DEV BLOG Gebnar's Mechanical Madness: Behind the Dice

Discussion in 'Myth Maker News' started by Gebnar, Feb 13, 2016.

  1. Behind the Dice
    Today I'm going to discuss Myth Maker's core dice mechanic. I'm going to jump right into the probabilities first, and discuss the design decisions that led to them afterward. All of the probabilities are fractional percentages, so I've rounded them to the nearest whole number for simplicity's sake.

    Presenting Probabilities
    I'll be expressing all of the percentages in the article like this:
    • Success% (Mastery%)
    Success% is the total chance a character has of succeeding at a task, including mastery. Mastery% is the total chance a character has of mastering a task.
    For example: 72% (8%) means that out of 100 attempts, about 28 will fail, about 64 will succeed without being mastered, and about 8 will be mastered.

    Basic Math
    As you probably already know, characters in Myth Maker have three core attributes: Mind, Body and Spirit. These attributes (when undamaged) can range from 6 to 10, and are added to a 2d6 roll to determine the success, mastery, or failure of a task. When thinking about attributes, we can associate the range of each attribute to how able a character is in that area. Here are the probabilities for an untrained character:
    Attribute       Success (Mastery)
    6 (Pitiful)      28%      (0%)
    7 (Mediocre)     42%      (0%)
    8 (Average)      58%      (3%)
    9 (Good)         72%      (8%)
    10 (Excellent)    83%     (17%)
    The Takeaways: As you can see, characters are more likely to succeed than they are to fail. This is backed up by the attribute assignment system, which limits a character to only one attribute of 7 or 6. In fact, if you run the math for all of the possible attribute arrays, the average success chance across all three attributes is always about 62%. Myth Maker is a game where characters are supposed to do things. It's not fun to roll repeated failures, especially for things your character is supposed to be good at. We feel these probabilities allow a character to be meaningfully weak or strong in one area, without hindering their overall potential.

    So how effective is that +2 from training? Well, here's the math:
    Attribute       Success (Mastery)
        6             58%      (3%)
        7             72%      (8%)
        8             83%     (17%)
        9             91%     (28%)
       10             97%     (42%)
    The Takeaways: As you can see, training makes a huge difference regardless of a character's attribute score. Even a pitiful attribute becomes success-favored with a little training, and even gets a small chance to master a task. Stacking training with strong natural talent not only makes success extremely reliable, but drastically increases the chance of mastering a task. Again, if your character is supposed to be good at something, the probabilities should (and do) back that up. You won't see a master swordsman running around missing attacks left and right in this game. Unless of course, he's opposed...

    Being Opposed
    Let's assume the master swordsman has 10 body and Sword Training. Here are the odds of the swordsman hitting an opponent who tries to parry him. These percentages include the chance of mastering the parry, which makes the swordsman automatically miss.
                              Untrained             Trained
    Opposing Attribute    Success (Mastery)     Success (Mastery)
            6               88%     (33%)         75%     (23%)
            7               83%     (28%)         67%     (17%)
            8               75%     (23%)         58%     (13%)
            9               67%     (17%)         48%      (9%)
           10               58%     (13%)         37%      (6%)
    (If you want the complete picture for how effective opposition is at various attribute levels and training, I've attached a text file of many tables.)
    The Takeaways: As you can see, opposing a task can be reasonably effective even for untrained individuals. This isn't the whole story though. Some talents allow characters to boost checks, which turns any success into a mastery, unless the task is opposed. Opposing a boosted task is almost always a good idea, even if you're not likely to prevent it.
    For example: Joe Moe has 6 body and no training. If the master swordsman boosts his attack, he has a 97% chance to critically hit Joe. If Joe tries to parry the attack, he'll probably still get hit (88%), but he's FAR less likely to be critically hit (only 33%).

    Attribute Damage
    Sometimes characters get hurt. We tried many different injury systems in the past, but we don't want to force players to deal with any more gore and grit than they feel comfortable with. We decided to just let injured characters attributes speak for themselves. Here are the basic success rates for attributes 5 and below. Mastery is not possible at these attribute scores, even with training.
                   Untrained    Trained
    Attribute       Success     Success
        5             17%         42%
        4              8%         27%
        3              3%         17%
        2              0%          8%
        1              0%          3%
    The Takeaways: Characters who get hurt show it by being unable to do things that would have otherwise been a cakewalk. Even minor injuries can be crippling. We like this because characters don't have to be killed to be defeated. In fact, it gives characters a really good sense of when to flee or surrender, since there's little chance of wining once you get hurt. Mind and Spirit damage will probably be a lot less common than Body damage (depending on what you do with the game), but they'll be no less crippling.

    One thing we wanted to avoid with our core mechanic was a scalable difficulty number. But we still had to include a way for players to modify the difficulty of a task to account for circumstances that arise in the fiction. To do so, we add a third die and use the higher or lower two dice. I've added Attributes and Training together to cut the number of tables in half. Here are the probabilities:
    Attribute           Easy                Standard                Hard
    +Training     Success (Mastery)     Success (Mastery)     Success (Mastery)
        6           52%     (0%)          28%      (0%)         11%      (0%)
        7           68%     (0%)          42%      (0%)         19%      (0%)
        8           81%      (7%)         58%      (3%)         32%      (0.5%)
        9           89%     (20%)         72%      (8%)         48%      (2%)
       10           95%     (36%)         83%     (17%)         64%      (5%)
       11           98%     (52%)         91%     (28%)         80%      (11%)
       12           99.5%   (68%)         97%     (42%)         93%      (19%)
    The Takeaways: There are two stories being told here: At the lower end of the attribute+training scale, success rates are affected dramatically by the circumstances. At the upper end of the scale, success isn't as much of an issue, but mastery is heavily influenced by circumstances. As with standard checks, being opposed makes a huge difference for easy and hard checks.

    I hope you enjoyed learning a bit about Myth Maker's action resolution mechanic. It took us years to nail down a core mechanic we were really satisfied with. We evolved a number of design goals over the years, and this mechanic satisfies them all:
    • It's Success-Favored.
    • It's produces results on a curve (some of the time it's actually a triangle...).
    • It uses friendly numbers. (Attributes range from 1 to 10. Only one difficulty number to remember, and it's a nice round 15? Sign me up!)
    • It has narrow scaling. (There aren't a huge number of probable outcomes, so we can design a tighter system.)
    • It's easy to use. (There aren't a lot of numbers to keep track of, and they're small enough to process easily in your head.)
    • it's "game-able". (We want all kinds of gamers to enjoy Myth Maker - even the ones who like to "game the system" to their advantage. As you can see above, stacking a high attribute, training, and favorable circumstances can virtually guarantee success. Of course, that's the kind of thing people do in real life, so it shouldn't be much of a stretch when game characters do it too.)

    Attached Files:

    #1 Gebnar, Feb 13, 2016
    Last edited: Feb 15, 2016
  2. For anyone interested, I made a small AnyDice program that can calculate all of the above probabilities, and quite a few more. If for some reason the link breaks, I've also attached a text file containing the code.

    To play with the program, edit the four variables at the top, and click the calculate button.
    There are comments explaining what the variables do, and their default values.

    The results are expressed as two tables, Success and Mastery. The "1" row in each table is the number you're looking for.

    Have fun!

    Attached Files:

Share This Page