Wizards HATE Her! How to Play D&D for Free, Part 2: Player Interface

Did you miss the last post? (link)

Player Interface

I don’t think it’s unfair to say that the overwhelming majority of content published for playing games (that aren’t adventures?) are player-facing rules for game interactions, i.e. rules for characters and how to use them as sort of your avatar in the game-world. I think people really overcomplicate it when it's probably the easiest thing to wing, provided that you have a structure simple enough to wing it.

Abilities as Heuristic Frameworks

Typically, in a D&D, your character is going to have scores that represent their capabilities with respect to different things. The traditional six ability scores are strength, dexterity, constitution, intelligence, wisdom, and charisma. These are the ones everyone knows, even if opinions on them vary. I don't think these are important specifics, and that the traditional six serve as a fine basis for the sort of fictional world supposed by D&D as a cultural thing. However, it's worth considering what exactly the point of ability scores is. Why do we take it for granted that characters, as player interfaces, have them?

I've discussed this before in greater detail [1], but ability scores have categorically different functionalities across different versions of D&D. This isn't just with respect to how they're applied, e.g. now dexterity applies to armor class whereas it didn't used to, but what their purpose is within the scope of the game. In the original D&D, you have three ability scores whose main functionality is to modify how quickly a character will level up as one of the three classes (fighter via strength, magic-user via intelligence, cleric via wisdom). The other three ability scores have pretty different applications. Constitution gives your character more or less hit points per level. Dexterity modifies your missile-shooting capability, and it is said that it can help decide whether one character moves before another (though no specific advice is given for how to do this). Charisma informs how many special units you can have in your service, and how loyal they will be to you.

In later D&D, your character's numbers became more central to how well they could perform in the game. By Third Edition, or maybe earlier than that, there was a single scheme for converting scores to bonuses which you could add to dice rolls. Most commonly, this is for ability checks or skill checks where your character is attempting something difficult or dangerous. You look at the situation and figure out which of the six abilities is most relevant; then, you add your bonus (or minus) to a die roll to see whether your character succeeds at the attempt, and you're trying to get a total that's equal to or greater than some target number. Ability bonuses also apply to damage rolls in combat; for example, you add your strength bonus to melee weapons or your dexterity bonus to ranged weapons (just like you add these bonuses to the attack/to-hit roll). This is a far cry from the original D&D where combat capabilities were determined entirely by class, with little if any differentiation between characters of the same class.

Many rulesets that call themselves 'old school' actually tend to use ability scores in the latter sense. On one hand, this is because it's how we're used to using ability scores since Third Edition, and so that's most people's expectation for how to use them. Whether you do it the old way or the new way speaks, in this context, to whether you want less or more mathematical differentiation between characters. You can imagine that it's handy to have all clerics do the same math for their attacks, as in OD&D; yet it's also handy that, if players are the ones making their rolls, they should have the information necessary at their fingertips and it should be information they have control over. How much do you, as a player, want to deal with the math? Some people really like it.

On the other hand, there is a world of freedom opened up by ability scores in general, and this is sort of what later D&D editions have realized in their rulesets. Ability scores (and bonuses) don't have concrete or specific use cases. What I mean by this is that you don't have specific numbers for Melee Attacking, Horse Whispering, Flirting, et cetera. You have instead six categories of contexts where your character might act in those capacities, and you (and/or the game master) decide which context works for the current situation. This allows the use of ability scores to be very flexible and extendable, and it's what helps D&D be a game with an open set of interactions rather than one with a hard-coded, closed set. It's part of what facilitates your character being able to do or try anything, because the ability scores are heuristics for how well your character will perform in broad contexts rather than at specific actions.

I think this is why there's the popular conception that D&D Fifth Edition is easy to learn; I agree 100%. The rulebook sucks, sure, but all you need to know is your character has six numbers, and every time you roll (i.e. for every single roll) just one of them will apply. Any other circumstance is handled as rerolling the die rather than doing more math to it. It's easy and it's intuitive and, by its nature, it's infinitely extendable to any situation. This is what I think people mean when they say they don't want to learn another ruleset than D&D Fifth Edition to play whatever scenario they're considering. What other people don't realize, I think, is that these players have found a flexible and intuitive interface for them to play whatever they want. It’s not about anything else Fifth Edition has to offer, to speak broadly. However, this is not because of D&D (though WOTC certainly relies upon this brand relationship to make money now), especially considering that the official materials aren't very helpful, but because people have learned an easy ruleset very quickly, often learning from each other [2].

In Practice, What I'd Do Differently

I mentioned that when I played with my friends, I used a Frankensteined ruleset whose “player interface” was built mainly off of Knave. You have the typical six ability categories, and you have two types of numbers attached to each: bonuses and defenses. Bonuses are determined by rolling three six-sided dice and picking the lowest, doing this for each ability. Each ability defense is equal to that ability's bonus plus 10. Players use bonuses to add to their rolls, whereas the game master uses defenses as numbers to roll against on behalf of monsters et cetera.

Let's consider combat as an example. Besides the six abilities, characters also have an armor pseudo-ability which is designated by a bonus and a defense, the latter defined by the former. Let's suppose that a player's character is attacked; there's two ways of handling this in Knave. The traditional way is that the game master rolls d20 and adds the monster's attack bonus, and the goal is to reach a total equal to or greater than the character's armor defense. Alternatively, players can be the ones to make all the rolls. The player can roll d20 plus their armor bonus, and try to reach a total equal to or greater than the monster's attack defense. The two methods are mathematically equal; the only thing that changes is who rolls and who has a static number.

I think this works fine under the hood, but it's a bit extraneous for players. At least, I felt like my friends were a bit confused at having the two numbers, and we didn't use defenses all that often except just for armor (in which case, it also feels weird to have armor bonuses). It also felt like an extra layer of complexity to character creation and character sheets.

Like 5E, but Without Scores

You could do 3d6 rolls for each ability score, or some variation thereof, but when those scores apply to every possible roll you can make (keep in mind that this wasn't the case originally), it can feel unfair. This is probably why D&D Fifth Edition has a standard array (more like a set?) of ability scores which players can assign to their characters' scores at will, though then I don't like how you have both scores and bonuses where you rarely if ever use the former. I did some math and it looked like given every possible combination of the standard array and ability bonuses from character race, characters are going to have the following optimal sets of ability bonuses:

• {-1, 1, 1, 1, 2, 2}
• {-1, 0, 1, 2, 2, 2}
• {-1, 0, 1, 1, 2, 3}
• {0, 0, 1, 1, 2, 2}
• Outlier: {-1, 0, 1, 2, 2, 3}

What this looks more like to me is that you distribute 6 points between your bonuses, at a maximum of +2, and then if you reduce any one bonus to -1 then you get an extra point to distribute at a maximum of +3 per category. You can make it simpler and just say distribute 6 points at max +3, and make one bonus -1 for an extra point somewhere.

I think this would work very well if (1) you want to use bonuses as a catch-all rule, (2) you don't like overly random generation, and (3) you want a heavier hand in character creation as opposed to character generation. It also encapsulates the mathematical effects of the questionable racial bonus, so now you can just apply bonuses wherever and if you want to justify your numbers on being an elf then that's on you. What if, though, you want more randomness but not too much randomness? Let's say, for example, your campaign is a typical dungeon crawl where you don't care about characters dying too much (them serving more as an interface or avatar for yourself than as a character, and also more as an instrument of the party as a whole).

Bounded Random Abilities

So, the effect of using standard ability bonuses is that everything is already kinda bounded. Let's say you wanted to convert ability bonus rolls to statistically equivalent roll-under ability checks, where let's say someone with a bonus of +0 is equivalent to someone with an ability score of 10. By defining an ability bonus as, basically, dividing an ability score in half (and normalizing the average to 0), you're dividing the range of success probabilities in half too, with respect to one target number. The equivalents are [3]:

• -4 (Score of 3) → roll ≤ 6, or ≥ 15 (30%)
  
• -3 (Score of 4-5) → roll ≤ 7, or roll ≥ 14 (35%)
  
• -2 (Score of 6-7) → roll ≤ 8, or roll ≥ 13 (40%)
  
• -1 (Score of 8-9) → roll ≤ 9, or roll ≥ 12 (45%)
  
• 0 (Score of 10-11) → roll ≤ 10, or roll ≥ 11 (50%)
  
• +1 (Score of 12-13) → roll ≤ 11, or roll ≥ 10 (55%)
  
•  +2 (Score of 13-14) → roll ≤ 12, or roll ≥ 9 (60%)
  
• +3 (Score of 15-16) → roll ≤ 13, or roll ≥ 8 (65%)
  
• +4 (Score of 17-18) → roll ≤ 14, or roll ≥ 7 (70%)
  

So, that's the nice thing about how it already works. But, if you determine each score randomly, you can still get screwed by bad luck. What if we bound the random generation itself, in total, and guarantee fair distributions without oversight (though while still guaranteeing characters have things they're bad at)?

3d6 is an iconic thing, so let's keep that. Roll 3d6 and look at which abilities are indexed by the numbers. Subtract -1 from each of those ability scores. Then, roll 3d6 again and add +1 to each of those ability scores. Finally, let's say you can add +1 to any score of your choice, just for fun. [4]

1. Strength
2. Dexterity
3. Constitution
4. Intelligence
5. Wisdom
6. Charisma
   

The effect is that the sum of all your ability bonuses is +1. The lowest one can be, if you're unlucky and decide not to fix it with the extra point, is -3. The highest one can be is +3 randomly or, if you go all-in, +4 with the extra point. You can refer to the equivalent roll under/over scores above to see that this is a pretty fair distribution.

However, this depends on you having those six ability scores to roll d6 on. Maybe you're fine with that or maybe you aren't. I still hope you can see how this isn't a difficult thing to wing so long as you know what you're going for. The goal is ultimately, if you want light-weight player-side math, to figure out something which can be applied to infinite contexts and which is easy to grasp without modifications or calculations made on the fly.

Character Powers

I think people like having unique powers or abilities (not in the same sense as above) that set apart their characters from others. In Fifth Edition, your selection of such powers depends on your class, meaning that the function of character class is to select a set of possible powers rather than to stringently classify characters as one group or another (so you can make tables indexed according to class, for example, for combat rolls or saving throws). Now, I don't like this, and I also don't like how many complicated choices there are in Fifth Edition or in similar rulesets. I don't like having a big character sheet that I have to read like a menu of things my character can do. It's complex and, often, restrictive in that formalizing abilities often prevents anyone from trying them even when it makes sense for anyone to do. It also makes the set of interactions feel more closed, since I'm relying on my sheet for knowing what to do rather than imagining what my character can do as a real person in the game-world.

For our home game, I meshed together the racial and class abilities from Here's Some Fucking D&D with abilities I found from a blog post called "Knacks for Knave" [5]. Players picked one racial ability for their character as their "Level 0" knack; humans, rather than having a specific ability, could pick one learned ability as their Level 0 knack. Then, players picked one learned ability for their Level 1 knack. Each level, you get another knack. There's the risk that you might get drained of levels by an undead monster, but your Level 0 knack is the one that always stays. This is what gives humans a sort of advantage, I think.

The racial knacks were as follows (and some of these weren't used, so who knows how they would turn out), with many of them being derived from Here's Some Fucking D&D:

• Dragonborn: Breathe 1d6 acid, cold, fire, or lightning damage.
• Dwarf: Have sonar vision in the pitch black dark.
• Elf: Heal at midnight without sleeping, and you experience no exhaustion. 1 free spell per day.
• Gnome: Understand and speak to animals.
• Halfling: Reroll results of 1, and add +2 to dexterity checks while being sneaky.
• Human: [Pick any learned knack.]
• Kenku: Perfectly mimic any voice heard in the last 24 hours.
• Orc: Advantage on saves against psychic attacks.
• Tabaxi: Always land safely on your two feet.
• Tiefling: Hands burn for d8 fire damage.
  

You can look at the "Knacks for Knave" article for what I used as the learned knacks, except that I changed the magic-oriented ones to not use the magic dice system from GLOG (since I thought it would just be one more thing to learn, in contrast to Knave's simpler "one cast per spell book") [6].

I think that this is a really nice and freeform way to incorporate what people expect now from class powers in later D&D, and without much additional complexity. It makes it easy to say, "I'm an elven bard", without the sort of mechanical baggage you have to deal with in D&D as written. I think it's also more intuitive than starting with just one knack, as suggested in the original post. Having just one knack doesn't actualize the character differentiation often desired, but with two you can actually start having concrete character concepts whose components synergize in interesting ways.

Objectifying Characters

Often you'll find people advocating for totally non-mechanical character records. Instead of quantifying how smart or strong or sexy a character is, you can just say "My character has kissy lips, but is also sort of a pushover." This is a popular approach in the FKR scene, which advocates for minimizing the 'visible' parts of the player interface. You only know the diegetic aspects of your character, not the numbers that make other numbers work. The game master is the one who runs that side of things. It necessitates simpler interfaces per character since there's ultimately one person handling all the characters, so either you need very simple character records (i.e. quantified attributes) or totalizing character classifications (as in original D&D). It can't be too finicky or having too many formal customizable or character-specific details (i.e. as opposed to diegetic character descriptions).

However, people seem to like the quantified character attributes that Fifth Edition uses. I see plenty of people who turn their original characters ('OCs') into D&D ones, or people who start with Fifth Edition as the very basis of making their original character. You can chalk this up to the cultural popularity and prevalence of Fifth Edition as a nerdy thing, but I also really think that people like the sort of objective definition of characters that Fifth Edition affords. It makes characters graspable, at least to people in the know, and also comparable between other characters who are defined using the same 'language'. I can easily, easily imagine a situation where someone playing Wanderhome recasts their character as a D&D Fifth Edition one for the sole benefit of lending their character an objective reality, even if that is never brought to the table playing Wanderhome. Again, we can criticize WOTC for profiting off of an intellectual property that has basically become common, but having that common creative basis is not a bad thing per se. At least, it shows that having an objective basis can be desirable and not unwelcome.

Now, having simple character powers gives us the best of both worlds. They do not have to be defined formally, i.e. written to interact directly with specific rules of the game. In fact, it might even be worse if they were written formally; it would make them less creatively applicable (not that it wouldn't be fine to just say "my character is good at fighting, so they get +2 to attack" or whatever). However, since they most likely interface with the character's ability numbers in practice, they already have at least an indirect basis in the objective reality of the game-world. This makes them more intuitive and less of a hassle to read or keep in mind, but it also makes it easy to just come up with new ones. My partner, for an ancient world D&D game, made a character who was a wine genasi (i.e. a wine elemental). It was not difficult to treat her character as a water genasi in all respects, except substituting wine for water whenever it came down to the rules. It'd be as easy to phrase this as, "My character is a wine elemental, and can shape wine with her mind." The specifics can be negotiated in action.

At this point, it's the premise of the game that should be taken as the premise for possible characters that can exist in the game-world. For party games or typical dungeon fodder fare, I don't think it's unfair to use random tables to quickly generate characters; besides, it's fun to play as someone unexpected, and it's a hassle to have to come up with a new 'Original Character' when your current one dies. You see, a lot of this depends on what you want or expect from your session or campaign, and none of these factors are universals that should be taken for granted. They're, obviously, either expectations you and your friends hold in common or ones that you should take care to establish before playing together.

 

[1] https://traversefantasy.blogspot.com/2021/10/kinetic-and-potential-abilities-across.html

[2] Under "Moving Forward": https://traversefantasy.blogspot.com/2021/09/critique-4-catching-up-to-speed-forge.html

[3] Though, keep in mind that a DC 10 check in D&D Fifth Edition actually has a 55% chance of success with a +0 bonus, or a 60% chance with a +1 bonus (which is statistically more likely). This means you can increase the roll-under target numbers by 2 or decrease the roll-over ones by 2, which improves all likelihoods by 10%.

[4] For a solution with equivalent outcomes to the previous, you need only roll 6d6 and treat all as +1 rather than treating half of them as -1. You can even reroll dice if it turns out you get more than three applying to the same ability. However, I think 6d6 is a lot of dice to roll at once. That's my only qualm.

[5] https://themanwithahammer.blogspot.com/2019/11/knacks-for-knaves.html 

[6] Something interesting might be to let there be multiple spells per spell book? Each book can still only be cast from once, but you can cast any one spell that is in the spell book. Players might then be incentivized to copy spells between spell books to have the security of redundancy, in case they lose one book or in case they might need to cast one spell more than once. Each spell book is a lot to carry and has to be, like, swapped out from your backpack if you need to switch between them.

You can even say scrolls have just one spell, but books have up to six spells or something. It'd be worth copying spells from a scroll into a book just to reduce how much you're carrying, though at the expense of only being able to cast one spell from that book per day. You don't even need a limit of spells per book, since I think being able to only cast one is its own self-regulating restriction. This basically makes spells like fidgety single-function devices, and spell books like reprogrammable devices with bad batteries.