Hi, all, :)

I have no idea if you guys do this sort of thing or not, and if so, whether this is the right forum for it, but it seemed like a good idea at the time. So...

What am I talking about? Well, I came up with a (seemingly) interesting idea for a resolution mechanic, and I decided to toss it out here for the purpuoses of: usability discussions; and appropriation by others for their own games, if they want it.

The mechanic is based on an exploding d3. PC skills are relatively low numbers (say 3-10). Target numbers are also relatively low numbers (say 1-12). The procedure is as follows:

Start with your skill level.
Roll the d3.
- If it comes out 2, you're done.
- If it comes out 3, add one to your skill level. Roll again.
- - If it comes out 1 or 2, you're done.
- - If it comes out 3, add one to your skill level. Roll again.
- - Repeat untill it doesn't come out 3.
- If it comes out 1, subtract one from your skill level. Roll again.
- - If it comes out 2 or 3, you're done.
- - If it comes out 1, subtract one from your skill level. Roll again.
- - Repeat untill it doesn't come out 1.
Take your final result and compare it to the target number.
- If it is equal, you have succeeded nominally.
- If it is lower, you have failed altogether.
- If it is higher, you have completely dominated the task.

So... that's it. Thoughts?

Cheers,

J.

I had a very similar idea in the past, which basically went:
1- subtract one from skill, roll again, keep going if 1's keep showing up
2-5- just your skill
6- add one to skill, roll again, keep goin if 6's keep showing up

Of course, the idea was that each skill rank was a significant jump from the last.

I like how it becomes fairly predictable as to the outcome, but I junked it because the mechanic didn't appeal to me as a player, it just lacked that bit of punch that makes rolling dice fun. Perhaps if combined with some other mechanics, it might make things work, but it seemed too blah to work by itself.

Other than that, it's a fairly sound mechanic, as long as you recognize that players have 1/27 chance of jumping up or down 3 levels in skill. In my system, ranks went: unskilled, novice, trained, expert, master, grandmaster, superhuman, supernatural, monstrous, legendary, godlike, so jumping 3 levels is kind of sick. All you'd need to do is make 3 levels a rare, but still plausible thing by scaling your skill ranks accordingly.

Chris

Here's my mechanic working similarly. Note that it's very heavy on the centering:

Roll 3(?) dice:

Count 1's and 6's. Subtract # 1s from # 6s.

if the result is 0, your skill is performed at your skill level

if the result is less than 0, reroll all the 1s, if you roll new 1s, keep subtracting the number of new 1s rolled from the original result and reroll the 1s

if the result is above 0, reroll all the 6s, if you roll new 6s, keep adding the number of new 6s rolled from the original result and keep rerolling the 6s.

Add the result to the skill level.

You can vary the number of dice for different distributions. This scheme is simpler to do in actual play than to explain it. Which incidentally is why I didn't use it for Ygg.

Example: I roll 5,6,2. result is 1. Reroll the 6. I roll 1. Ok final result is 1. I perform at skill level +1.
2: I roll 1,6,3 result is 0, I perform at skill level.
3: I roll 1,1,1 result is -3, I reroll 3 dice, 1,6,3. Result is not -4, I reroll 1 die, result is 3. I perform at skill level -4

Cool, exponential decay functions!

I kept the probabilities in fractional form so you can see the patterns.

[code]
STAT CHANCE OF CHANCE OF CHANCE OF
DIFFERENTIAL TIE (MARGINAL DOMINATING FAILURE
SUCCESS) SUCCESS

-4 2/243 1/81 193/243
-3 2/81 1/27 76/81
-2 2/27 1/9 22/27
-1 2/9 1/3 4/9
0 1/3 1/3 1/3
1 2/9 4/9 1/3
2 2/27 22/27 1/9
3 2/81 76/81 1/27
4 2/243 193/243 1/81
[/code]

This is a very stat-senstive mechanism. When the stats are against success, each additional point of stat difference makes success one third as likely. When the stats are in favor of success by at least two points, each point of stat difference reduced makes failure three times more likely. So in general, one can say, approximately, that each point of target difficulty represents a threefold increase or decrease in the difficulty of a task. Similarly, each point of character skill represents a threefold increase in effectiveness (as it will, in most cases, make the character either three times less likely to fail, or three times more likely to succed).

This sort of consistency is advantageous, which is the reason I use exponential mechanisms most of the time. But the threefold increase/decrease per point of stat is rather extreme, yielding a lot of very high and very low odds unless all the stats are grouped very close together.

For a different mechanism with exponential behavior, see this thread. It has the same kind of mathematical behavior, but much less steep decay curves, and you can do all the die rolling at once.

- Walt

PS Christoffer, here's an exponential mechanism that incorporates your randomness knob.

Set a randomness level, from 0 (totally stat-sensitive) to 9 (highly random).

Roll a d10. As long as you keep rolling less than or equal to the randomness level, roll it again. Count how many rolls you make before failing to roll the randomness level or less. (Don't count the final roll that exceeded the randomness.) Note that there's no point in actually rolling if the randomness is 0; the number will just be zero.

Now roll any 50-50 chance, to determine whether the number you just got is added to or subtracted from your skill level. Do the addition or subtraction, and compare the result with whatever the target is.

Hey guys,

I've got a question: what's the value add? It seems like an exercise in gratuitous handling time, with no value add other than the fun of rolling dice. Couldn't you just map success levels to percentages on a chart, and use d100 (or d1000) to achieve the same result without the rolling and rolling?

I can see it if successive rolls and calculations were introducing fresh qualitative data to the resolution, but they aren't. You're just crunching numbers.

Paul

Paul Czege wrote: Hey guys,

I've got a question: what's the value add? It seems like an exercise in gratuitous handling time, with no value add other than the fun of rolling dice. Couldn't you just map success levels to percentages on a chart, and use d100 (or d1000) to achieve the same result without the rolling and rolling?

I can see it if successive rolls and calculations were introducing fresh qualitative data to the resolution, but they aren't. You're just crunching numbers.

Paul

[:-)]Well, all these dice mechanics are just idle speculation for me. I much prefer to use spinners. You can create any probability distribution you want, just by firing up the pie chart option in Microsoft Excel.[/:-)]

Seriously, and speaking only for myself, I doubt I'd ever use any roll one die at a time mechanism. I don't even like exploding dice for that reason (just when you thought you were done, there's more work to do).

But some of the mechanisms I do use, I've only arrived at after contemplating several impractical variations first. I can probably come up with a single-roll version of JM's mechanism (though it will require custom-marked dice, and won't yield a success or failure margin) as a Symmetry variant. But first I wanted to find out how he likes the outcome distribution he's got.

- Walt

Hey, :)

wfreitag wrote: Cool, exponential decay functions! <...> For a different mechanism with exponential behavior, see this thread.