Flat Math, Deep Characters

I have played and built primarily class based systems. I grew up playing various incarnations of d20, most notably Dungeons and Dragons, always eager to try out some different class that I hadn’t played before, some occupation or career or appellation that bestowed new mechanical advantages. I reveled in it and I didn’t think that RPGs had any other option. Yet, as I grew older and my exposure to various systems increased, I discovered that a large number of classless systems were in print and in play, even amongst some of my friend’s tables.

The enormous advantage of a class-based system, or one that categorizes characters into easy to describe chunks and skillsets, is that it is very easy to organize and to delineate roles. Having roles in a party is a relatively old concept, one which has been streamlined and incorporated more heavily into more recent incarnations of RPGs (such as D&D4e). Classes make it easy for players to fall into their role, and provide enough structure to allow for newer players to simply select their class and progress along its track. Players and gamers are used to selecting classes as seen in Diablo II and World of Warcraft and therefore it is a salient idea that is easy to grasp.

But one of the greatest disadvantages of a class system is that it is inherently restrictive. The entire purpose of the class is to naturally “balance” certain characters against others, fulfilling the roles that are necessary within a typical party makeup, and to limit those character’s options so that they can not excel at everything equally. Yet what if a player wishes to make a character that is not easily winnowed into one of the ascribed classes? The answer for years was simply to publish more classes, or to build individual classes, but as D&D3.5 evolved, multi-classing and Prestige classing (a form of a focused sub-class) became almost necessary to optimize a character and to fit in with the player’s original concept.

I remember building and playing a character that was made up of none less than eight classes, which was perfectly “legal” within the system. Yet, it became such an exercise in cherry-picking that I realized the very nature of the class system just made everything convoluted and stifling. It made the generation of creative, optimized characters a nightmare for those who didn’t wish to spend dozens of hours poring over seedy forum threads talking about DPR and max to hit by level.

When I originally built OMEN, I intended for it to be a universal tack-on for d20 and then eventually D&D4e. I had settled on six classes, each of which was balanced through the combination of technical skills, durability, and combat prowess. Yet that felt entirely boring to me and to my playtesters, but adding on more than six classes was a logistical and systemic challenge. It was something that I didn’t really want to pursue; slicing hairs into incrementally finer fractions continues to create the problems I discussed about D&D.

And so when the major redesign of OMEN took place, one of the first things to go was classes. But how to replace such a staple of the RPG genre? Most of the other classless systems in existence use a method to “buy in” or “opt in” or “opt out” of certain abilities that allows the player to focus his or her character’s skills. This seemed logical enough to me, but as I began to tinker with and study more of the popular systems, the complaints that I had heard about classless systems began to surface. It was entirely possible to make a character that was overwhelmingly powerful compared to others or genuinely wreck a character enough in the generation process that they were more of a liability than an asset to the party.

One of the largest reasons for this was the wide array of choices and combinatory mechanics. The ability to stack up bonuses from multiple sources is inherently logical, and an idea that I rather liked, but also could create some wickedly strong monsters of a character. This isn’t an entirely bad thing, but it could get very out of hand for a system that espoused heroic action, and only one or two of the characters were optimized enough to act heroically.

OMEN is not designed to promote Homeric heroes: supernatural beings with godly strength and power. Instead, I wanted the characters to feel alive and organic. I wanted the players to believe that their character could actually exist within the universe, and that that universe was founded upon reality with minimal assumptions made. People are fragile, people are fallible, and some people are simply better than others. These were things that I looked at, as well as other classless systems that I had experience with (and others that I didn’t) to examine how best to approach it.

I had tried percentiles, but the huge numbers became a problem, and the disparity between the dice roll and the bonuses grew wider and wider as the characters advanced in level. Since I had already determined I wanted to use 3d6 as the only dice mechanic in OMEN, I wanted the numbers to stay pretty consistently less than 50. And then I stumbled upon micro-lite systems; ones that use small amounts of foundational mechanics but don’t have to fill a six hundred page tome to get the concepts across. This fell in line with the design principles that I wanted for OMEN, and rather copy or implement how those games accomplished the task, I set certain limitations on myself as a game designer to accomplish the style of game that I wanted.

I like round numbers. One of my favorite numbers for mechanical reasons is ten. Multiples of ten are easy to manipulate, and serve as a good representation of percentages. Ten also happened to be relatively close to the average number for an unmodified 3d6 roll, and therefore the entire pillar of the system was formed: ten.

There would be ten points to allocate into the various aspects of a character’s creation and “build” or “skills” or whatever one wanted to call them. I wanted the aspects of a character to become increasingly more focused, starting with a large category at the top, such as the physical and mental attributes, and ending with the specific zones of expertise for a character, such as his specialties and skills.

Since I wanted my players to be able to add easily, I wanted the numbers to modify the dice roll (with an average of ten) directly. Some systems will add half level, or three-quarters of one this, or one quarter of that, or one every seventy five points of this, and so on. I’m no mathematical genius, nor are most of the players that I have gamed with. I wanted people to look at a handful of numbers and add them instantly rather than dividing and manipulating them in some complicated manner. So I determined that OMEN would become a “flat math” system. By flat I mean that the numbers stayed relatively low, and that they didn’t fraction, split, or multiply. Things would increase by one at a time, and the numbers would stay manageable.

This meant that different aspects of the character would also have to add directly together, and while I knew this could create stacking, I was completely aware of its potential and so built the system to inherently encourage and reflect that, while naturally limiting that highly focused character in other areas.

To start off, the big broad strokes that would define the character’s genetics and natural abilities. I decided to call them Attributes, and would have to keep a relatively low number of them. If the number on the Attribute itself would apply to the roll, with an average of ten, then the numbers should be between one and six, with maybe ten being the absolute highest. And since I wanted to utilize the number ten, I wanted players to have ten Attribute “points” to allocate, and that those numbers would directly influence a roll falling under those attributes. To prevent complete and utter unnatural “min-maxing,” or the tendency to build an extremely slanted character, I set the first level cap for a single attribute to be four. Setting the limit to four also meant that there would be six points left over, which split nicely into three categories. The four Attributes therefore were created: Strength, Resilience, Agility, and Mind.

Strength for all of the physical prowess and athleticism, Resilience for a character’s ability to withstand punishment and harsh conditions, Agility to represent his dexterity and natural coordination, and Mind to reflect his wisdom, intelligence, and natural ability to understand and influence people and things. Having ten points also meant that a standard “array” could be used: 4, 3, 2, 1. This created a balanced character that was interesting and dynamic, but the 10 split into 4 has many combinations that can create many different variations.

This same principle would be applied to the other aspects of the character, which I wanted to keep to a relatively low number. I settled on just two (for creation purposes) that would utilize this same mechanic: Skills and Specialties.

Skills would be the general purpose training and natural affinity in certain tasks for a character, while the Specialties would be specific and focused areas within the Skills. This meant that a character could be generally good at something, such as Ranged Combat, but be exceptionally good at a certain type of Ranged Combat. From there I determined the various Skills and Specialties, all with an eye for what made sense and worked for OMEN, but was generic enough to be utilized in most if not at all “realistic” science fiction or military-style RPGs.

Ten points would be allocated into the eight Skills, and ten points would be available to allocate into the Specialties of those Skills. Maintaining the same balance of a maximum of four points (at first level) this created characters that would have to spread into at least three Skills and three Specialties (4, 4, and 2) but encouraged even more widely invested characters. The dilemma of stacking and creating characters that had unnaturally high numbers was relatively solved, and keeping the amount of points handed out for future levels and the level cap low, a character’s roll should always be under 50 (barring certain circumstances.)

What developed was sort of an epiphany for me and for many of my friends and colleagues: the simplicity and flexibility of the system allowed for truly and inherently creative character creation and play. The low number of Attributes, Skills, and Specialties kept the option overload to a minimum for new players, but provided enough nuance for those who enjoyed the optimization process to truly build the character that they wanted without having to dabble in ten classes. We wanted people to build the character they wanted to play at level one, and have that character naturally and flexibly evolve as they progressed.

We feel pretty proud that we accomplished the goal; and the players so far love it.

Read more »

Dice Light, Fun Heavy

One of the greatest advantages of playing a video game over a tabletop game is that all of the complicated formulae are calculated instantly. I’ve known many gamers over the years, who are otherwise entirely intelligent individuals, that have had an almost unbreakable detestation of numbers. One of my colleagues has suggested that small number addition is as difficult to some as reading is as difficult to those with dyslexia, and I’ve seen enough hesitation and trouble with adding up the different dice and their modifiers amongst my tables for years to believe this is probably correct.

While numberless systems are typically elegant and provide for excellent storytelling, they tend to lack the amount of control that most of the players I’ve come across enjoy. There is that certain competitive edge that people seek, even during the character creation process, that numberless systems do nothing to little to promote. I am sure there are exceptions, but in my experience, that has almost entirely been the case.

Due to the design goal of regulating combat, a numbered system for OMEN was completely necessary, and as stated in an earlier article, I had settled for the OGL d20. It was most familiar to me throughout the years, and had a substantial customer base, so why not? Yet, as many of you know, the d20 system is rather like using twenty-five tools when two will do. I abandoned it shortly after beginning the design of OMEN.

I had tinkered in the past with percentile systems, and had played in several that used “roll-under,” but I thought it somewhat counter-intuitive, especially to newer players who couldn’t grasp the vagaries of probabilities and the like. I had GMd games more than one in which people asked, “But shouldn’t a higher roll be better?” Even the older editions of d20 had used a roll-under mechanic, but had moved on to “higher is better” thinking as the system evolved. I too determined that for OMEN, higher numbers should always be better.

One of my greatest issues with many systems is the dependence of many different dice. This I felt was one of the greatest impediments to including new players, and one of the things that many players got snagged up on during the calculation process. Certainly the dice added a sort of romantic, mysterious appeal in the early days, but the esoteric nature of such peculiar dice has become a running joke to those who are unfamiliar with RPGs. As OMEN is intended to garner as much of an audience, if not more, from brand new players, settling on a streamlined dice mechanic was deemed appropriate.

I had done quite a bit of research regarding the different probabilities of dice mechanics, as I dabbled in various MMORPGs and their tabletop counterparts. I thought for many years that the “cleanest” were percentile systems, the higher the percent the better your likelihood of success. The issue with this was the rather large numbers that came along with it; while adding twenty seven and thirteen is not entirely difficult, adding multiple rolls with bonuses and penalties culminating in a number in the hundreds is quite another matter.

I had even completed a percentile system for a science fiction game I intend to re-work in the future, but it was unnecessarily complicated and entirely the wrong direction for OMEN. So I settled on using a die that nearly every household possesses: the classic six sided die, or referred to as the d6 in the RPG industry. The d6 was ubiquitous, cheap; even free with many other games out there that were common to the house hold. The numbers were easy to read and add, and people were used to rolling and using them. Cubed dice also seem to have one of the better physical rolling mechanics, they clatter around clamorously and then settle quite nicely; it is almost impossible to have a d6 land crookedly.

Another benefit to the d6 is that the numbers remain low. I was a fan of “measured randomness,” and one of my favorite parts of the d20 system was using multiple d6 to generate character attributes. One of the benefits of this design mechanic is that the roll probabilities are bell-curved, meaning that rolling an average is much more likely than in a linear mechanic such as the d20. This would mimic reality somewhat better in my opinion: a degree of randomness that is definitely measurable, but the great majority of the time a character could behave rather consistently.

I settled on using three d6, hereby abbreviated as 3d6, to create this clean bell curve figure and keep the numbers manageable. Since I decided that I strictly wished to use d6s, attacks, damage, and defenses would all have to be approached differently than the standard. The majority of systems I have experience with use two distinctly different mechanics for resolving successes and determining the degree of that success.

For example, in the d20 system, a twenty-sided die (the d20) is rolled, modifiers are applied, and then that roll is compared against the target’s defense or difficulty. Once success is determined, the effect comes into play. This effect typically involves rolling multiple and different dice than the d20, adding them together with any additional modifiers, and applying it to the target. This could cause a couple of things to occur: primarily slowing down the game, but more significantly abstracting the success of that roll to a point that just didn’t make sense.

For example, with the d20 system as described above, a warrior could hit his enemy with a very high roll of the attack (using the d20) against his target’s defense, but then roll very low damage (with all of the other dice.) This essentially robbed the fun for the player who rolled the great hit, as the payoff wasn’t directly related to the success of the roll.

I wanted to get rid of this double mechanic, which worked in seamlessly with my goal of having conflict resolution be fast and using a single dice set. As opposed to rolling once (or multiple times) to determine success, and then rolling again to determine effect, I would combine the two into a single roll and compare that roll against the target. This could be used with static target numbers or rolls. The difference between the action and the target would determine the degree of success.

The system immediately clicked for me. It was fast and it made sense. The significance of the weapon in the d20 system, generating the larger die values, was paramount in determining combat effectiveness. If small dice were used, high modifiers had to be added, or large amounts of attacks had to be made in order to balance it out. This kind of system heavily favored huge powerful weapons. I didn’t want that in OMEN.

The people of OMEN typically used guns, or small and improvised melee weapons that wouldn’t slow them down. These kinds of weapons would still be deadly, as the difference of the weapons would not be as significant of a factor as the skill of the wielder. An extremely skilled combatant would consistently beat the average thug, even with a better weapon; but that threat would always exist. A lucky shot, in this case embodied by the roll of the dice, could land in that perfect spot and kill even the most skilled.

This went well with the realistic and somewhat gritty tone I was going for with OMEN; it played fast, it played smooth, and it made combats extremely fun and dangerous. This made tactics, cover, and strategic choices all the more important. The math was relatively simple, with numbers in the ten to forty range, and the math was fairly flat. The playtesters who had slogged through the convoluted mess that was the OMEN d20 system now had hours of fun in conjunction with combat, role-play, and all manners of obstacles requiring rolls. The system began to evolve and become refined, and I was happy with what I had created.

In another article I will discuss in more detail how the skill and character progression system was built, and why.    

Read more »