Dice methods 2

However, in an objective, rules-heavy system, incorporated effects present some challenges for the designer. For one thing, it is harder to differentiate strength-type modifiers from skill-type modifiers, because they modify the same roll. Rolemaster has this problem; and D&D armor has it too, as Ran has said (even though D&D is not incorporated). A Champions fight of a super-strong vs. a super-fast hero has a certain feel, which sometimes appears in movies and comics, and which Rolemaster doesn't capture very well for this reason.

So I would say in a rules-light system, incorporated is the way to go. In an objective system, incorporated is cool, but you've got your work cut out for you as a designer.

RON EDWARDS: Nice to get some feedback on those archive threads. I think this list has seen some pretty high-quality RPG discussion over the last year or two.

Nelson wrote, > However, in an objective, rules-heavy system, incorporated effects present some challenges for the designer.

I definitely agree. Usually, "objective, rules-heavy" indicates a Simulationist perspective, and in my experience, most simulationist role-players really don't mind taking a long time to work out what happened in two seconds of game-time. They roll, and fiddle, and look things up, and enjoy a debate about a particular rule, and talk about similar situations in a game they played two years ago ... and then, when it's all settled, THEN they consider the actual result.

So from that perspective, incorporated effects must seem kind of like cheating: two outcomes (did-it-hit and how-hard) in one roll. It's hard for me to sympathize with this outlook -- because I really enjoy that moment of dramatic outcome in a role-playing situation, and I've always hated rolling real good to hit and then rolling real bad for damage.

Good point.

RAVEN: Would someone mind explaining to me what seperate mechanics are, and what incorporated mechanics are? I am unfamiliar with the terms (and wondering if I missed a discussion somewhere).

NELSON: Raven wrote: >Would someone mind explaining to me what seperate mechanics are, and what incorporated mechanics are?

For anyone who wants to know: Seperate mechanics use two different rolls to determine success and degree-of-success, such as "to-hit" and "damage." You can read all about it on the "dice mechanics" thread on the Sorcerer home page.

RON EDWARDS: Raven wrote,

> Would someone mind explaining to me what seperate mechanics are, and what incorporated mechanics are? I am unfamiliar with the terms (and wondering if I missed a discussion somewhere)

To clarify further: in the list archives, the thread called "Dice methods" rambles on a couple of topics before getting rolling on this issue.

Specifically, we discussed how two aspects of event resolution are related: (1) whether the BASIC TASK ROLL'S potential values have a curved or flat statistical distribution, and (2) whether the EFFECT is determined by the same task roll or by a new roll or other mechanic of some kind.

More discussion on this issue would definitely be welcome.

NELSON: on linear vs. bell curves:

This is kind of tough. Here are some things to think about:

Consider a flat system where you have to roll a target value or higher on a d20. Alexander, Bob, and Clovis, need to roll 2, 9, and 20 respectively. Now, how big of a difference does a +1 modifier (to the die roll) make to each one?

Alexander's chance of failure is cut in half, from 10% to 5%. That's a big deal. Clovis's chance of success doubles, so he's got to be happy about that. For Bob, the +1 is hard to get excited about: his chances of success and failure both stay about the same, as a proportion of their values before the modifier was applied.

Now suppose you roll 2d6 instead. Alex needs a 4, Bob a 7, and Clovis a 11. Again the +1 halves Alex's failure probability (from 6/36 to 3/36), and doubles Clovis's success probability (from 3/36 to 6/36); but now even Bob has something to get excited about , because his failure probability went from 15/36 to 10/36. That's is a 33%reduction, as opposed to only a 10% reduction in the previous case.

Basically, with a bell-type distribution, the +1 modifier takes bigger bytes (in terms successes per attempt) in the middle of the spectrum, because the middle values are the most likely. But this is precisely where bigger bytes are needed in order to make a perceptible difference in the player's before-the-roll expectation. So I see that as an advantage.

That's just one thought on the issue.

RAN HARDIN: My thinking is kind of the opposite: on a linear curve, +1 means +5%, always. One person's +1 is the same as any other's. With a bell curve, that +1 can mean very different things to characters of different, let's say "skill levels." In play, Player One's +1 can be "worth" twice what Player Two's +1 are worth. I prefer linear, although it's not so strong a preference that I automatically one for the other.

NELSON: RAN wrote: >on a linear curve, +1 means +5%, always. One person's +1 is the same as any other's.

Adding the same amount to the success probability is a straightforward sense in which the modifier can "have the same effect" for everyone. But that's not the only way to think about it.

Suppose we take a vaule of 20%, and apply a +1 modifier and the success probability goes to 30%. Now what would that +1 modifier do to a 40% in order to have the same effect on both?

If the 40% becomes 50%, then we have added 10% both times. On the other hand, if 40% is modified to 60%, then the success probabilities have increased by the same proportion. On yet another hand, if the 40% is modified to a 53%, then we have decreased the failure probabilities by the same proportion. (All this talk of proportions may sound contrived, but it's really the way we intuitively measure differences: $100 seems far from $250; but $100,100 seems close to $100,250.) There are many more hands we could consider, all equally valid from a mathematical point of view.

Anyway, thinking about the probabilities is tricky business and in many cases is counterproductive, because it can obscure the intuitions we get from playtesting. Those intuitions are usually right. The mathematical machinations are usually flawed.