[SP: Freelancers] Designing Dice Mechanics

[TracerFox]

I was wrestling with this rather late last night, and have come to the conclusion that I am not enough of a dice math guru to answer this on my own.
So I have decided to post this for public review and commentary.

Here's my problem. I have made the decision and assumption that within my game the player characters are well trained, above average competency individuals, and that they will have a 60% chance of success right out of the gate.
This means that I have to design both chargen and dice mechanics in such a way that characters right out of chargen can succeed 60% of the time, without any buffing or improvement of their starting scores.

So, I'm asking for examples of dice mechanics. I have a fondness for roll-under target number systems (where ability determines a TN, and roll the dice and try to roll under). If possible I would also like to have the result range be divisible by 5 (1-5, 1-20, 1-100, etc), for some reason it seems easier on the brain.

So, any suggestions or questions?

[Tobias]

Roll under a 7 on a d10?

While this may seem too obvious and silly, it does work. And why make a system that requires more?

UNLESS you want that 'more' to matter or interface with something else, somehow. In which case, you'll need to consider that.

[Ron Edwards]

Hello,

Tobias is right. Given what you've said, just give'em all 60% to start, in everything. If you want them to have a bit better chance at a couple of things, to customize characters, give'em 70% or 75% at two abilities of their choice. Easy, effective, done.

Best,
Ron

[Sydney Freedberg]

Not to complicate things needlessly, but there is a basic question being begged, I worry:

characters right out of chargen can succeed 60% of the time, without any buffing or improvement of their starting scores.

Succeed at what? At doing ordinary, nonstressful things that you see in daily life (e.g. drive down the street, shoot a paper target on the range, talk with your spouse)? At doing extraordinary, stressful things that you see in dramatic situations (e.g. chase another car through busy traffic, shoot an enemy who's shooting back, get the prom queen/quarterback to go out with you)? Or at doing over-the-top crazy things you see in action movies (e.g. seduce your archenemy while shooting off-handed at a car you're chasing through speeding traffic driving the wrong way)?

[groundhog]

What Sydney said about difficulty levels is very important. The death of many games I've seen die was due to poorly chosen difficulty levels.

If you're looking for a cinematic style, you still need to determine just what kind of cinema you're talking about. Do you want elite troops like in Predator, or do you want each character to be the like commander of those troops? Do you want a Chuck Norris style movie where he takes down dozens of enemies one or two at a time, or something like The Matrix in which Neo can take down hundreds of opponents dozens at a time?

Are you looking for The French Connection, James Bond, Jason Bourne, Batman,  Knight Rider or Gone in 60 Seconds types of car chases? Maybe something more along the lines of Cannonball Run or Dukes of Hazzard? Jason Bourne or Luke Duke do amazing things with normal cars. Michael Knight, Batman, and James Bond have amazing cars. There's a big difference in style and feel here. There's also a big level of skill difference in driving a sports car on an empty dirt road running from one sherriff vs. driving a Mini in crowded city traffic and running from a number of other vehicles.

Basically, saying that you want a 60% success rate out of the box doesn't mean one thing to every game. Fox Mulder, James Bond, and Bruce Wayne all have fascinating careers fighting bad guys. They all three are bright, athletic, and well-funded. All three see lots of success for their respective film genres. There's much different among them, though. A smart guy with a gun and some level of government backing need not be self-financed likie Bruce Wayne. He need not have gadgets like Bruce Wayne or James Bond. He need not have the full cooperation of the government that backs him. He also need not be an atheletic phenomenon -- a good runner who can hold his own hand-to-hand when his pistol is empty can be just as exciting. Fox Mulder would be just another mostly average person in a James Bond or Batman movie. However, in a grittier genre in which the great unknown drives the plot, he's an amazing step above most people in his own Bureau.

[TracerFox]

Succeed at what? At doing ordinary, nonstressful things that you see in daily life (e.g. drive down the street, shoot a paper target on the range, talk with your spouse)? At doing extraordinary, stressful things that you see in dramatic situations (e.g. chase another car through busy traffic, shoot an enemy who's shooting back, get the prom queen/quarterback to go out with you)? Or at doing over-the-top crazy things you see in action movies (e.g. seduce your archenemy while shooting off-handed at a car you're chasing through speeding traffic driving the wrong way)?

Good replies here Sydney, Groundhog. I see your points about both overall setting difficulty/lethality, and individual task difficulty.

Okay. I think the closest answers to the overall setting angle is that the PCs should be analogous to a skill/training level of the crack troops from Predator. Not every one of them will be Dutch (the leader), but they'll be at *least* as good as the worst member of that group.
Now that that gives and idea of the capabilities of the average PC, let's set up a 'normal' difficulty for individual tasks.

Using Sydney's examples above, I would say the 60% out of the box success would be during stressfull, combat/action oriented actions. Shooting an enemy during a firefight would be the most common example.

This game isn't 100% about combat, but it is heavily focused on the actions one would see in a potential combat situation (Hostage rescues, infiltration of enemy bases, defending a location/item against attack, escort missions, etc...). I want the 'default' success rate in these kinds of situations, for a group of well trained and above-average capability 'troops', to be right around 60%.

[Lance D. Allen]

To set your base level of difficulty, you have to determine what counts as an action that requires a roll. From what you've said so far, shooting at paper targets wouldn't require a roll at all. If the player wants to hit, he hits. If he wants to miss, he misses. If he REALLY wants to roll, he can, but the rules shouldn't call for it.

In aid of helping you determine what counts, I offer my definition of a "task" from Mage Blade as an example.

: A task is anything that a character does that has a reasonable chance of failure, and significant consequences for failure. What this means is, if the consequence for failure would not have any real impact on the events of the game, then it is not a task. Likewise, if the chances of failure are extremely marginal (less than a 1% chance of failure, say) then it is also not a task. For instance, even if failing to walk across a steady, fairly broad suspension bridge would have a momentous effect (falling to your doom, for instance) but the chances of failure are extremely tiny (because it is a fairly broad bridge that is not moving nor is it falling apart, and because you've been walking since early childhood) this is also not a task.

You may very well want to scale the level up from what I have, or possibly even down. But, if I have to say so myself, my definition here is a good place to start.

[Tobias]

Not every one of them will be Dutch (the leader), but they'll be at *least* as good as the worst member of that group.

How true. We can't ALL be Dutch. ;)

Sorry. Couldn't contain that one. On a more appropriate note:

Still no need to deviate from the 1d10 roll. Just add -x or +x modifiers as appropriate for difficulties that deviate from the standard you've mentioned.

(With 'appropriate' decided by player (for instance, by his setting skills during character design, or during the game to 'raise the stakes' in some manner), GM, rules table, group agreement - what you like).

[eef]

How do you want the overall game to play?
For instance,

-- Conflict or task resolution?
-- How many rolls to a resolution?  i.e. "I take him out!" vs "I hit him again, three more and he should be down"?
-- Can any X have a chance to beat any Y?  I.E., if a high school dweeb gets into a fistfight with the Bruce LeeBot, does HSD have any chance at all?

Why do you want people to play your game?

==Ed

[Sydney Freedberg]

-- How many rolls to a resolution?  i.e. "I take him out!" vs "I hit him again, three more and he should be down"?

Good point. The probabilities are very different depending on the number of times you roll (the Burning Wheel rulebook has a nice little section on this):

a) If you roll once to win the fight/get the girl/sneak into the castle, and you have a 60% chance on each roll, obviously, you have a 60% chance.

b) If you have to roll, say, four, times to work step by step to your desired result, and each step only succeeds if all the others succeed (e.g., roll to spot target, roll to hit, roll to penetrate armor, roll to do damage), then you're actually looking at 60% of 60% of 60% of 60% (0.6 to the fourth power), which is only a 12.96 % chance overall.

c) If you roll four times to achieve your desired result, but you achieve the overall goal if you succeed on any one attempt (e.g. I attack each round, I have a 60% chance to succeed each time, any successful attack will neutralize the foe*), then you have -- uh -- uh -- a really high chance. (It's late, I have an early-rising baby, I was never good at combinatorics, anyone want to help me on this one?)

Which of these three you pick makes a huge difference.

* Note that one successful attack effectively guarantees victory in a lot of systems where being wounded (or disgraced in a social conflict, or whatever) gives you a penalty to later rolls to avoid being wounded again -- a phenomenon called the "death spiral." Even if I hit you lightly and impose a 10% penalty, that 10% penalty makes me more likely to hit you for another 10% next time, and then you're at -20%, so I'll probably hit you for 40%, and so on; once you took that first hit, you were already dead, you just didn't know it yet. Which is somewhat realistic, but sometimes in real life and fiction alike, that first hit just gets you mad and makes you come back harder and slam me down, which is trickier to write a rule for.

[xenopulse]

Four rolls with 60% chance is a 97.44% chance of at least one success. You can do this in sequence (40% failure at first roll, of which 60% will be success on second roll, so you're left with 16% failure, of which...) or simultaneously (4D10, no roll over 4, there are 256 possible combinations for that (4*4*4*4), and the remaining 9744 possible combinations contain at least one number between 5 and 10.

So I agree, it's not only about conflict vs. task resolution, it's also about the amount of rolls involved in a task.

Alternatively, you could make a system in which the PCs always succeed, but have to determine the cost of success. That seems reasonable for veteran characters. Most of the time, they'll make it just fine. Sometimes they take a hit or trigger some other unwanted consequence.

[giblfiz]

I'm going to take a wild leap here, and guess that your looking to make a kind of crunchy, gameist system, and that by 60% success rate you mean "a 60% chance of hitting a single opponent and doing 1/3 of a normal opponents HP or 1/10th of a heroic villains. I'm also assuming that your primary concern is combat.

as recommended better than a 7 on a d10 works well, particularly if you want a flat advancement curve.

better than an 8 on 2d6 will net you 58% chance of hitting. it also has a different curve to it, so that buying from 8-9 is much better than buying from 11-12.  


Another way to go is if you require 1s to hit, and give the players a starting dice pool of 5D6 ( ~58% of getting at least one 1) Once again, if you give them more dice, then there are natural diminishing returns ( the sixth die in the pool buys them an extra 7%, while the 9th only gets them 4%) this also makes a lot of gameist players happy because they get a big fist full of dice.

You have to ask yourself a second question as well... do you want your damage to be an independent roll? or is that  60% of 1/3 HP rolled up in your one roll.. There are lot of ways you might do this...

I.E. if you use 2D6 and make players get a 7 or lower, you could also say that they do as much damage as they roll. They will on average do 5 damage on average when they hit.. The guy who needs an 8 or lower (because of his better training) will also be able to deliver more damage in the form of an 8 damage hit. (his average damage will be much closer to 6)

So the real question is what else do you want? how do you want your advancement slope? what about your damage slope? how do you want to hit and damage to play together? how many shots do you want a player to have?

I hope I gave you something in the way of help, if you want to post more about your game, I might be able to generate more insightful comments.

Best luck
-Harry