Tasks and skills

 I have thought a lot about this while waiting for Eero to write up his CWP and thought I might put mine in the ring first. A lot of it is based on the stuff I've seen in the RPG Theory discord, but not all of it.

Some small preparatory notes. The skill system, whatever it is, and the task "system", are totally separate. Their formalisms are just not the same. The task "system" is not a system at all, it is a vocabulary with which you can construct individual task resolutions. Various tasks are so different it would be nonsense to treat them all the same.

Simple skill system

In the simple skill system, you either have a skill or you don't. What the skill is is completely free-form; anything you can be good at is a skill. At any point, you can declare that you had a skill the entire time, if the skill is consonant with your background and current situation.

If you want to train a new skill, go train it, for however long that would take. Maybe the skill requires certain stats or resources to train, and you can't train any without them. That's fine.

If you ever need to really nail down what the skill does, make a writeup for the skill, something like the following:

Forestry. 6 months to learn, requires Int or Wis 12+ or a wilderness background. Allows tracking, navigation, foraging, and hunting in coniferous forests, and tracking and navigation in most other biomes.

If you want to spice things up, add a single adjective to your character's skill, representing their skill level: beginner, intermediate, advanced, fully competent, et cetera.

Complex skill system

Keep the skill writeup from before. Every character with a skill should put a number next to it, between 1 and 100. Higher means greater familiarity with the skill.

Skill levels. Stolen verbatim from Eero

Don't start with the number, start with the skill level, and use the appropriate number.

Skills can't increase above 100. When they get to 100, you can write down successor skills, things that are more complex and require prior familiarity. So for instance most adults have Chess 20, no better than a child who has learned the rules. If you get to Chess 100, you're a decent player and you can wipe the floor with the average slob. But your journey has just begun; now you can take skills in opening, midgame, and endgame. And after you've hit Chess Openings 100, you can start studying Bongcloud Attack. And so on.

In every session that you use a skill, increase it by 1. (If you have a specialist character class in your game, every session they can distribute their level in % between all the skills they used, too. If you keep track of martial arts as skill, fighter-type characters can do the same, but just with martial arts.)

You can still advance skills via training, obviously. We just need to figure out how long it takes to get from basic competence to advanced competence, or whatever. Again, worry first about the skill level, and then after you've figured it out, match up the skill level to a number.

The time it takes to train from "ignorant of the skill" to "some practice", say, does not have to have anything to do with the time it takes to go from "some practice" to "basic competence". Anybody can pick up a bow and start shooting in the general direction of a target within 15 minutes, and after an hour or two of practice, they'll be hitting the target occasionally. So they went from 0 to 20 in about a day. But it takes several days or more to get to basic competence (50), the point where you reliably hit the target but aren't scoring bullseyes. And longer than that to get to full competence. (I don't actually know how long, because I never made it past basic competence, and that was many years ago.)

I said above that skills might require certain resources or ability scores. They might also grant resources or ability scores. If a character goes to the archery range every day for a few hours, that should count as strength training, pumping them up a bit, and it should give them a few casual social contacts in the archery community.

Task construction vocabulary

Ok, on to tasks. When a character is trying to complete a task, ask a few or all of these questions, in no particular order:

• What resources, including time, will the character require to even attempt the task?
• Is this the sort of task affected by randomness, complexity, and incomplete information?
• Does this task depend on the raw ability of this character, and if so, how much?
• Does this task require specialized knowledge to complete?
• Is this one atomic task, or several? Do we want the higher resolution from breaking it down into several tasks, or do we want the abstraction of a single task roll?
• Is this task affected by prior linked tasks? Will its success or failure affect future linked tasks?
• Is success on this task a matter of accumulating individual successes, or is it binary?
  

 is this something where randomness, complexity, and incomplete information will have an effect? Or is it the sort of thing where we can simply figure out what is going to happen?

Fortune is a task-resolution method using randomness, most often a die roll. For instance, attacks are usually resolved with fortune -- roll a die, hit a target number.

Karma is a task-resolution method using some static value. For instance, in a tug-of-war, the heavier side -- the side with the most static friction -- nearly always wins. (There may occasionally be some strategy, but it's not likely in most games of tug-of-war.) So if you want to find out the winner of a tug-of-war, just figure out which side is heavier. Similarly for arm-wrestling -- just see who has the higher strength, and they'll be the winner.

Karma-based tasks are usually resolved by either direct comparison or some short calculation. You can do that on your own. The rest of this vocabulary concerns fortune.

When you're making a task check, the usual first step is to ask, what are the odds that a normal person under normal circumstances would succeed at this? Say, 2 in 3. Then you'd roll 1d6 and succeed on a 4 or less.

You can model advantages and disadvantages through bonus and penalty dice. When you have a bonus die, in the example above, you'd roll 2d6, and if either of the dice were 4 or less, you'd succeed. When you have a penalty die in the example above, you'd again roll 2d6, but you wouldn't succeed unless both dice were 4 or less.

A few pointers for bonus and penalty dice:

• You can get multiple bonus or penalty dice, nothing restricts you to just one.
• They cancel out, so you can't have both a bonus and penalty die.
• If you can retry a task, you don't reroll all the bonus or penalty dice, you just get one bonus die added to the original pool
• Circumstances, ability modifiers, and skills are normal ways to get bonus or penalty dice
  

Skills apply unevenly to tasks. We might give you a bonus die, or several, for having a skill, or we might decide that the task needs to be reconstructed from the beginning, that it's totally different when a skilled person is trying it versus an unskilled person. For instance, an unskilled slinger -- someone who picks up a sling for the first time -- has a less than 1% chance of hitting any target at all. A skilled slinger doesn't just get bonus dice for the task, the entire task is evaluated differently. (Probably we'd ask the skilled slinger to make an attack roll.) Similarly for hacking. An unskilled "hacker" simply cannot hack anything. A normal hacker might have a 30% chance of breaking into some system given a month's time. (I'm making this up, to be clear, it's all bullshit, I don't know anything about hacking.) Maybe a really skilled hacker gets a bonus die or two, or maybe they reduce the time necessary for the task, or both.

Ok, that's the basics of it. Everything else is just a special topic, in no particular order.

Ability checks are d20 rolls under a character's ability score. Usually they aren't directly useful for task resolution, because they don't take into account the inherent difficulty of the task, just the character's raw abilities. Oftentimes you'll make an ability check to see if you can get bonus dice from a character's ability score in a normal task resolution.

Cumulative tasks are tasks where each individual success or failure counts for little -- the goal is to stack up a number of successes. Multiple degrees of success count as multiple successes. Understanding a difficult book can be modeled as a cumulative task, where the reader needs, say, 10 understanding checks, and can make 1 check per 4 hours reading. 

You might wonder, why not just keep trying the task until you accumulate enough successes? There's a few answers possible, depending on the situation: because of the opportunity cost of attempting this task versus anything else; because the task relies on limited resources (money, rare chemicals); because, based on some arcane particulars, the task will automatically and irrevocably fail after a certain number of failures.

Degree of success or failure. When you're rolling a dice pool (because of bonus or penalty dice) your degree of success is the number of successful dice (when you have bonus dice) or failed dice (when you have penalty dice). Keep track of this when you care about more than binary success or failure, you want to see how good or bad the result was.

On an open-ended roll, your degree of success is the number of times the target number can be divided by the result.

On sequential tasks, take your degree of success or failure as bonus or penalty dice, respectively, to the next task roll.

The disruption die is 1d20 optionally rolled alongside any other task roll, to model extraordinary good or bad fortune. Some unexpected good thing happens on a 20, some unexpected bad thing happens on a 1.

Exploding dice. If you roll 2d8 exploding 7, then you roll 2d8, and if either of the dice are 7 or 8, you roll an additional 1d8 for each 7 or 8, and if either of those are 7 or 8, then you roll an additional 1d8 for each... and so on. Add up all the dice at the end to get a total result. Works well if there's a small chance of a large increase in performance, like for an open-ended task.

Open-ended task. If there's no theoretical limit to you character's performance at a task, then instead of rolling under some target number, try to roll over it, with exploding dice. Useful if you're making a great work of art, for instance.

Sequential tasks need to be completed in, well, sequence. You can use degrees of success or failure from prior tasks as bonus or penalty dice, if it makes sense. Fully resolve one task before moving on to the next!

 Examples

Still to be written, unfortunately, I got too tired making this post. Planning to have a street chase complex task, low complexity, and a research complex task, medium complexity.

In the street chase, you could do simple running tasks (resolved by karma), navigation checks, try to topple shit over to slow the other party, stuff like that. Each component task has to be short, and the whole complex is resolved within initiative.

In the research task, you can look for books in libraries, try to translate books, find experts, bribe people to help you, whatever. It takes months to complete, and there's probably a few people helping out on each task.