innerdude
Legend
Okay, so I recognize that fiddling with the baseline mechanics of Savage Worlds may be considered heresy in some circles. The whole point of using SW at all is to make things FAST! FUN! FURIOUS! and take out a lot of the nagging "fiddling" with numbers that slows other systems down.
But hear me out here.
After playing GURPS for a little while (first experience ever with the system) the last 4 or 5 months, I came to appreciate a bit some of what it does. As a self-proclaimed "skill monkey" type when playing D&D, I really appreciated GURPS' ability to model well-roundedness.
However, to model that effectively, GURPS gets infinitely more granular than Savage Worlds--which, as much as I like Savage Worlds, is an admitted weakness of the system (SW is anything BUT granular to a high degree).
The second thing GURPS made me realize is that even though I detest "roll under" systems, and specifically some of the decisions about the way GURPS implements it, I really, really, REALLY like bell curve dice mechanics.
It's purely personal preference, I realize, but to me the bell curve is really "how the universe works." You have middle of the road, and you have outliers, and all things being equal, results push to the middle of the curve.
So, in a fit of, I don't know, insanity? Slap-happy craziness? I started thinking of how you could model a die-step mechanic like Savage Worlds, make it a bell curve, but also make it a bit (not a ton, but just a bit) more granular.
Here's the results of my efforts so far.
Dice Mechanics Changes
Change the flat distribution "either/or" trait die/wild die into a normal distribution.
In other words, instead of rolling both dice and taking the best result of either, we now add them together.
Wild die is still always a d6, and explodes normally.
Trait dice progress normally, but have more options at the top end. Since we're moving to a bell curve, this obviously changes the math (more on that later), but in effect it doesn't change the core mechanic--roll dice, roll high, and try and beat the target number.
In addition to the standard dice (d4, d6, d8, d10, d12), trait dice now also have d14, d16, and d20 at the high end (the d14 and d16 you can buy as part of a "zocchi" set, or simply use a digital dice rolling program).
However, to even out the normal distribution slightly, trait dice no longer explode, only the wild die.
Target Number Changes
Obviously, making this change totally screws up the assumed TN probabilities--but here's the thing, it doesn't necessarily matter of you account for them.
By default, SW assumes a TN 4 for everything, though it's a bit disingenuous, as situational modifiers are added and subtracted from the player's roll. They could just as easily simply raise or lower the TN appropriately.
Essentially, that's what will happen here.
Bad Axe Game's Trailblazer breaks down the assumed inherent probabilities for D&D 3.5, with the baseline assumption that at all level-appropriate challenges, the player should have a 65% chance of success. However, GM's change DCs all the time circumstantially. "Well, the floor is wet with rain, so that DC is actually an 18, not a 16."
Having run some initial probability numbers (thank you, thank you to the creator of Anydice.com. As soon as I can reasonably afford to pay you some money I will do so), it's fairly easy to create a reasonable "curve" for a target number using trait die + d6 wild die.
Starting with a d4 + d6 at the lowest, a player has a 58% chance of rolling AT LEAST a 6.
d6 + d6, 58% chance of rolling at least a 7
d8 + d6, 58% chance of rolling at least an 8
d10 + d6, 57% chance of rolling at least a 9
d12 + d6, 58% of rolling at least a 10
d14 + d6 57% of rolling at least an 11
d16 + d6 58% of rolling at least a 12
d20 + d6 60% of rolling at least a 14
Each step up or down the scale raises or lowers the probability between 6-13%, with the range of change narrowing the higher up the die scale we move (i.e., a 1-step change to a TN with a d4 skill die is a 13% variance; a 1-step change with a d14 skill die is closer to 7%).
So what's the point of all this again, you might ask?
Granularity.
By increasing the number possible die steps, and not being beholden to a static TN, we can create a much more fluid, and in my opinion, realistic way of resolution, without losing the speed and streamlining a die-step mechanic offers.
As a side note, if someone wanted to codify a "strict" TN rating, it's actually fairly easy to do for this system:
TN 4 = Very easy
TN 6 = Easy
TN 8 = Average
TN 10 = Somewhat Difficult
TN 12 = Moderately Difficult
TN 14 = Very Difficult
TN 16 = Limits of Human Capacity (i.e., someone who is one of the top-10 greatest ever in history only has a 45% chance of success).
The point behind this, though, is that since we're operating on a bell curve now, it's much easier to keep the TN closer to the center, without disadvantaging the low end too much, or making life too easy for the high end.
A TN 8, for example, has a 29% of success for someone with only a d4 trait die, but is still only a 79% for a d16. Yeah, that's a big difference--but a d4 is the effective equivalent of a Level 1 adventurer, and a d16 is considered a paragon of that skill / trait.
I've done some math on the combat portion as well, but I'll have to post them later on when I get more time.
But hear me out here.
After playing GURPS for a little while (first experience ever with the system) the last 4 or 5 months, I came to appreciate a bit some of what it does. As a self-proclaimed "skill monkey" type when playing D&D, I really appreciated GURPS' ability to model well-roundedness.
However, to model that effectively, GURPS gets infinitely more granular than Savage Worlds--which, as much as I like Savage Worlds, is an admitted weakness of the system (SW is anything BUT granular to a high degree).
The second thing GURPS made me realize is that even though I detest "roll under" systems, and specifically some of the decisions about the way GURPS implements it, I really, really, REALLY like bell curve dice mechanics.
It's purely personal preference, I realize, but to me the bell curve is really "how the universe works." You have middle of the road, and you have outliers, and all things being equal, results push to the middle of the curve.
So, in a fit of, I don't know, insanity? Slap-happy craziness? I started thinking of how you could model a die-step mechanic like Savage Worlds, make it a bell curve, but also make it a bit (not a ton, but just a bit) more granular.
Here's the results of my efforts so far.
Dice Mechanics Changes
Change the flat distribution "either/or" trait die/wild die into a normal distribution.
In other words, instead of rolling both dice and taking the best result of either, we now add them together.
Wild die is still always a d6, and explodes normally.
Trait dice progress normally, but have more options at the top end. Since we're moving to a bell curve, this obviously changes the math (more on that later), but in effect it doesn't change the core mechanic--roll dice, roll high, and try and beat the target number.
In addition to the standard dice (d4, d6, d8, d10, d12), trait dice now also have d14, d16, and d20 at the high end (the d14 and d16 you can buy as part of a "zocchi" set, or simply use a digital dice rolling program).
However, to even out the normal distribution slightly, trait dice no longer explode, only the wild die.
Target Number Changes
Obviously, making this change totally screws up the assumed TN probabilities--but here's the thing, it doesn't necessarily matter of you account for them.
By default, SW assumes a TN 4 for everything, though it's a bit disingenuous, as situational modifiers are added and subtracted from the player's roll. They could just as easily simply raise or lower the TN appropriately.
Essentially, that's what will happen here.
Bad Axe Game's Trailblazer breaks down the assumed inherent probabilities for D&D 3.5, with the baseline assumption that at all level-appropriate challenges, the player should have a 65% chance of success. However, GM's change DCs all the time circumstantially. "Well, the floor is wet with rain, so that DC is actually an 18, not a 16."
Having run some initial probability numbers (thank you, thank you to the creator of Anydice.com. As soon as I can reasonably afford to pay you some money I will do so), it's fairly easy to create a reasonable "curve" for a target number using trait die + d6 wild die.
Starting with a d4 + d6 at the lowest, a player has a 58% chance of rolling AT LEAST a 6.
d6 + d6, 58% chance of rolling at least a 7
d8 + d6, 58% chance of rolling at least an 8
d10 + d6, 57% chance of rolling at least a 9
d12 + d6, 58% of rolling at least a 10
d14 + d6 57% of rolling at least an 11
d16 + d6 58% of rolling at least a 12
d20 + d6 60% of rolling at least a 14
Each step up or down the scale raises or lowers the probability between 6-13%, with the range of change narrowing the higher up the die scale we move (i.e., a 1-step change to a TN with a d4 skill die is a 13% variance; a 1-step change with a d14 skill die is closer to 7%).
So what's the point of all this again, you might ask?
Granularity.
By increasing the number possible die steps, and not being beholden to a static TN, we can create a much more fluid, and in my opinion, realistic way of resolution, without losing the speed and streamlining a die-step mechanic offers.
As a side note, if someone wanted to codify a "strict" TN rating, it's actually fairly easy to do for this system:
TN 4 = Very easy
TN 6 = Easy
TN 8 = Average
TN 10 = Somewhat Difficult
TN 12 = Moderately Difficult
TN 14 = Very Difficult
TN 16 = Limits of Human Capacity (i.e., someone who is one of the top-10 greatest ever in history only has a 45% chance of success).
The point behind this, though, is that since we're operating on a bell curve now, it's much easier to keep the TN closer to the center, without disadvantaging the low end too much, or making life too easy for the high end.
A TN 8, for example, has a 29% of success for someone with only a d4 trait die, but is still only a 79% for a d16. Yeah, that's a big difference--but a d4 is the effective equivalent of a Level 1 adventurer, and a d16 is considered a paragon of that skill / trait.
I've done some math on the combat portion as well, but I'll have to post them later on when I get more time.
)
