Second Thoughts on Attributes

I’ve been thinking about attributes and I have to say I’m not really happy with the solution I settled on. While a modifier-only attribute system starting at 0 is very straightforward, it introduces two problems. First, even low-level characters have really high bonuses to everything, which definitely fails on the elegance front. Second, it creates major issues with a few aspects of the game math, such as hit points and damage. Therefore, I’ve realized that I really want a modifier of 0 to be typical.

To define a modifier of 0 as average, I have two options. The first is to go back to a score and modifier system, with the average score as 10 and average modifier 0. With a 1-to-1 correlation between score and modifier, however, the score itself is basically useless. The second option is to base a modifier-only system around 0. I avoided this on the first pass because I disliked the arbitrary negative score limit; but what if this limit applied in both directions? Attributes could range from -10 to 10, with an average at 0. This feels much cleaner to me, and satisfies my conditions of clarity, simplicity, and elegance.

But wait—I have another condition, verisimilitude. I’m going to look at a few test cases to make sure this system feels right. Most human attributes will be 0 +/- 2; this means that the difference between the best and worst of typical humans is a tier’s worth of experience or a training talent. This seems right—training and experience can overcome weakness, and someone with professional training should be on equal footing with a very talented amateur. Ranging farther out, an attribute of -5 could be defined either as a severe weakness in that area, such as extreme clumsiness (Dexterity) or social ineptitude (Charisma), or as a notable but non-crippling disability, such as a lame leg (Dexterity), lame arm (Strength), or a missing eye or ear (Awareness). An attribute of -10 indicates extraordinary weakness and/or severe disability, such as lacking an immune system (Constitution) or widespread paralysis (Strength/Dexterity). Going in the other direction, an attribute of +5 reflects standout ability, such as that shown by Olympic athletes or genius-level intelligence. An attribute of +10 represents near-supernatural talent and ability such as Herculean strength.

That all feels good for humans. What about creatures and monsters? There are more opportunities for extreme attributes here, but I think it is still workable. The strongest and toughest creatures are also large, which means they have a size bonus that can effectively increase their Strength and Constitution beyond the cap. On the mental side, creatures with superhuman Intelligence or Charisma could certainly exist; however, the upper 3-5 points of the scale are already defined as a supernatural level of ability, so these should work as well.

So there we are, hopefully for good this time. Attributes range from -10 to 10, and most humans have scores of 0 +/- 2. With that fixed, time to move on and discuss some of the system math.

Core Design Goals

I’ve touched on some of my design goals, but I haven’t set them out clearly anywhere. I’d like to do that before going any farther, so here goes. Roughly in order of priority:

1) Clarity: There should be minimal opportunities for the rules to be misunderstood. One key area that I’m targeting is overloaded words such as level and ability; wherever possible, rules terms should have unique names.

2) Simplicity: The game should be playable with minimal reference to the rules. Nothing breaks the tension like having to look up a rule or a modifier. My main target here is combat mechanics, though the skill mechanics could do with a good deal more uniformity as well.

3) Depth: The game should have a wide variety of meaningful options. This is one that requires a delicate touch. Too many options and you wind up with system bloat that impacts the goal of simplicity. Too few, and characters all start to feel the same. Spells, weapons, and armor are all areas that this goal is especially relevant for.

4) Elegance: This one has two meanings. On one side, it really overlaps with clarity and simplicity; the rules should be clean and easy to use. In addition, the system math should be clean and functional at all levels of play.

5) Verisimilitude: Realism is a problematic term when fantasy is involved, but at the very least internal consistency needs to be maintained. At low levels this can mean comparison to real-world physics and human ability, while at higher levels internal comparisons take over.

Now, back to design work!

Size Categories

The issue of creature sizes was an important consideration as I was determining the system math. In Pathfinder, creature receive bonuses and penalties to all of their physical attributes based on size. This creates problems in a few cases--for example, the Strength bonus given to very large creatures completely negates their size penalty to hit. In order to address this, I decided that for Telharis all physical attributes would be measured relative to a typical creature of that size. To compensate, size modifiers apply to more things. I've also inverted the size modifiers to minimize the number of calculations that involve subtraction; small creatures receive a negative size modifier, and large creatures a positive one. The size modifier is added to maneuver bonus, maneuver defense, damage rolls, hit points, and Fortitude saves. It is subtracted from attack rolls, defense, Reflex saves, and Stealth checks.

I've also adjusted the size categories compared to Pathfinder. Flipping through the Bestiary, I found that the Fine size category was effectively limited to creatures that made up swarms. Overall, the size modifiers seemed too low. After some tinkering, I came up with the following categories:

Size           Modifier     Space     Reach       Examples
Fine                -10           1 ft.         0            Mice, coins, a typical insect--little more than points.
Tiny                -5             2.5 ft.      0           Cats, rats, most birds.
Small              -2             5 ft.       5 ft.         Halflings, dogs.
Medium           0             5 ft.       5 ft.         Humans, black bear.
Large              +2            10 ft.      5 ft.        Horses, cows, lions, a small dragon.
Huge               +5            15 ft.      10 ft.      Elephants, a medium dragon.
Colossal         +10          25 ft.       15 ft.      The proverbial broad side of a barn.
Immense        +15          35+ ft.    20 ft.       Kraken? Really old dragon?

I should note, I don't foresee the Immense category getting much use, and when it does, it may require special mechanics--a creature of that size isn't going to be hitting anything with a -15 on attack rolls. On the other hand, a creature of that size's attacks are probably more like an area of effect that would require a Reflex save. I'll deal with that if and when I come to it.

Also, I should maybe explain my 'proverbial broad side of a barn' example. What I mean by that is that an average level 1 human literally could not miss an attack against a stationary object of that size.

That's all for now.

Skills

I gave a brief preview of skills while I was discussing attributes; I’d like to spend a little more time on them now. First, I will list the skills I intend to use and give a brief description of each:

Acrobatics: jumping, balancing, tumbling.

Climb: what it says.

Deception: misdirection, lying, and disguise.

Endurance: ability to run or fight for long periods, resist heat, cold, and fatigue. Distinct from Fortitude saves, which are used to resist poison, disease, nausea. Endurance lets you ignore affliction, Fortitude lets you resist it.

Escape Artist: escaping bonds, squeezing through tight spaces.

Handle Animal: train and direct animals.

Heal: what it says.

Intimidate: what it says.

Lore: any application of the Pathfinder Knowledge skill that doesn’t fit under an existing skill—e.g. history, legends.

Perception: how sharp your senses are.

Persuasion: what it says (replaces Pathfinder Diplomacy)

Profession: making money with your skills. Also includes crafting.

Ride: what it says.

Security: lockpicking, trap disarming, trapsmithing.

Sense Motive: opposes Deception; can also be used to make a gut assessment of a situation.

Sleight of Hand: stealing.

Spellcraft: knowledge of magic.

Stealth: what it says.

Streetwise: getting by in the city, gathering information.

Survival: getting by in the wilderness, tracking, general outdoor skills and knowledge of nature.

Swim: what it says.

I intend to emphasize that the key attributes listed for each skill are only the attributes that are most commonly relevant; in many cases, substitute attributes can and should be used. For example, what in Pathfinder would be a Knowledge: Nature check is in Telharis a Survival check using Intelligence instead of Awareness. Another example would be using a Strength-based Acrobatics check to jump.

One major difference in the handling of skills between Telharis and Pathfinder is that I have eliminated skill ranks. Instead, skill training will be part of the tiered talent structure. The first tier will grant a +4 competence bonus and unlocks trained-only uses of the skill. Additional tiers grant higher competence bonuses and various abilities. For example, high-tier Stealth might enable a character to actually turn invisible for short periods, or high-tier Perception might grant enhanced senses.

One final change is in terminology. I’m not fond of the term DC (difficulty class). I want many of my skills to utilize degrees of success, and I’d like a name that better reflects that. For now, I’m going to go with success threshold (ST); it’s clear and doesn’t conflict with any existing abbreviations.

Talents and Tiers

Today I want to discuss the heart of the character generation system—feats and other class abilities. Most of these will reflect training and skill growth in the character, and I want a name that reflects that. Feat doesn’t carry the right connotations at all. Skill makes some sense, but it already means something else. I think taking a cue from the Pathfinder rogue and calling them talents is best—it has a dual meaning of learned and inherent ability, and thus encompasses the widest range of potential talents.

Next, then, I must decide how to organize talents. To start, I’ll look at how Pathfinder handles feats and class features. Feats in Pathfinder are arranged haphazardly; some stand alone, others form short progressions, and some are arranged into complex interconnected trees. Some modify or enhance class features, while others are relatively generic. While flexible, this approach is inelegant and the vast selection of feats can be overwhelming to new players. Conversely, class features come in simple progressions, with the exception of occasional stand-alone abilities. They improve at regular intervals, anywhere from each level to every sixth level. Those that improve at every level are limited to simple numerical bonuses, such as base attack bonus and base saves; these are covered by the experience bonus. Spellcasters gains access to a new spell level every other level, but even in early versions of this rewrite I wanted to reduce the number of spell levels to seven or even fewer; this would slow spell level acquisition to every 3rd or 4th level, in line with the majority of other class abilities.

This then suggests to me a system where all abilities are scaled by either three- or four-level steps. The steps can be given a universal set of names, similar to the Pathfinder Basic/Improved/Greater feat naming convention, or the 4th edition D&D Heroic, Champion, and Epic tiers. Actually, I rather like the concept of tiers—the levels of talents will be identified by the tier names, and they can be used as a general measure of power level. A three-level tier system could start at level 0, and have seven tiers bringing it to level 20. A four-level tier system would start at level 1 and have five tiers to bring it to level 20. The first would give me seven spell levels, which I like, but I think would prove problematic for most other abilities; so I’m going to go with a four-level progression, and adjust spells as necessary.

So I have five tiers which need names and general descriptions. One of my main concerns in tier definition is making sure that non-player characters are well represented. Without classes, I need a way to build relatively unskilled NPCs—commoners, guards, craftsmen, and the like. My immediate thought is to reserve the first tier for these types of characters. Player characters could typically begin at the second tier, with their first-tier talents encompassing abilities such as weapon and armor proficiencies, good save progressions, and extra hit points. Next, I want to establish roughly where the limit of ‘real human ability’ is. This is very useful for calibrating skill DCs and in defining what abilities should be available in a given tier. In my opinion, real-world people can be sorted roughly into three groups—typical people, extraordinary people, and historic individuals—so I will use the first three tiers for ‘realistic’ characters. Now that I know where the boundary lies, I can go about naming my tiers.

To start, I’ll make a list of potential descriptors: apprentice, novice, journeyman, expert, master, heroic, mythic, legendary, champion, epic. Heroic stands out to me as a good name for the fourth tier, just past the limits of realistic human ability. Novice, expert, and master sound nice for the first three tiers. I think I like mythic for the fifth tier; legendary and epic could be good descriptors for a sixth and seventh tier, should I want to expand the progressions further.

To summarize, each talent will have a progression through the tiers. In order to select a talent, a character must have already taken the lower-tier versions of that talent, and must be of the appropriate tier themself; for example, a 5th-level (or higher) character with Novice Toughness could take Expert Toughness. The tiers are defined roughly as follows:

Novice: Typical people, such as farmers, craftsmen, and regular soldiers or guards.

Expert: Particularly well-trained or talented individuals, such as elite troops or master craftsmen. Characters of this tier typically influence a town or small region.

Master: Characters of this tier represent the peak of real human abilities. Such characters leave an impact on history. Real-world examples include Genghis Khan or Leonardo da Vinci.  

Heroic: Characters that bend the limits of possibility. Examples include ancient heroes such as Achilles or Hercules, or the protagonists of most action movies.

Mythic: Characters of this tier have the power to affect entire worlds. They perform the impossible on a regular basis.

Finally, I’ll leave you with a sample talent.

Precise Shot

Novice: When using the Focus action with a ranged weapon, you receive an additional +1 to hit.

Expert: You can make ranged attacks into melee with no penalty.

Master: You can attempt disarm, sunder, or trip maneuvers using your ranged attacks. Your attack bonus for such a maneuver is equal to your Awareness + experience bonus + size modifier - 4, plus any bonuses you receive with the weapon used to make the attack.

Heroic: Your ranged attacks ignore anything less than total cover or concealment.

Mythic: As a double action you can make a single ranged attack that ignores armor.

Attributes: Application

Last time I talked about the mathematical side of attributes; now I want to discuss the descriptive side. Attributes or ability scores are featured in many RPGs, both on the computer and tabletop. An ideal set of attributes should encompass all aspects of a character that are relevant to the game in question. That said, some amount of restraint is also required for the sake of playability. A system with a large number of very specific attributes has added a great deal of complexity with relatively little benefit to the play experience. For comparison, I’m going to summarize the attribute system of several RPGs:

D&D: Strength, Dexterity, Constitution, Intelligence, Wisdom, Charisma

GURPS: Strength, Dexterity, Intelligence, Health

Elder Scrolls: Strength, Agility, Speed, Endurance, Intelligence, Willpower, Personality, Luck

Diablo II: Strength, Dexterity, Vitality, Energy

Fallout: Strength, Agility, Endurance, Intelligence, Perception, Charisma, Luck

So it appears four to eight attributes are typically considered suitable to define a character. However, it should be noted that of the two four-attribute systems, one is based entirely on combat (Diablo) while GURPS allows detailed fine-tuning of ‘secondary attributes’. I only want one set of attributes and Telharis will include non-combat situations, so I’m probably looking at six to eight attributes.

What then should these attributes be? I find it helpful to think in terms of physical and mental attributes, so I’ll start with physical. All of the systems I’ve listed share similar physical attributes: Strength, Dexterity or Agility, and Constitution or Endurance. Elder Scrolls makes the additional distinction of speed, but I’m going to stick with the Pathfinder treatment and have speed be something inherent to your race. An argument could be made for separate Dexterity (fine motor skills) and Agility (reflexes and coordination), but I think that might be getting too detailed. I’ll return to it later.

Mental attributes are where things get more complicated. Intelligence works well; I don’t feel any need to make changes there. Representing other mental traits is more difficult. In D&D, the exact meaning and application of the Wisdom and Charisma scores has long been problematic. The best suggestion I’ve seen has been to define Charisma as both force of will and personality, making it the key attribute for Will saves, while moving ranged attack bonuses from Dexterity to Wisdom, since Wisdom includes senses and intuition. This equalizes the relative importance of the attributes, but introduces new problems—why should clerics be good at archery, and what makes the fast-talking rogue so resistant to enchantment? My solution is to divide Pathfinder’s Wisdom and Charisma into three new attributes: Willpower, Charisma, and Awareness. Willpower is the character’s force of will and mental resilience—a mental equivalent to Constitution. Charisma describes personality and leadership skills. Awareness, meanwhile, is sort of a hybrid mental-physical attribute. It incorporates all aspects of a character’s senses and situational awareness.

Thus, we come to the following definitions:

Strength: physical strength and knowledge of how to use it.

Dexterity: agility, reflexes, and fine motor skills.

Constitution: physical resilience.

Intelligence: memory, learning ability, cleverness.

Willpower: mental resilience.

Charisma: personality, leadership potential.

Awareness: senses and intuition.

And an approximate list of what each applies to:

Strength: melee attacks, carrying capacity, skills (Climb, Swim)

Dexterity: combat defense, Reflex saves, initiative, skills (Acrobatics, Escape Artist, Ride, Sleight of Hand, Stealth)

Constitution: hit points, Fortitude saves, skills (Endurance), combat abilities (e.g. rage)

Intelligence: skills (Heal, Lore, Profession, Security, Spellcraft), some types of magic, possibly influences feat/talent acquisition

Willpower: Will saves, skills (Use Magic Device), all types of magic

Charisma: skills (Deception, Handle Animal, Intimidate, Persuasion, Streetwise), leadership abilities, some types of magic

Awareness: ranged attacks, Insight saves, initiative, skills (Perception, Sense Motive, Survival)

There are still some minor issues here; the three mental scores are now spread a little thin, particularly Willpower. However, they are each critical for characters pursuing certain skill sets, such as magic or leadership abilities, so that might be okay. As with my original rewrite attempt, my ultimate goal is an elegant and intuitive ruleset. If something doesn’t make sense, please let me know and I’ll try to explain and/or fix it. 

Attributes: Scale

In this installment I’m going to talk about attributes. Telharis attributes are the same as Pathfinder ability scores in their purpose and use, but differ in several key ways. First of all, there will be no separate score and modifier; you simply have an attribute and add it directly to any relevant rolls. The odd ability scores in Pathfinder were basically empty space; why use two numbers when you can use one? There are then two choices to make—what is considered the base or average score and what is the minimum score?

Basing the system around 0 as the average has the obvious advantage that it’s fairly intuitive; someone who’s bad at something gets penalties and someone who’s good gets bonuses. However, the worst possible score is then some arbitrary negative number, and that seems rather inelegant. The alternative is that the scale starts at 0 and only goes up: everyone gets bonuses, but those who are good get bigger bonuses. This has some potentially problematic implications for some of the system math (particularly damage and hit points), but I think those can be dealt with.

With the minimum score established as 0, we must now determine what constitutes an average human score. Two options jump out at me: 5, which maintains the status quo, or 10, which increases the distance between the minimum score and the human average. In the context of skill checks, a difference of 10 is the difference between reliable success (while taking 10) and complete incapability. This seems about right for the difference between an average person (10) and someone with a crippling disability (0-2). For a reality check, let’s look at some actual skills. Based on the Pathfinder rules, the average person should be able to climb “a surface with ledges to hold onto and stand on, such as a very rough wall or a ship’s rigging”. That sounds like something that someone with minimal strength (the key attribute for Climb) would not be capable of doing. According to Pathfinder, an average person should also be able to hear a creature walking within 10 feet. Someone with severely impaired senses would probably struggle to do this. Finally, I want to compare the effect of starting attributes vs. bonuses gained from leveling and training. Characters gain between +1 and +2 to skills per level depending on how heavily they focus on them. This means that, using 10 as a base, a level 6 or 7 character with a crippling attribute can perform actions based on that attribute as well as an average level 1 character. For comparison, in a base 5 system this would happen around 3rd or 4th level, which feels a little too early to me.

So, to clean all that up: Attributes range from 0 (crippling), to 10 (average human), and technically have no upper limit, though scores above 20 should be highly unusual. The vast majority of humans have scores in the range 5-15, with typical scores maybe even more tightly constrained.

More on attributes next time.