score of 17 or higher. The Monk class defined in Blackmoor (1975) required a character to have a Wisdom and Dexterity of at least 15 as well as a Strength of 12. With limitations this severe, a player might only rarely have an opportunity to experiment with these classes: a score of 15 or higher on a given die roll occurs less than 10 percent of the time on average. We therefore should not be surprised to read in A&E 12 that Steve Perrin devised an alternative character-generation system to apply when you are “rolling for Monks and others who need characteristics of 15+.” Rather than rolling three 6-sided dice, Perrin rolled one 20-sided die to determine where the ability falls between 15 and 18: “1–10 is 15, 11–16 is 16, 17–19 is 17 and 20 is 18.”

Glenn Blacow practically sputtered in outrage when responding to Perrin: “This is remarkably depressing, and I hope I’m mistaking you, but it looks like you’re saying that instead of rolling 3d6 for characteristics and trying to get the character type from them, you simply decide, ‘Oh, I want a monk’—and roll d20 to see how good a monk you got” (AE 14). Blacow summed up his feelings about this variant with a single word entirely in upper case: “VERBOTEN.” Perrin confirmed that there was no misunderstanding: “There is no use in having exotic character types if it takes forever to get one. The idea, good people, is to have fun. Playing duds can have its own joys, but I prefer for them to balance out” (AE 16). He added icily that “no one has the right to say my way of playing with these ‘guidelines’ is verboten.” Referees and players here openly mutinied against the system, which D&D itself concedes should be no more binding than “guidelines,” and fixed the problem on the tabletop.

For purists such as Blacow, a method where players first select a character type and then apply alternative character-generation methods to secure any needed characteristics remained controversial in 1976. Most of the earliest role-playing games followed the character-generation precedent of D&D. Even Perrin’s own Runequest (1978) had each player roll three 6-sided dice for the game’s seven characteristics, albeit Runequest lacked classes altogether and did offer ways for characters to increase some characteristics in game. But certain in-born characteristics, such as intelligence and size, could not be augmented through any normal means in Runequest—players were stuck with the initial dice rolls.

Even before 1978 role-playing game systems had to acknowledge widespread practices for dealing with unwanted characters burdened with wretched attributes. Space Patrol casually mentions that “we have also known people who are fond of sitting down and generating some 20 or 30 characters and then throw out all but those who are unusually good. They then claim that all the characters in their stable were, in fact, randomly generated.”13 They recommend that referees require all generated characters be played at least once. Traveller (1977) similarly insists that “each player should use the character as it is created” even though “it is possible for a player to generate a character with seemingly unsatisfactory values.” Because some of the backgrounds in military-service careers, which characters in Traveller dice through before play begins, can readily prove lethal, for truly hopeless characters Traveller suggests “the low survival rate of the Scout Service may make it the best career choice.”14 In Tradition of Victory (1978), where players roll percentile dice for the six canonical character attributes—with Social Level standing in for Wisdom—the system is more lenient: it advises that “if a character is particularly unpromising, the player may roll again with the permission of the referee, but I urge referees to be fairly strict.” Tradition of Victory also proposes “another method might be for the player to roll three characters and choose one from among them.”15

Rather than condoning suicide or condemning players to the caprice of fortune, Advanced Dungeons & Dragons embraced the principle that its characters are exceptional. Players Handbook (1978) justified this decision by stipulating that “the premise of the game is that each player character is above average—at least in some respects—and has superior potential.”16 The Dungeon Masters Guide (1979) thus formally legitimized four alternative character-generation methods, including one that allows players to effectively roll twelve characters and choose the best one for play. It similarly allowed methods such as rolling four 6-sided dice and discarding the lowest roll, which augmented all of the characteristics. Without these measures, due to “quirks of the dice” players might be assigned “rather marginal characters” who “tend to have short life expectancy—which tends to discourage new players, as does having to make do with some character of a race and/or class which he or she really can’t or won’t identify with.”17 Where Chivalry & Sorcery saw the fun in playing a dimwitted character, AD&D optimized for making characters extraordinary so that players will more readily identify with them.

In both game design and play, the two cultures publicly grappled with the impact of abilities on playing roles throughout the late 1970s. Among “games people,” some welcomed the impartiality and fairness of the dice and accepted the requirement to simulate the random inequalities of people even when the dice delivered disappointment; others, however, sought to maximize their tactical advantage and found ways to ignore or circumvent any constraints imposed by abilities by exercising personal ingenuity. Among “story people,” some recognized the comic potential of playing a fool or klutz and relished the challenge of improvising to the abilities that dice generated—yet others identified more with extraordinary characters or wanted to assume the role of a particular, exotic character type and so availed themselves of alternate character-generation mechanisms. Designs could insist that players and referees abide by the dice, but that insistence had no enforceable consequences—players and obliging referees who wanted an extraordinary character could justify their practices no matter which culture they came from.

At stake in this choice was the nature of character generation as a creative act of the player. Consider the character-generation system of the era proposed by Bob Frager in A&E 34: a process of guided meditation. He encouraged players to close their eyes, breathe, and envision a door marked “Adventurers.” After opening the door and allowing a few adventurers to emerge, players should select an interesting one and speak to it, listening as it explains its history, its motivations, its strengths and weaknesses. “Now let yourself become the character. Identify with it and experience what it’s like to be it. Ask this person, how do you feel? What is the world like to you? What do you want?” Returning to reality with a specific person in mind, players then roll characteristics and “assign the six rolls in whatever way best fits their character.” At that point, though, rolling dice hardly seems adequate as a means of specifying the character.

A more radical design solution was to let players choose characteristics rather than leaving them to chance: to give players true self-determination. Such a system appeared as early as 1977 in the published design of Superhero ’44. It entirely eliminated die rolls for characteristics and instead granted characters a pool of “power points” to allocate as players saw fit. The rules postpone generating these abilities until after class selection, stating that “once a hero has created a background and selected a character type, he assigns prime requisite points.” Each player divides an initial sum of 140 points across the character’s seven requisites, though the choice of character class offsets certain requisites by a predetermined amount. Furthermore, “at the discretion of the referee, up to 50 bonus points may be added” for various circumstances, including “characters who accept weaknesses or disabilities (Kryptonite, for instance).”18 This design would be among the earliest to allow characters to take on some form of flaw during character generation in exchange for more initial purchasing power. But it ultimately created a level playing field for characters, where all superheroes have the opportunity to be equally super in place of the random inequalities yielded by dicing for attributes.

This design property proved to be attractive. When Steve Jackson included it in the first published components of his Fantasy Trip role-playing system, Melee (1977) and Wizard (1978), he explained about attributes, “Players don’t roll for these. Instead, each character starts out with 8 in each attribute, and the player gets another 8 to split between them as he likes. You can have an average fighter, a dexterous weakling, or a powerful clod. But you never start with a superman or a total oaf; all beginning fighters are equivalent” (SG 12).

The character-generation system of Bushido (1978) closely followed the point-buy precedent of Superhero ’44. “In Bushido, unlike most other role-playing games, the concept is for the Player to design a Character according to his ideal vision of a Profession. Thus, Attributes and other Abilities are custom-designed by the Player after he decides on a Profession.”19 Players distribute 60 points across the six attributes of Bushido, though the choice of class (Profession) offsets attributes by some positive or negative amount. Lee Gold, reviewing Bushido in A&E 41 , would call that “a fairly workable system” but warned that “it may result in a lot of carbon copies as the best mix because known for each type” of character. Randomness at least does guarantee a certain amount of variety in characterization.

Point-buy character-generation systems would inevitably inspire variants of D&D as well. Late in 1979, Andrew Gelman expressed in A&E 55 his dissatisfaction with the sanctioned alternative methods enumerated in Dungeon Masters Guide, assessing that although they gave “players better chances to roll higher characteristics,” they ultimately merely “had an effect like inflation—people just tried to roll better characters.” Gelman therefore proposed simply allocating 63 points to players for distribution across the six character abilities within the 3 to 18 range—he explicitly chose that figure to correspond with “the average sum if they were rolled with dice.” He noted that a 73-point system would deliver the same for Chivalry & Sorcery. From his implementation experience, Gelman insisted that “the characters do not become carbon copies. There are many possibilities of characters and there is no ultimate character type.”

In point-buy systems, the choice to play a “fumble-fingerered boob” rests entirely with the player. Those who find it fun can select a role with those properties, while others can avoid it, but every starting character has access to the same pool of talent, so none dominates others at the whim of the dice. Securing the consent of players to these characteristics rather than assigning them arbitrarily makes constraining characters to conduct appropriate to their statistics a matter of choice rather than imposition. It furthermore depends on players exercising creative control over the character-generation system—a clear departure from the original D&D guidelines, which stipulate players are not even allowed to roll the fatal dice to determine their own attributes.

Ethical Calculus

Constraints on what might be attempted arose naturally when players generated characters who were not as smart or wise as themselves and then diligently played to their dimwitted personalities. But another important early system element further restricted the statements of intentions given by players, forcing them to consider their character sheet before deciding on a course of action: alignment. The original Dungeons & Dragons rules are very clear that alignment is something players choose, one of the rare qualities of a character that players actively select without the help of the dice. Adherence to the constraints of alignment would be another practice linked to what it meant to remain “in character” and thus to role playing.

The “line up” of the Chainmail wargame originally sorted its combatants into the divisions Law, Chaos, and Neutrality in order to simulate the alliances depicted in fantasy fiction, so that elves might fight alongside dwarves but never orcs. D&D inherited these three categories and rebranded them as “alignment,” assigning a stance to fantastic creatures but granting humans the latitude to choose any personal ideology. As originally conceived, alignment was a statistic selected by the player during character creation. Alignment governed the compatibility of adventuring parties and the use of certain magic items that themselves held an alignment bias, such as intelligent swords. Players were largely free to adopt the alignment of their choice, though the D&D system did levy some restrictions: Clerics could choose only Law or Chaos, not Neutrality, and Elves, Dwarves, and Hobbits could not be Chaotic.

In its margins, however, the original D&D shows that alignment is not fixed at character creation; it warns that if a powerful Cleric “changes sides, all the benefits will be removed!”20 This might happen accidentally, when a Lawful Cleric dons a cursed “Helm of Chaos,” say, but the text elsewhere provides the germ for a system of morality: a Lawful Cleric must not take evil actions, such as misusing the “Finger of Death” spell, because that “will immediately turn him into an Anti-Cleric.”21 When selecting actions for a Lawful Cleric, a player must thus weigh the potential in-game consequences. D&D took this a step further in the Greyhawk expansion, which introduced the Paladin and Thief character classes, the former of which must be Lawful, and the latter must not be. Greyhawk warns that “any chaotic act will immediately revoke the status of paladin, and it can never be regained.”22 Paladins must also “give away all treasure that they win” to appropriate charitable endeavors, live a life of modest means, and associate only with Lawful characters. In recompense for adhering to this behavioral code, Paladins receive considerable in-game advantages—so players must wisely choose actions for their Paladins, under a constant threat of lapsing into banal Fighters.

These rules strongly imply that the referee must evaluate character actions to ascertain whether they violate the constraints of alignment. But, as usual, the original “guidelines” left much for the community to sort out. In the summer of 1975, in only the second issue of A&E—well before the term role playing had gained any currency—Joel Davis reported on the implementation of alignment in his Colorado group: “I wonder if most referees pay attention to Law/Chaos divisions—other than dangerous or safe use of a few magic items such as swords. Local custom here makes alignment quite important . . . and several referees keep track of each player’s law/chaos points. Progressing either way leads to special protections, occasional divine intervention, etc.” Davis here suggested that local referees at the time kept some kind of quantified tally of how Lawful or Chaotic a character’s actions had been and that certain game benefits or penalties could result from that score.23

Distinguishing Lawful acts from Chaotic ones was also a matter left largely to the discretion of groups. Lee Gold explained in A&E 9 that players in Los Angeles “tend to play that the Lawful’s aims are to rescue the unfortunate, kill Chaotics, and get loot. The Chaotic’s are to torture the unfortunate, kill Lawfuls and get loot. The Neutral’s are to get loot.” She added that “Lawfuls and Neutrals get along fine most of the time.” But the specification of these categories admitted of enough ambiguity that there were even disputes over whether sexual violence should be considered Lawful.24 Some reported to A&E on attempts to break the alignments down into components that would further clarify the expected behavior: Glenn Blacow, for example, divided the Chaotic alignment into three subcategories, Meanness, Sadism, and Dedication, each of which had an associated percentile score (AE 14). A character with low Meanness “tends to avoid trouble,” but one with high Meanness will never flee from a fight.

However loosely or strictly referees and players conceive of alignment, characters have some in-game obligation to adhere to the ethos chosen by their players: when crafting a statement of intention, the alignment of the character is something to take into account. But acting against alignment is not simply an error in role playing, like acting smarter than one’s rolled characteristic for intelligence, say, which a referee might veto. Characters implicitly possess free will in the game world and thus have the capacity to make ethical choices that contradict their stated affiliation and have consequences in the game. Potential consequences for transgressions against alignment render the decisions a player makes for a character meaningful. A referee observing lapses in character can take a number of actions, including adjusting alignment accordingly. As Glenn Blacow recommended in A&E 10, “If people persist in acting Chaotically, change their alignment, dammit!”

In that crucial sense, alignment in practice might not be something chosen by the player but instead a judgment of character behavior made by the referee. This was true of a variable quantified alignment system in use in New York in early 1976 recorded by Scott Rosenberg in the first issue of his fanzine, fittingly titled the Cosmic Balance. Rosenberg admitted twenty degrees of alignment: “Alignment on the general framework is measured from 10 to −10: 4 to 10 being varying degrees of Law, −3 to 3 of Neutral, and from −4 to −10 of Chaoticality,” as he disarmingly put it. Rosenberg insisted that a character’s ethical behavior determined alignment and that cynical schemes to elevate alignment superficially, such as “paying off the monastery,” would not inflate the number. “Your alignment is determined completely by your actions, interpreted by the GM.” Not only did the referee score players’ alignments to mirror their actions, but the exact tally also remained a secret. “You will only know whether you are Lawful, Neutral, or Chaotic—never a specific number.” In Rosenberg’s system, players were not privy to the execution of the morality system; it was conducted exclusively by the referee, who could surprise the players when their mounting transgressions—or virtuousness—relocated them on the alignment spectrum. Later in 1976, Howard Mahler from nearby Princeton documented his sprawling system wherein referees track quantified character alignment between 100 and −100; although he kept that precise value a secret, at least he advised that “if they start getting towards the boundary line a warning should be given to the players” before their alignment changes (QQG 1).

The uncertainty of these alignment-tracking systems made players internalize an ethical inner voice. By isolating players from their quantified score, it brought them ever closer to their characters. Mahler stressed the impact this had on play and in particular how it might require players to formulate actions for their characters with care. “Remember,” Mahler says, “that very lawful characters cannot just go around playing the game like a game of chess. Sometimes, the best strategy from a game standpoint does not coincide with the lawful thing to do” (QQG 1). And in 1977, Kevin Slimak explicitly linked abiding by alignment to the concept of remaining “in character” in a way that he would not for the similar question of acting within abilities such as Intelligence: “In fact, I do insist that people stay in character. The difference is in the definition of character. I demand of the Lawfuls that they remain Lawful, especially the Clerics” (AE 21). Slimak added constraints for the nonreligious as well—for example, “FM giving oaths, especially on their weapons, are expected to keep them.” Even for a games person such as Slimak, doing the “lawful thing” and remaining in character trumped the wargaming impulse to prevail in the tactical situation.

Dividing the alignment system into a continuum with many incremental steps only clarified so much about what it meant for an action—or a character—to be Lawful or Chaotic. In February 1976, Gary Gygax unveiled in Strategic Review a famous revision of the alignment system, which added a continuum for good and evil and thus raised the number of alignments to nine, all displayed in a two-axis graph between “Lawful Good” and “Chaotic Evil.” He began that article by acknowledging the opacity of the original alignment rules and the understandable confusion expressed by the community: “Many questions continue to arise regarding what constitutes a ‘lawful’ act, what sort of behavior is ‘chaotic,’ what constituted an ‘evil’ deed, and how certain behavior is ‘good’” (SR 2 (1)). He even supplied a helpful list of a few dozen qualities to associate with the four endpoints of his axes: Law claims terms such as “principled” and “uniform,” whereas Chaos will be “unrestrained” or “disordered”; a Good character will be “honest” and “beneficial,” in contrast to the “injurious” or “corrupt” Evil character.

What is less commonly remembered about this revised “fourfold way” alignment system is that Gygax recommended that referees literally pinpoint the location of each character on the graph and revise these positions according to how well in-character behavior corresponds to the listed adjectives. A character’s alignment thus is not absolute and atomic; each character exhibits a graphable degree of Lawfulness, Goodness, and so on. “The actions of each game week will then be taken into account when determining the current position of each character,” Gygax continued, stressing that referees consider the intensity of actions to weigh how drastic a correction certain behavior might require. This essay revised the previous guidance for Paladins: rather than a single Unlawful action irrevocably erasing Paladinhood, Gygax now graphed an “area in which a paladin may move without loss of his status,” in the Lawful and Good corner of the chart.

This two-axis version of alignment was soon equipped with its own quantification. Along with its “Installment L” in the spring of 1977, the Judges Guild shipped a “Character Checklist” that transformed alignment into a pair of statistics locating a character from +20 to −20 on a scale of “Law/Chaos” and +30 to −30 on a scale of “Good/Evil.” It stipulated, for example, that “each act adjudged to be either Lawful or Chaotic will move the character up or down 1 in that column,” so that “players not meeting alignment expectations can be charted and then penalized accordingly.” Systems with these properties soon captured the imagination of the community: Lane Whittaker provided in A&E 33 his own hack for the Judges Guild quantifications, along with concrete guidance on properly playing the ethically challenging Paladin, Cleric, and Thief classes.

Quantification systems for alignment became commonplace in fantasy role playing by the end of the decade, though in some cases the quantities were fixed rather than variable. We can find fixed alignment values in a system such as Bifrost (1977), which rates alignment in two dimensions: a numeric axis ranging from Good (1) to Evil (7) and an alphabetic axis ranging from Law (A) to Chaos (G). Players in Bifrost select their alignment, but the system acknowledges that a player can then “decide to take action contrary to that alignment.”25 Bifrost leaves it to the referee’s discretion what punishment if any should apply in those circumstances.

For those more interested in the tactical situation than in exploring fictional personalities, alignment serves merely as a challenging impediment, a handicap that a referee has to enforce. D&D uses the threat of a dire reduction in power to enforce proper behavior in Paladins, but some systems offer a trade-off instead of a direct reduction in ability. Buccaneer (1979) has only a rudimentary alignment system, but it encourages characters to conform to the system through incentives. “Buccaneers have a ‘public’ image to foster,” which may be either a good or bad one, and both paths confer advantages and disadvantages. A “bad” buccaneer receives a combat bonus, but a converse penalty to the roll made to determine the severity of punishment when captured; a “good” buccaneer receives a commensurately greater chance of being pardoned if captured but trades this for less force in combat. Buccaneer enumerates a set of good and bad actions, such as aiding wounded enemies versus torturing them; “to reach and maintain these reputations, each player must consistently perform certain positive or negative acts on the personages he captures.” If the buccaneer’s behavior is erratic and inconsistent, then “he is neutral with no die roll modifiers,” though no way of tabulating behavior is recommended by the rules.26

Some quantified, fixed-alignment systems require players to roll for alignment like other attributes during character generation—and then to role play the consequences as a constraint. Chivalry & Sorcery is a notable example: players roll a 20-sided die, which yields a common result such as “Worldly” or a rare extreme such as “Saintly” or “Diabolic.” In directing character actions, players are expected to defer to the system and accept this randomly generated alignment because it is fun, if not indispensable to the idea of role playing as Chivalry & Sorcery understood it. Whereas D&D bribed Paladins with special powers in exchange for obedience to a moral code, Chivalry & Sorcery contained language such as “alignment is merely a guide to players so that they can build their character’s personality.”27 As another path to “letting the characters play themselves,” voluntary submission to ethical tenets became one of the starkest points of demarcation between adherents to the new genre of role-playing games and their predecessors in the wargaming community.

The Arduin Grimoire (1977) strikes a similar posture, providing a chart that let players roll percentile dice to determine one of its thirteen character alignments and then giving explicit cues for the resulting character behavior. This could assist referees in deciding the actions of nonplayer characters, but it has equal applicability to played ones. On average, one-quarter of all characters will turn out “Moderately Lawful,” the most common outcome; only one percent will become “Amoral Evil.” Each alignment is assigned a specific score in five behavioral circumstances: “Kill Factor,” “Lie Factor,” “Tolerance Factor,” “Loyalty Factor,” and “Cruelty Factor.” A “Moderately Lawful” character has only a 5 percent chance of lying or engaging in cruelty and should remain loyal 85 percent of the time. By contrast, a “True Chaotic” character flips a coin for most behavioral decisions but has only a 35 percent chance of remaining loyal. A brief set of “General Notes” gives further tips beyond the five categories on role playing each alignment: for “Lawful Evil,” the text reads, “Fanatical, bigoted, arrogant, nasty.”28

Those sorts of cues for character behavior begin to stretch our understanding of the scope of alignment. Indeed, in prescribing actions associated with both alignment and abilities, early role-playing games often strayed into the more nebulous region of specifying personality. Mark Swanson proposed in 1975 a set of randomly generated minor abilities that would help to differentiate starting characters, which was immortalized in A&E 1 as the “Swanson Abilities.” Over the next year, several authors refined his proposal: in A&E 4, for example, Jack Harness presented an assortment of character tweaks, which included various potential skill bonuses for combat, perception, defense, and so on, but then also dices for “idiosyncrasies” that give a sense of the character’s personality. These idiosyncracies range from being unflappable to excitable, paranoid to overconfident, and merciful to bloodthirsty. Beyond differentiating characters, they give explicit cues to the player on how to direct character behavior. By the spring of 1976, Grant Louis-d’Or had expanded this into a sprawling “Set of Special Characteristics” in A&E 11, a table spanning six pages and 144 different randomly generated outcomes. A character might be “moody, temperamental, easily intimidated,” which would induce that character to retreat from a fight at at half hit points; or a character might be a “dynamic talker but lazy; ambitious; unromantic,” which bestows a bonus against Charm spells but a penalty to combat; or a character might be someone who “thinks everyone else is insane” or who “has a buried subconscious personality.” Louis-d’Or’s personality traits illustrate how, even before the widespread acceptance of the term role playing, systems extending the baseline alignment design of D&D could challenge players to adapt to some behavioral constraint imposed by a die roll during character generation.

Any aspect of personal disposition could be diced and quantified in this manner. Greg Costikyan had already invented a “Sex in Dungeons & Dragons” system by 1975.29 An initial “affiliation” roll would assign characters to the categories of heterosexual, bisexual, homosexual, transsexual, or “extraordinary,” where the latter encompasses a variety of fetishes. A “fixation” table gives new characters a 65 percent chance of being “normal” and a 2 percent chance of being obsessed with armpits. Three 6-sided dice determine a character’s “sex drive,” which must periodically be satisfied to avoid desperate acts; Costikyan noted the implications of this determination for alignment in that “any sadist or sado-masochist with a sex drive of above 16 must be chaotic” and that Paladins may not have a sex drive higher than 14. Die rolls determine the likelihood that a nonplayer character will welcome advances, whereas “players may, of course, fuck among themselves without checking sex drive/charisma, so long as they are of the appropriate sexual affiliation/sex.” Scott Rosenberg, who played in Costikyan’s game, observed “Once, I was a clerical phase-spider, homosexual with an oral fixation, masochistic (the sex characteristics are determined randomly, not by choice)—and soon dead” (AE 12).

With the proliferation of behaviors a system could dictate, some players wondered why it was necessary for a system to cover concepts such as alignment at all. “Why not just do away with alignments and have the players decide what their characters are like by writing a paragraph or two about the character on their character sheets?” Michael Troutman wondered (AE 51). Runequest abolished alignment in favor of setting-based factional affiliations, though some did come with certain constraints or responsibilities. But others found value in an alignment system, including Jon Pickens, who would counterpropose augmenting alignment with additional factors to create a situation “of the player balancing conflicting impulses to produce the action most consistent with the character he is trying to create” (AE 52). Promoting alignment as a source of tension in playing a role, where ethical choices may compete with other incentives, began to expose ways that these games could tap into more fundamental questions about human nature than the simple “line-up” sorting fantastic combatants into sides.

Trisha La Pointe, writing in 1978, reported on her experiences playing “one of the ‘unacceptable’ types of characters,” a greedy, Lawful Evil character who “will cheat the unsuspecting of their due and will double deal to her advantage,” but who nonetheless “will associate willingly in an adventure with characters of other motivations” (AE 32). She found her character ostracized by “good” parties, and she ultimately concluded there were deeper matters at stake: “There is more to playing D&D than the game itself. There is the social nature of the game which brings together a divergent lot of people for, it is to be hoped, an evening of fun and challenge. There is also the meta-game where the players own values drift into their characters’ values.” In what would become a common sentiment regarding the “meta-game” of interpersonal dynamics that supervenes on role-playing games, La Point rueful noted that “dealing with this is one beast of a challenge, and not much fun at all.”

Thus, alignment and the variant systems extending it provided another vector in which playing “in character” could override the traditional goals of wargaming. Fidelity to alignment, like constraining actions based on abilities, was not always a winning move, but some saw it as an obligation built into the system of D&D. Even where a rule set has no element called “alignment,” ethical calculus can still exert leverage and steer character behavior. Bushido departs from a number of Eurocentric fantasy-gaming conventions by adopting a mythical Japanese setting. Instead of having alignments the players can choose, Bushido instead assigns to players a value for on, or “face,” which the designers associate with honor and reputation. Newly created characters receive an initial on value based on their family’s circumstances and accumulate more through honorable actions. On might accrue from behaving properly in Japanese society: by exhibiting courage and skill in combat, by solving mysteries, by performing various courtly activities, or just by having “outstanding style,” according to the referee’s discretion. More so than classic alignment systems, on demands that players constrain their actions to conform to the Bushido code as the game systematized it. And Bushido would light upon a key way to incentivize players to conform to on: by linking it to the game’s experience system.

Personal Goals

If playing “in character” gets in the way of some ultimate objective, what is it? A traditional board wargame ships with achievable victory conditions for the players, but during the play of Dungeons & Dragons fidelity to the characterization of abilities and alignment chafes against something else. A D&D game is, to borrow the words of the Western Gunfight authors, “an informal ‘campaign,’ open-ended,” without some preordained goal. So if you are not playing to win, what is a character’s purpose in a game like D&D? What’s the point of the whole game, anyway?

D&D does not specify any ultimate objective of its play—unlike its close imitator Tunnels & Trolls, which would state quite bluntly in its 1975 edition that “the true object of this game is to accumulate as many experience points as possible and by this means advance your first level character into as much of a superperson as you can.”30 Gygax in 1976 would say much the same of a D&D campaign: “progression, rather than winning per se, is the object” (SFF 87). If the life of a D&D adventurer may be compared to a game of pinball, then the experience point total for each character is the glowing score in the back box; players sneak glances at its steady rise throughout the game.31 The progression system of D&D implies a character arc: characters begin as inexperienced, weak, and undifferentiated yet will over time grow in power, gain confidence, and develop a personal history, if not a legend.

The earliest commentary on the play of D&D celebrated the novel way characters persisted over game sessions and advanced with success. Reflecting on the key moment when he first played in Arneson’s Blackmoor, Gygax remarked on how he found “the idea of measured progression” that it had introduced to be “very desirable” (DR 7). Progression was measured and quantified in the form of experience points accumulated by characters, which would in turn allow them to advance in level. Mike Wood, who observed a Minnesota group in early 1974, attested that he “was intrigued by the way the results of one game could be carried over to future games: a warrior could advance in rank by virtue of number of orcs killed, etc.” (MN 39). Players participated in the exercise of the progression system so prominently that it was obvious not just to players like Gygax, but to a spectator such as Wood.

Wood here identified two crucial dimensions of progression: how it allows successful characters to advance in power; and how this occurs episodically, spanning game sessions. D&D signaled its episodic nature with the word campaign in its subtitle: Rules for Fantastic Medieval Wargames Campaigns. H. G. Wells documented wargame campaigns in Little Wars (1913), where he explained that a campaign is a series of wargame battles, as a “rubber is to a game of whist” (that is, a best-of-three series).32 A wargame battle might involve only one evening of play, but ambitious gamers who simulate a protracted war could stage many evenings of related battles as opposing forces in various locations around the campaign map discover one another, fight battles, and regroup afterward. D&D explains in its foreword, “While it is possible to play a single game, unrelated to any other game events past or future, it is the campaign for which these rules are designed.”33

The episodic nature of D&D reverberates throughout the rules, which divide a campaign into a series of “adventures,” illustrated in the text with constructions such as a “campaign of adventures.” The scope of a single adventure is only indirectly given in the rules, as during a dungeon delve “the adventure will continue in this manner until the party leaves the dungeons or are killed therein.”34 Any such bounded adventure might have its own purpose: a referee might design a dungeon where players have some direct objective for a given descent. D&D furthermore scoped the accumulation of experience to the adventure explicitly, as “it is also recommended that no more experience points be awarded for any single adventure than will suffice to move the character upward one level.”35 Progression is thus restricted during the course of a single adventure: only through surviving repeated episodes in a campaign can a character rise to the heights of power—it keeps players coming back for more. As for when exactly the advancement of level takes place, Kevin Slimak observed in July 1975 that “there’s some question as to whether you go up during the expedition down in the dungeon, or when you come up” and leave the underworld (AW 2 (12)).

D&D grants referees almost total discretion in determining how to award experience, but it softly recommends that referees grant experience for slaying monsters and accumulating treasure. Empire of the Petal Throne recasts this as a restriction, making it clear that those are the “only two ways” to obtain experience points and that “no points are granted for casting spells or other types of activity.”36 Such a policy would ostensibly steer the characters into a life of aggression and acquisitiveness, and surely many played D&D and other games in its tradition that way from the start.

But by 1977 the experience-point system of D&D and the implications of “becoming as much of a superperson as you can” for role playing had become a subject of increasing controversy and consternation. Peter Cerrato, writing in the Wild Hunt 22 in November that year, noted the growing interest in competing titles and hypothesized, “One of the main reasons for the reduced interest in D&D is the fact that it is becoming a zero-sum game, with the person who reaches the highest level the fastest ‘winning.’” He believed, “This is in direct opposition to the concept of ‘role playing’” and that successful D&D games were those where “the GM encourages the players to ‘characterize’ their characters.” The very design of the original game weighed against this approach, however: “The direct cause of the lack of role-playing is how the D&D level system is set up. The whole game becomes a hunt for ‘experience points.’” The result is that “the players themselves are connected too much with the mechanics of the game instead of with their characters.” Cerrato did have a remedy in mind that would sound familiar to Sandy Eisen: he proposed that the referee unplug the scoreboard and be “the only person who knows a character’s experience points, his level, and even his hit points.” This solution echoed similar sentiments being expressed about quantified alignment and other system properties. Cerrato maintained that this would lead to a situation where “the players can pay more attention to their characters instead of to game mechanics.”

Convincing players to kick the progression habit would prove a tough sell, however. Many instead experimented with systems that restructured the relationship of progression to the goals of the game. These systems included realigning the award of experience points to incentivize role playing, abolishing experience points and levels in favor of other measures of progression, replacing unidirectional progression systems with fluctuating ones, and finally substituting another explicit objective for the baseline goal of accumulating power.

First, many published and reported practices for awarding experience points broadened the set of game events that yielded rewards and more finely tuned the purpose of awarding experience. When Howard Mahler surveyed the readership of his Quick Quincey Gazette about experience points before its third issue in 1976, the community immediately demonstrated that referees bestowed experience points for all sort of activities. Jim Cooper reported, “I favor giving experience for everything. Say maybe 1 or 2 points for just riding your horse.” Peter Cerrato explained that “experience will be received for just that, experiences. Even if you come out of the dungeon without any treasure, just going down will give you experience points.” Cecil Nurse one issue later added “throwing spells” to his own list and gave experience points “for each day in the dungeon or wilderness.” Experience was a commodity these referees disbursed with largess, constantly rewarding players with small, reassuring amounts of it.

Other contemporary play reports show a similar diversity of character actions that warrant an experience award and give some inkling of the ways referees could use such award to steer characters. Early in 1977, Jim Thomas submitted to A&E 20 seven categories of behavior that he as a referee encouraged with experience, including “using a new professional skill successfully,” as, for example, “a Magic-user gets points the first time he detects evil with a detect-evil spell,” for “helping others,” which Thomas deemed “particularly important for good neutrals,” for Clerics successful in “converting the heathen,” and even for simple “survival.”

Others would more subtly groom experience rewards to encourage helpful conduct. Although Robert Clifford reported in Wild Hunt 12 that he tallied “group experience” for the entire party rather than tabulating individual awards, he did personalize each allotment: “During the adventure, each character is assigned positive or negative percentage points depending on how much of an aid (or a hindrance) that individual was to the party and on how well they acted in character.” Tying rewards to acting in character used experience specifically to encourage role playing. This modifier, Clifford noted, “is also very handy to dock the turkey or pseudo-lawful characters who are always endangering the party or causing mayhem.” Methods for coping with a “turkey,” a problem player in the eyes of a group, form a major branch of the literature of the day.

Among published designs, Tunnels & Trolls had in 1975 already broadened the circumstances in which referees should award experience. It instructs referees to grant experience for “using magic,” regardless of whether the spells contribute to the death of an adversary.37 Making a saving throw also grants experience. Even exhibiting “daring” by going deep into the dungeon merits such a reward. Tunnels & Trolls was only the first of many commercial role-playing systems to tweak the incentive of progression, and others would transform it in more radical ways. Chivalry & Sorcery remained fairly close to the earliest precedents for progression through experience points and levels, insisting, “There must be some system by which the success of a character can be translated into an improvement of his abilities.”38 But because Chivalry & Sorcery aimed for a deeper level of fantastic medieval simulation than competing titles, it scaled experience awards based on how well character actions correspond to the source literature of fantasy. Dwarves, who have a documented hunger for treasure in the fantasy genre, gain one-to-one experience points for acquiring gold, but magicians, who presumably have less passion for wealth, get only a tenth as much experience for it—a “Magick User” derives more experience from learning new spells, enchanting items, raising demons, or even simply meditating. Clerics receive experience for “doing good works,” while Thieves accrue it for successful sneak attacks or picked pockets, and Knights advance through chivalric victory and gallantry.

Awarding experience points to players for adhering to characters’ expected behavior in the setting was a key step in the development of role-playing games. The Bushido rulebook notes well that “in strict gaming terms, the object of any role-playing game is to gain experience,” but in Bushido “the Level of a Character is determined by his Experience score and his on score.”39 As the surrogate for alignment in the Bushido system, on dictates the proper behavior of characters in its mythical Japanese society, so Bushido effectively links progression to adherence to that role. A character must meet a target sum in both experience and in on to advance. Certain in-character actions devastatingly affect on: “breaking one’s word reduces one’s on score by 25%.”40

Other designs followed the lead of Bushido by rewiring experience points into something that encourages remaining in character. Heroes (1979) closely follows the number-and-letter quantification of alignment in its fellow British title Bifrost with a system that measures Good and Evil on one axis with a point value from 1 to 5 and rates Law and Chaos on a second axis as a letter range between A and E, such that an entrenched Chaotic Good character would have an alignment of 1E. “Once chosen,” Heroes instructs us, “a character should attempt to act in character with his alignment throughout the game.” To ensure compliance, Heroes equips referees with a crucial tool: “Actions in line with a character’s alignment should be rewarded Personal Experience Points (PEP) by the umpire, but actions out of character should be met with a loss of PEP.”41 In games where the ostensible objective is to accrue experience, nothing motivates players quite like systems that will increase or diminish their progression. If accumulating sufficient experience to become a “superperson” is indeed the closest thing to victory, systems like Heroes and Bushido show how progression can be coupled to acting in character: role playing itself becomes the object of the game.

Second, early designs began to reject the pinball-style accumulation of high scores in favor of less-quantified ways to progress. Among self-identified role-playing games, Bunnies & Burrows pioneered a progression system that eliminated experience points and levels entirely. It instead implemented the concept of “experience rolls.” A character earns an experience roll by succeeding at the use of a particular skill while under duress; the game gives an example of a rabbit fighting off a badger as justification for awarding an experience roll for Strength. An experience roll does not guarantee progression: the probability of advancement depends on the innate value of the ability as determined during character generation. A rabbit with an innate Strength of 8 has a 10 percent chance of advancing, whereas a rabbit who originally had a strength of 17 has a 30 percent chance. If the target is rolled, then the “level” of that characteristic is increased by one.42 Players can still seek to become a “superperson,” or superrabbit as the case may be, but the process comes across as more organic. By eliminating the visible scoreboard of experience points, these progression roll systems liberate characters from the implicit goal of reaching the high score and create a space for other potential campaign goals.

Moreover, experience rolls in Bunnies & Burrows do not result only from fights: a character earns them as well for notable success in rolls involving Wisdom or Intelligence or Charisma. The designers of the game recognized how progression could entice players into role playing rather than into aggression and acquisitiveness. The game places a particular emphasis on in-character story-telling, where rabbits relate “the fabulous adventures of some heroic rabbit of the distant past.” As “players should be highly encouraged to invent and tell such stories,” so “the gamemaster can encourage such stories by giving experience die rolls to the rabbit of a player that tells a good tale.”43 Experience rolls could can function just as effectively as experience points in steering players into desired activities.

No small part of the effectiveness of progression derives from dispensing it in measured doses, especially when the improvement will not come into play until the next game session within a campaign. The Bunnies & Burrows rules stress, “It is best for the Gamemaster to actually save these die rolls and make them at the end of each adventure, rather than allowing rabbits to advance during the same adventure, thus slowing down the flow of the game.”44 Players can then look forward to exercising these augmented abilities, an incentive to return for the next game session. Gains should be rare and precious enough to feel like a substantial reward to players, and thus episodic language survives in Bunnies & Burrows: “Not more than one Level in any given category should be awarded per Game Day.”

Versions of the experience-roll mechanic appear in many subsequent role-playing games. Space Patrol has no experience levels, but consistently accurate use of guns awards to players a roll that, if successful, will grant a permanent combat bonus with the weapon in question. Famously, Runequest also dispenses with experience levels in favor of skill increases: “To learn a skill by experience, a character must use it successfully in conditions of stress. The player may then try to make a roll of (100 – current ability with skill) or less on d100. If he makes it the character gains up to 5% in ability in that skill.”45

Not all skills could be improved through use: in Runequest, characters must increase some basic abilities through spending in-game time and money. These mechanisms subtly rechannel progression into a targeted self-improvement decision made for the character. Training is the most obvious way to translate money into a more lasting form of progression: games such as Superhero ’44, Traveller, and Runequest have characters invest in education and exercise to raise their skills. Traveller, which takes a long-term, even tedious view of the character arc, recommends that characters enter a four-year degree program costing tens of thousands of credits in order to learn a new skill. In Buccaneer, a character in the regular navy can gamble £200 “for a chance to roll a six-sided die to advance himself by study and perseverance,” raising one ability, but “a roll of 1–3 accomplishes this, and a roll of 4–6 indicates the effort was wasted.”46

Even games that completely disavowed the principle that adventuring itself would somehow intrinsically improve characters still permitted progression through the accumulation of material advantage. Science-fiction role-playing games of the 1970s seem to have fallen into a pattern of designs of this form. For example, Metamorphosis Alpha simply has no concept of either experience or training. A character’s core statistics are fixed as originally rolled—apart from in-game events that might induce new mutations in the character, which can be either benign or malignant. The early science-fiction game Starfaring (1976) similarly lacks any progression system. The only advancements in Metamorphosis Alpha and Starfaring derive from money and equipment: acquiring better guns, armor, or technological artifacts through either plunder or purchase. Commerce usually happens only outside of an adventure, back in a town—or, as the case may be, in a spaceport—hence reinforcing the notion that progression happens between adventures and thus within the context of a campaign. But wealth and possessions are not intrinsic to a character like experience levels are; they can be lost, stolen, or otherwise surrendered in the course of a story.

Third, gamers began to reimagine progression as something impermanent, even reversible.47 The notion of losing levels does appear in the margins of the 1974 edition of D&D, notably as the result of attacks from certain level-draining undead creatures, but in the course of play early adopters devised a more fitting cause for experience loss: death. Although a character’s demise might result in the total annihilation of all progression, in practice resurrections and reincarnations abounded in these games and had little practical consequence, albeit the Greyhawk rules placed a limit on the number of times any single character could return from the grave. But Lewis Pulsipher reported in 1976 that in his games “when someone dies he loses 6–10% experience,” a penalty that could set a character back a level under the right circumstances (AE 13). Lee Gold, for her part, affirmed in response that “all my D&D life I have played that a revived/reincarnated character loses 0–99% experience.” Around these tabletops, progression was not a one-way street, and experience points were not an irrevocable asset inscribed on a leaderboard that characters could accumulate without check, even in the face of death.

Earlier games had already transformed experience points and levels into almost unrecognizable forms. En Garde, for example, replaced experience points with status points and levels with social levels. The system is quite explicit that “the player’s object is to accumulate as many status points as he can . . . in order to raise his social level.”48 At a given social level, a player requires a target number of status points to advance: to rise from social level 4 to 5, 15 points are required. Progression is episodic in that status points reset to zero after every game month, and characters advance in social level only if they meet the necessary quota at the end of that interval—but no amount of excess points permits characters to gain more than one social level in a month.

Liberated from both the fantasy setting and the original connection of “experience” to combat, the activities in En Garde that accrue status correspond to the expected activities of characters in the libertine Renaissance setting it simulates. Characters can gain status by joining social clubs, by carousing and gambling, or simply by holding a military rank or noble title. But social level can decline as well as advance. A minimum income of status points per month is required to maintain a social level; characters who fail to meet it will fall one social level. Any wager gambles with status as well as with cash: a winning bet grants a status point, but losing subtracts one. Although duels can grant status points, “any character who has cause and an opportunity to fight and does not will lose status points equal in number to half his current social level.”49 Those in a military profession gain status for warranting a mention in the dispatches but lose status for public cowardice. For En Garde, level is a fluid commodity; moreover, the system makes progression a consequence of something other than a character’s aptitude for aggression and acquisitiveness: it is instead a reflection of how closely a character’s behavior conforms to its literary antecedent. Tradition of Victory similarly has no experience point system but focuses on advancement in naval rank through a social system of notice and promotion reminiscent of En Garde.

Even in a heroic fantasy setting, why should campaigns explore only character arcs about improvement over time? After certain types of experiences, surely an adventurer could get worse rather than better. As early as the fall of 1975, Matthew Diller circulated rules for “Psychotics in Dungeons & Dragons” through the early Diplomacy zine the Pocket Armenian.50 Diller proposed the addition of a “Sanity” ability to D&D, rolled with the other basic attributes during character generation, which fuels a sort of nightmare version of experience levels. Characters with a less than average Sanity score will have between a 10 and 25 percent chance of “turning psychotic” when confronted with certain traumatic conditions as determined by the referee: examples include long periods of confinement, being charmed or polymorphed or petrified, or especially being raised from the dead—the latter trauma adds 30 percentage points to the chance of a psychotic break.

Repeated episodes will raise a character’s Psychotic Level and lead to recurring mental illnesses that surface under milder stress situations, such as routine D&D combat, manifestations of which include hallucinations, paranoia, catatonia, and other incapacitating effects. As characters advance in Psychotic Level, they accrue more illnesses and a greater chance of triggering them in response to stress. Upon reaching the tenth Psychotic Level, the character becomes permanently insane and liable to involuntary commitment to a monastery for treatment. The rules do allow Clerics above the fifth level to study “therapy,” which can aid them in rehabilitating psychotic characters by both assuaging symptoms of an episode and, through protracted therapeutic sessions, reducing the character’s Psychotic Level.

A character arc where adventurers become disabled by long exposure to trauma and stress, rather than relentlessly improving, offered a welcome antidote to the addictive self-aggrandizement of the D&D progression system.51 Yet the most interesting implications of Diller’s approach lay in the sort of role play it encouraged. In some outcomes, a character might become a hypochondriac who must claim to suffer from imaginary conditions or invent a wish-fulfillment alter ego or suddenly lose interest in acquiring material possessions in a fit of neurotic selflessness. Some immediately render a character’s alignment Chaotic. It challenges a presumably sane player to try out new experiences, even when they had the potential to be uncomfortable or even frightening. It implicitly sets characters the personal goal of keeping it together.

Fourth, and most fundamentally, games began to search for goals unrelated to progression. Why should the advancement of a campaign depend on characters getting either better or worse? Human beings can be motivated by desires other than just the base acquisition and retention of power.

Any given game session could, after all, be based on a situation with a particular objective. Traveller, which eschews traditional experience systems, emphasizes that “players can play single scenarios or entire adventuring campaigns set in any science-fiction situation.” Its rules elaborate that “the scenario resembles a science-fiction novel, in that some basic goal or purpose is stated, and the adventure occurs as the group strives to achieve the goal. Usually, the scenario is a one-time affair and ends when the evening is over or the goal is reached.”52 If progression factors only into long-term episodic campaigns, then playing through a single scenario requires a goal other than progression: early tournament scenarios such as the famous D&D “Tomb of Horrors” at the Origins convention in the summer of 1975 established the defeat of a powerful adversary at the end of a dungeon as the most enduring objective. Tournaments typically employed pregenerated characters that were discarded after use; the documentation used by tournament referees rapidly evolved into the single-scenario commercial product that would soon be called a “module.”

The designers of Traveller must, however, acknowledge the unsatisfactory nature of unconnected adventures: even though the intention of such “one-time affairs” is for “the characters and situation to be discarded at the adventure’s end,” they find that, “strangely enough, players generally become attached to their characters and usually want to continue their lives in further adventures. To this purpose, the campaign is designed.”53 Once players began to identify with characters through playing their roles, this naturally evolved into a lasting bond. Still, the scope of role-playing games would include both protracted and one-shot adventures: the scenario versus campaign terminology would become the standard, appearing as well in titles such as Superhero ’44 and Runequest; in its section on the time requirement for play, the latter notes that it ranges “from a couple of hours for a quick scenario to years for a long-running campaign.”54

Implementing a scenario within a campaign with recurring characters can benefit from linking the scenario’s objectives to the rewards system. Uuhraah! (1978) specifies four scenarios for its prehistoric setting: a hunt, a migration, a dinosaur kill, and a war between tribes—though of course it notes that “the total number of scenarios that can be played is limited only by the referee’s imagination.” Uuhraah! scenarios were supposed to run fast: its designer attests that sessions “took an average of a half-hour to play,” though “quickie games can be played during half time.” In addition to awarding experience points for killing monsters, the rules also grant “experience for scenarios,” stipulating that “each character successfully fulfilling the conditions outlined at the beginning of play gains 10 experience points if he retains more than half his hit points, or 5 points if he retains less than half.”55 Before starting the next scenario, characters can trade in 100 experience points to raise any core ability by one.

By the late 1970s, more radical systems had begun to enter the market. In the design of Legacy (1978), David A. Feldt recognized that for most role-playing games “the primary factor which differentiates a scenario from a campaign . . . is the lack of an ongoing pressure to increase the characters’ abilities and characteristics as the single significant avenue of advancement.” Instead, “scenarios are always organized around some specific goal or task.”56 But there is no practical reason why a specific goal or task cannot similarly become the objective of an episodic campaign.

So, Feldt reimagined the concept of progression entirely: he proposed that instead of gaining experience levels, characters should accumulate “Information Levels” on specific topics integral to the game world. He explains this design decision in the rulebook: “I didn’t like experience points because there was no such thing even approximately in reality; because I didn’t agree that killing some pirates who happened to have a ship full of gold made me smarter, wiser, or better than studying magic in the library or practicing swordsmanship in the courtyard.” Instead, “I personally feel that it is a lot more interesting to wander around in the wilderness or in a dungeon searching for clues and information regarding what the deal is with the crazy world I am in and trying to take advantage of information I discover to increase my powers and abilities than I do wandering around in the wilderness and looking for treasure and killing monsters and wild beasts.”57

Determining “what the deal is with the crazy world I am in” recast the object of the game from gaining in power to resolving a mystery. Legacy ordinarily rates these Information Levels between 0 and 10, though for some topics the level may go as high as 25. Only at Level 7 do characters reach “the beginning of the real clues as to what’s going on and what might be in store at the higher levels of information.”58 As the referee informs players of the Information Level they have reached, this bestows a sense of quantified accomplishment not unlike gaining levels of experience.

Whether in a protracted campaign or a one-off scenario, liberating characters from their addiction to aggression and acquisitiveness opened new spaces for role playing. But we should probably understand the Legacy Information Level system not as a replacement for personal progression but as a way of measuring a different and perhaps more compelling axis of campaign advancement. Even Feldt had to confess, “I as well as many other gamers harbor a secret liking for experience points,”59 and his system does permit advancement through a complex system of “enablers” with various skill levels. The impulse for self-improvement would remain a key tool for motivating players and encouraging role playing when wielded properly by a referee.

Beyond gathering intelligence on a world or a scenario, characters might have their own personal purpose in a game world: the literature contains many echoes of players’ search for a personal goal beyond just becoming a “superperson.” Tom Smith, in a 1978 article, complained that his D&D players frequently pestered him with questions about their characters, such as “Who am I? Why am I here? Why do I have to get hacked up by these orcs when I could be herding cows or making shoes or some other menial (and safe) thing like that?” (AAW 20). Smith therefore set out to solve this “problem of character motivation” with a set of charts for generating backgrounds that would help players understand why their characters became adventurers. These include birthplace, social class, family situation and connections, and similar elements. But, most strikingly, it concludes with a chart about character motivation called “What the Hell Am I Doing Out Here, Anyway?” Outcomes of its percentile die roll include various reasons for seeking wealth or glory: for example, to impress a love interest or to escape criminal prosecution—which might be unjust or quite fair—or for being on some sort of religious pilgrimage or geas. On a roll of 00, the character is “working on a slow, subtle plan to conquer the world (30% chance of being insane).”

But should the direction of a character’s life be left to chance? Writing in the fall of 1979 in a piece called “Fantasy Role Playing: How Do You Play a Role?,” Paul Mosher sought more from role players than simply an understanding of “the character’s abilities, advantages, disadvantages, etc.” and how they affect play: he insisted that players invest characters with a personal goal (AE 50). The player “must decide what motivated the character to go out in the Big Wide World to make his way. Is the young woman leaving home to raise her family’s station in life or because of an unusual wanderlust? Is the Prince seeking a suitable wife so he may legally ascend the throne or has he been banished by the King? Once the player has determined this motivation, he is ready to breathe life into the character.” Mosher encouraged his players to fill in part of the backstory of the world, making them partners in the creative process of the game.

Tom Smith had titled his piece “Turning Numbers into People,” a catchphrase that summarized one of the main goals of the earliest players who talked about role playing. The idea of rolling against a chart to determine some overarching goal for a character fit nicely into the discussion of remaining “in character” that spanned the 1970s: just as characters are expected to act within the constraints of abilities, or alignment, so too can character behavior be guided by prescribed goals. The community immediately recognized that fidelity to these constraints on actions stood at odds with the traditional approach to wargaming and that this set these new games apart from earlier ones. But there was also a common tension across all these discussions around whether these constraints are imposed on players or chosen freely by them and indeed how much players even know about them. Characters with their own motivation for adventure, something beyond just self-empowerment, raised one of the most fundamental questions about role-playing games: Who or what really drives the events of a game?