Creating a New Table-top RPG System, Part 4


I had most of this blog entry written before GenCon, but it didn’t get published. I’m glad it didn’t, because my experiences at GenCon modified my perspective on a few points.

But before I dig in to “Decisions,” I want to answer one of the unanswered questions from Part 3 (thank you Jon for pointing out the omission).

What is the carrot for players and their characters?

As Jon mentioned, this is a tough question indeed! There are two significant parts to the question: What’s in it for the PLAYERS? and what’s in it for the CHARACTERS?

As important as this question is, I have difficulty distinguishing mine from other RPG games. There are many games on the market, but I must consider some other factors – the primary purpose of my game design is to support my novel(s). From that perspective, I think I can give answers.

For the characters in the novel, and by extension, the characters in the setting playing this adventure, their goal is to overcome specific challenges. This includes solving various mysteries, defeating evil forces, and bringing peace and prosperity to the world. This is their GOAL, but NOT their CARROT. My setting is a world without “gods.” In its history, all gods lost their powers and, over time, died off. Part of solving the mystery involves removing the curse that separates the gods from their powers, making them mortal. It is conceivable that one carrot is for the characters to become gods themselves. If not them, they can pave the way for others who do desire apotheosis.

Obviously, fame and fortune within the setting are carrots. Owning land, becoming rulers, gaining power and influence, etc., etc.

To answer this question with respect to the player, we need to look at why people play RPG games. And for me, I need to include writing a novel, or a series of novels, that accomplish the same, or similar, goals for the reader. RPG is a form of escapism. For the RPG player, it’s an opportunity to act, through a character, in different world and be a different person. While it’s true that many Player Characters are extensions of the players, there’s the feeling that the character can do and be something different than we can in our real lives. You and I can’t exactly go into a dark cave, face down and defeat a fearsome dragon, and escape with its treasure. Even if my character’s personality is only subtly different than my own, in the realm of imagination, it’s something I, personally, could never do.

The reader of the novel needs to have a similar experience. They need to connect with the main character(s) and experience their lives through their eyes. The reader wants to ride the emotional roller-coaster along with the characters. They want to see the character grow, change, and be the hero.

According to David Brin, my job, as an author, is to break up marriages and cost people their jobs, because they want to read the next page of the book rather than do whatever else it is they should be doing. As the author, and as the game master running the game, my carrot is to create that feeling in the readers and players. It doesn’t matter if it’s 3:00 AM and we’ve been at it for 9 hours straight. The reader doesn’t want to put the book down and the players don’t want to stop playing. That’s the carrot.

Now, let’s make some decisions. I’ll frame these decisions by asking questions and answering them. Here are the questions:

  • What terminology do I want to establish?
  • What are the core mechanics?
  • What values will be used to support the core mechanics?
  • What is the scale?
  • What is the randomizer?
  • What is the setting?
  • What rules are required?


For now, the name of my game is Sligo RPG. This name is meaningless other than it defines me as its author. It says nothing about the type of game, its mechanics, nor its setting. Likely this name will change later.

Characters will be described using Aspects. I like this term because it is broader than “attributes,” “stats,” or “traits,” which are more commonly used. There are two types of Aspects: Social Aspects and Physical Aspects. Within these two broad types, there are several Facets. Using these terms implies a holistic approach to the character’s description. We’ll cover this in more detail later.

Whenever one character is taking an action that requires task resolution, that character is referred to as the Actor. If the action being taken is against another character or object, the defender, or acted-upon, is referred to as the Target.

The player that runs the game is referred to in other games as Game Master or Dungeon Master. Again, I want to separate myself from convention and call this person the Conductor. This term has a different connotation; a Conductor leads an orchestra (or drives a train) which involves coordinating many parts with the goal of creating something wonderful. He’s as much a team member as he is a leader.

Core Mechanics

On the outset, I wanted to create a system where a single mechanic governed everything. I’ve since decided that this is neither practical nor necessary. Already, I divided the game into two components: Social and Physical. This implies that I need a mechanic for both.

Physical mechanics are easier to develop than Social mechanics. These are numeric abstractions based on statistical probability. The tough question is how great is the abstraction? Some players like a crunchy, detailed combat and task resolution system while other players want to breeze through this in favor of more narrative and role-play opportunity. It’s impossible to appeal to both camps. I can only build something that I like.

What I like compares to a tactical board game that resolves quickly but encourages the opportunity for role-playing. The combat and task resolution system must be easy to resolve, straightforward, logical, and not require a rules-lawyer or theoretical mathematician to determine how things should work.

Furthermore, I want greater participation. One thing I don’t like about other games’ systems is that when it’s not “your turn,” you have nothing to do. I want the system to be fast-paced and efficient. The last thing I want, however, is for combat encounters to become boring!

The system I have in mind will require some simple math and a chart. Two comparisons are made: The appropriate Facet value of the Actor minus the appropriate Facet value of the Target will determine what column is used on the chart. Dice are rolled by both the Actor and the Target. Again, the Target’s result is subtracted from the Actor’s result to determine the row on the chart. The outcome of the task is shown in the intersection.

There may be some situational modifiers. To keep math simple, all die-roll modifiers are positive. There are some situational modifiers that will shift the column to the left or right. My current idea is that environmental modifiers will adjust die rolls while role-playing modifiers will shift columns.

This leads us to the social mechanics. Social Aspect Facets won’t be numeric, meaning that determining success and failure can’t be resolved by numeric abstraction. It can only be handled through role-play. My goal is to emphasize role-playing, but there are times when the Conductor needs to rely on random chance to help determine how a social scenario should play out. For this, I plan to provide some simple reference charts as aids.

Values and Scale

The first draft of this article went through several iterations and brain-gushes on scale. I don’t want to go through that again here. The issue, as I understand it, has to do with logical probability. The questions break down as follows:

  •                 At what point is success assured?
  •                 At what point is failure guaranteed?
  •                 When two are opposing each other, at what point do we say that one will more likely succeed and the other will more likely fail?
  •                 What levels of gradient are required between the extremes?

To start this discussion, I’ll compare two extremes: Savage Worlds and Tunnels and Trolls.

In Savage Worlds, attributes are defined by die types: d4, d6, d8, d10, and d12. No character can ever have less than d4, and no character can ever have more than d12. This means there are 5 levels between “worst” and “best.” Through advancement, a character can upgrade their die type, but they are capped at d12.

Tunnels and Trolls starts with a 3d6 system (3-18) but gives both a “TARO” (triples add and roll over) and racially-based multipliers. Furthermore, character advancement places no upper limit on any attribute. It is conceivable, and quite probable, for a character to have stat values in the hundreds or even thousands.

In the proverbial one-on-one combat between characters, based on how the mechanics work, it is statistically possible, though very unlikely, for the minimum character in Savage Worlds to defeat the maximum character. Conversely, it is impossible for a minimum character to defeat a correspondingly “high-level” character in Tunnels and Trolls.

The question I have is: which paradigm do I want to emulate?

Part of this answer I’ve already covered in the section on Core Mechanics. For physical tasks, the Target’s Facet value is subtracted from the Actor’s Facet value. It is impossible for me to have infinite columns on my chart. It’s impractical to have more than fifteen or so. I need the chart to fit on a single page in landscape mode, with a large, readable font. This practical requirement enforces a limit on the chart’s size, which constrains the scale.

This means: the left-most column is “Actor can never succeed” and the right-most column is “Actor always succeeds.”

The next factor is my intention that every unique value is meaningful. I’ve always been annoyed by the D&D/Pathfinder model where Modifiers are calculated from the Attributes, and once that calculation is done, the Attributes become meaningless. This means that each column, representing the difference between Actor and Target, is incrementally different.

With fifteen columns, my scale of difference is 7. In actual numbers, the “best” is 8 and the “worst” is 1. If the Actor is best and the Target is worst, the column is +7, or far right. In my way of thinking, this is too close to Savage Worlds’ limited scale of 5 values. The variation is too extreme.

Though it strays a bit from my initial premise, I can create a bell-curve style spread in the column design. Say, for example, my best score is 25. The worst is still 1 (this is a constant.) My point spread is now 15. With a simple curve, I can make each column’s spread wider by one until the extremes are hit, and we’ll still have 15 columns:

-24 -16 -11 -7 -4 -2 -1 0 1 2 4 7 11 16 24
-23 -15 -10 -6 -3 -1 0 1 3 6 10 15 23 +

I like where this is going, but I don’t think the scale works. D&D 5e’s upper limit for player-characters is 20, which is achievable at first level. For monsters, the upper limit is 30, but with the flat-line scale, and the way the mechanics work, it is still possible for a first-level newb to defeat the mightiest beast of all. I know we’re dealing with fantasy, but even that suspends disbelief too far.

Let’s reign in my scale a bit, using a different numeric sequence:

-21 -13 -8 -5 -3 -2 -1 0 1 2 3 5 8 13 21
-20 -12 -7 -4 -2 -1 0 1 2 4 7 12 20 +

Fibonacci’s sequence, slightly hybridized, puts more emphasis in the middle and brings the extreme values a little closer. To get to the extreme column the difference must be 21 or greater. I like this. It gives a wider variation of stat ability but doesn’t impose a top limit. If the Actor’s value 30, the Target needs at least 10 in their value to move off the edge of the chart. Additionally, a difference of less than 3 isn’t going to be huge, but still has significance.

We can now focus on the range. I stated previously that 1 is minimum, no matter what. In my game, there are NO negative modifiers or values. Negative is the result of a calculation. If I have 1 in a Facet, it means that I am an untrained novice with no knowledge or skill. Give me a set of nun-chucks and I’m just as likely to whack myself in the back of the head with them as I am to hit someone else. By extension, the far-left side of the chart has a higher probability of fumbling.

With this decision, I now know my extreme ranges and I have a rough idea of the gradients. The next step is to estimate where my top value should fall. Why? I don’t want to be T&T and have unlimited stats. Because of this, T&T is impossible to balance at higher levels. The designer of the game cares little about this type of balance, but my experiences, as player, game master, and author of solitaire adventures, taught me that this type of balance must be considered to keep the game from becoming an exercise in rolling and counting hundreds of dice. And if you’re not great at math, this is a huge challenge. In my D&D 5e game, I have players who struggle adding damage dice and damage modifiers. I don’t want this to be any more of a barrier than necessary. The next number in the sequence, as shown in the chart, is 34. This seems good! Let’s go with it!

Decision made: 1 is no knowledge, skill, aptitude or ability. 34 is the ultimate godlike maximum where there is no better.

Next question: What is the Human average? It’s not 17. Humans aren’t “half as good as gods.” Again, I refer to the chart. Counting just the positive side, starting with zero and going right, and adding the unnecessary 8th column, the middle column is the 5-7 range. Works for me. Based on this arbitration, the “average” Facet Value for Physical Aspects should be 5-7. Do we have a consensus? I think we do. Let’s move on.


I learned a valuable lesson with my prior attempt at a role-playing system: some people can’t shuffle cards! The selling point of No-Dice was that the card system allowed the player to influence the random outcome at the meta-game level. It brought in a karmic balance that rolling dice can’t achieve.

But is that true? I played a proto-type RPG at Gen Con a few years ago where, at the start of the session, 10 (or so) six-sided dice were rolled. The dice were placed on the character sheet without changing the results across the attributes. During play, rather than rolling dice to resolve a conflict, the player chose which dice from the character sheet to use. When used, the die was put aside. When all dice were used in this manner, they were all re-rolled and re-assigned.

While clever, I thought the system sucked. It became a game of dwindling odds and diminishing chance. Do I use my better rolls now or save them until later? As a role-playing game, it didn’t matter. The GM had knowledge of what we players had and adjusted the challenges. This isn’t a bad idea for a board-game mechanic but has minimal adaptability to role-playing task resolution.

But can I devise a dice-rolling system that integrates karmic balance, and keep it simple enough? For that matter, do I want to do away with dice completely?

Here are some options:

(Right now, I’m comparing with a d20-based system)

  • Twenty cards, numbered 1 to 20. The player chooses which card to use in a given situation and once used, can’t be re-used until all cards have been used.
  • Twenty cards, numbered 1 to 20, drawn randomly. Not recycled until all have been used.
  • Roll several dice in advance. Choose which to use. Re-roll all dice when all have been used.

I don’t like any of these ideas. The card systems mean creating custom cards for all players. This becomes exponentially unmanageable for players running multiple characters and worse, for the Conductor, trying to manage NPCs, monsters, and the environment with this type of system. Rolling dice in advance presumes the results can be preserved. Obviously, there are easy work-arounds: writing down the results and crossing them off as they go.

All these ideas share a common flaw: what if all the numbers aren’t used? In an average 4-hour D&D 5e session, how many times does a player roll their d20? Legitimately? I’d say it’s less than 20. I’d say 10 to 15, and that’s a session with two or three combat encounters. Maybe I should keep track next week. (I will if I remember!) This means, for a player that knows this, the lowest results will never get used. Ever.

Another way to instill karmic balance is to roll multiple dice. While not true balance, this creates a pyramid or bell-curve result scale, where the middle result is more common than the extremes. The more dice that are added, the more rounded the curve and, by extension, the less chance of rolling extremes.

A third way to create the bell curve is to use “averaging dice.” This is a semi-customized type of die where the top and bottom faces are replaced with the middle faces. Instead of 1-2-3-4-6 on a d6 we have 2-3-3-4-4-5 instead. Let’s throw this one out now; I’m not going to use custom or non-standard dice. (Fate dice fall into this category.)

The final way is to use hit/miss architecture. Shadowrun. Roll a handful of dice. Count the 5s and 6s, count the 1s, ignore everything else. This merges the concepts of multiple dice and averaging dice. On the positive side, it uses standard dice. On the negative side, this system is all about adding more dice to the pool to improve your odds. I’ve seen it get ridiculous, where a 36-die cube wasn’t enough. Rolling combat for a monster with an MR of 1000 in Tunnels & Trolls is worse, as it requires 101 dice, but I digress.

Or, we can forgo the idea of karmic balance and just roll a d20 and be done with it. D20s are common and popular, easily recognized, and usually easy to read. Most people can math with numbers of 20 or less without too much trouble.

There is one thing I must consider, however, is my one-page chart. Subtracting the Target’s roll from the Actor’s roll means my potential range is -19 to +19, or 39 possible results. That’s a lot! In the chart, that equates to 585 cells! The other aspect to consider is the probability of the extremes. For d20s, the chance of each extreme is 1 in 400. This is extremely granular. Let’s run the numbers for the other die types:

Die Type Range Extreme Probability Cells in chart
d20 -19 to +19 (39) 1 in 400 585
d12 -11 to +11 (23) 1 in 144 345
d10 -9 to +9 (19) 1 in 100 285
d8 -7 to +7 (15) 1 in 64 225
d6 -5 to +5 (11) 1 in 36 165
d4 -3 to +3 (7) 1 in 16 105

If d20 is too much, d4 is too little. There is one factor that will help sway my decision: synergy. In the prior section, we decided that the average Facet value for a human is 5-7. The average roll on a single d6 is 3.5. A d8 is 4.5, a d10 is 5.5, and a d12 is 6.5. Based on this, we’re narrowed down to either d8 or d10. There are systems that use d10s, some in the form of d100 (percentile dice), but are there systems built around the eight-sided die?

Interesting side note, gold ore usually manifests as an octahedron.

Let’s roll with it! (Pun intended). D8 is our die.

I did a quick online search. There are a lot of systems that use d8, a series called “Free Style (Space Gamer)” that uses d8 as the primary die but uses others as well. I couldn’t find any that use d8 exclusively. This doesn’t mean they don’t exist.


I already decided to start with the Neuith Setting I’ve been using for D&D 5e. Most of you aren’t aware, but this setting was originally created to support a DT&T (Deluxe Tunnels & Trolls) adventure. Over the past three years, this setting has taken on a life of its own and of all the settings I’ve created over the years, it’s likely the most complete and detailed. That’s not to say more can’t be added!

About a year ago I came across a “Twenty Questions for Your Homebrew Setting” list. My answers to those questions are still valid:

Other entries in my blog discuss how the setting originated and how I developed it. I’m not going to repeat myself. Go read those articles.

Required Rules

This is the last section for this blog. I’m not writing the rules here, though. I’m just outlining the rules I need along with some commentary.

Character Creation

My concept of character creation will model the classic point-build system. The Physical Aspect Facets are numeric values ranging from 1 to 8. I haven’t yet determined what the Facets are, which will drive the point total and the build process. I want to use a non-linear increasing cost model for higher skill levels.

Social Aspect Facets are based on single word/phrase descriptors. This will be a challenge to create, but once designed, I think will end up being very easy for the players. There will be a set of Facets equating to social interaction types, and for each, the player will choose which word or phrase best fits the character’s emerging personality.

By separating the Aspects in this manner, players will not be able to create “min/max” or “optimize” by maxing the physical Facets and minimizing the social Facets. My objective is to remove the arbitrary numeric assignment for these Facets in favor of a descriptor, which I believe will enhance role-playing. Furthermore, as an author, I can better keep character personalities consistent with verbal descriptors than with numeric assignments.

Task Resolution

As discussed, I’m going with an interactive model where there is an Actor and a Target. The Actor announces his or her action, which defines what Facet is being used. The Target determines their response, which defines what Facet they are using. The Target’s Facet value is subtracted from the Actor’s Facet value. The result determines which column on the chart is being used. Both parties can add role-play effects. Based on general assessment, the Conductor determines if a column shift is warranted. Next, each player rolls an eight-sided die. Again, the Target’s roll is subtracted from the Actor’s roll. If there are any environmental factors that can adjust this roll, a modifier is applied. The chart is consulted, and the resolution is announced.

The rules will need to clarify what constitutes a column shift. Arguably, if both parties present good role-play, there would be no column shift. The brief list of environmental factors should be shown on the same page as the main chart. I don’t want page-flipping during game play if it can be avoided. Environmental factors include things like tactical advantages and disadvantages (cover, flanking, etc.)

If the Target or the Actor is an inanimate object, such as a character trying to open a locked gate or a scythe blade swinging from the ceiling, the process is the same. The object gets a set Facet value based on the relative difficulty. A table will be provided to give the Conductor guidance. In this case, an object is unable to role-play, giving the player character an advantage.

The third chart listed on the same page is the summarized explanation of the Resolution Codes. Each Resolution Code translates to several possible outcomes depending upon the situation. The general intention is to avoid the binary, yes/no result, but to allow for degrees of success and failure. This means the code needs to incorporate the amount of damage (for combat) and cover critical failure and success.

Equipment (Weapons and Armor)

This opens the door to how weapons and armor work. Armor and shields would be considered Environmental factors, but they should be enhanced by skills and abilities. A novice using a shield won’t be as good with it as a veteran. Most other games don’t make this distinction, and if they do, it’s over-simplified. The same is true for armor, but it is less apparent.

Different weapons do different amounts and distinct types of damage. Again, this is highly influenced by relative skill. Also, something that other systems rarely consider, is that some weapons are more effective against certain types of armor than others. While I’d like to include these considerations, I’m tempered with the need to abstract for game play.

The way weapons work needs to work seamlessly with the task resolution chart. Each resolution code specifies the amount of “damage” that is dealt for a successful it. This alleviates the need for a second die roll and it ties the damage to the severity of the hit. Conceptually, “barely hit” means minimal damage, where “solid, definite hit” means higher damage.

Ahh… Just had an idea. Greg Harter’s G-20 system is loosely based on RoleMaster (aka “ChartMaster”) where each weapon has its own to-hit and damage chart. I’m emulating this system slightly in that I combine to hit and damage into one chart, however, I only have one chart, not dozens like his system. What I can do, however, is have the resolution code refer to the weapon’s data, which would be recorded on the character sheet (or weapon card if I go that route… still not ruling out custom cards, I just won’t use them for random number generation!) The letters ranging from A to ?? correlate to damage. For a dagger, “A” may mean 1 point of stabbing damage, “B” is 2 points, and so on. For a longsword, the scale may be wider: A=1, B=3, C=5, and so on.

Another aspect I’d like to explore is Armor Ablation. Rather than using Armor as an environmental factor, Armor gains an Ablation value, meaning that certain types of damage are reduced.

This concept will translate well to non-weapon damage, such as damage from traps, falling, and magical attacks.

This leads us to the next set of rules…

Magic System

Magic is often very complicated and reserved only for experienced players. In addition, it is intricately tied to the setting. Before we create a magic system, we must establish the pseudo-science of how magic works. Well, we don’t have to, but doing so will help make the system more robust, logical, playable, and balanced.

In many systems, magic users are either too powerful or too weak. Often described as “glass cannons” they can deal a lot of damage in a limited fashion, but if they are attacked directly, they go down easy. This isn’t necessarily a terrible thing, but it’s not something I want to enforce.

Something else that drives me nuts about some other systems is the concept of “once per day.” I prefer a mana-based derived power/fuel system. I also like a system where a caster can focus on a single, common thread or force, and expand upon it as they gain experience. This is where the understanding of how magic works within the setting becomes important.

I was going to start a conversation about the “science” behind magic in Neuith, but upon reflection, that’s not a subject appropriate for this discussion. This is something that will get its own blog entry. For now, we just need to understand that magic will use the same core mechanics for both physical and social interactions as previously described.

Rules of Grit

One thing I learned at Gen Con is that the omission of a rule is just as important as inclusion. There are numerous potential rules that add realism to the game but add little to enhance role-play. Many of these rules don’t even impact combat, or if they do, they become an inconvenience.

I’ll list some here and talk about how I’ll handle them in my game:

Encumbrance: This is likely the most ignored rule in all RPGs. For this game, the player and the Conductor need to have an agreement on what is practical and logical. There will be no number crunching.

Damage Types: I’ve already decided that I will use damage types, because this is part of the realism I want to convey during combat. I’ll strive to avoid commonly used names. Also, it’s my opinion that most weapons can do more than one type of damage. I believe I can reflect this with the weapon data without breaking the rhythm of play.

Tactical Maneuvering: My preference is to use a hexagonal grid where each hex is 5 feet across. While this is common across different systems, it’s also easy. When we have tactical maneuvering, the system will require a set of relatively simple rules for movement, facing, positioning, and so on. I will probably write an entire article on this subject.

Mass Combat: Every attempt I’ve seen for this fails. Mass combat, simply put, is a different game. If you’re willing to step out of role-playing for a session or two and play a board game, go do it. If you want mass combat in a role-playing game, it’s not going to work. Mass combat should be handled through role-playing. Prior to the “battle,” the Conductor should already know the outcome if the characters don’t participate. Let the players write the story and let their involvement influence the outcome. As it should.

Setting-Specific Rules: This is a broad topic that must be covered in its own blog entry. This covers everything from Encounter Generation to Economy. Much of this ends up as reference material for the Conductor, with the caveat that it can be modified to suit the needs of the adventure, story, and group.

Character Advancement and Specialization: This is an extension of Character Creation. With both Physical and Social Aspects, everything that can be done here becomes an extension or adjustment to the existing rules. Conceptually, everything a character can do should be accomplish-able within the ruleset. The thought I have regarding Social Aspects is that the basic personality of a character either doesn’t change at all or if it does, the changes are minimal. On the same token, I want to be able to show how a character has grown over time and through experience. I have some thoughts on how to accomplish this, but I need the core character creation system developed first.


This article outlined most of the major decisions required to build my new game. I’m sure I’ve left out a few things, but there’s room to expand as well. I don’t want to get too far ahead of myself; my instinct is to start putting numbers into charts and compiling lists and crafting spells and all that. I must resist that urge because it’s easier, in the long run, to build things right the first time than to have to go back and fix things because I missed an important detail.

This is where I open the floor to you, my avid reader(s) and teeming fan(s). Tell me what I’m overlooking. Give me your thoughts on where I’m headed. Do my ideas make sense? Am I crafting a game you’d be interested in playing?





























Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s