Randomize Difficulty/Skill in Rolling

[Evalyn]

I've got a system coded up that's superficially similar to the cortex system-- wherein increasing a skill will increase the dice "value," up until 12, at which point you "start over" with a new die tacked on.  A skill level of 2 represents a coin toss, 6 a similarly sided die, and so on-- until you reach a skill of 14, at which point you have one 12-sided die, and one coin toss (12+2).

From what I recall (I lost my book, and can't remember for sure), you roll your skills against static difficulties, and if the rolled skill-die is greater than or equal to the difficulty, you've passed the roll (with varying degrees of success and failure depending on by how much the roll's passed).

Compare this to (for example) DSS, wherein the difficulty is the electronic die.

The immediate differences are in where their respective 100% values lie; in the former, any difficulty higher than your skill is considered an auto-fail, whereas any skill higher than difficulty is an auto-success with the latter.

This is, I'll profess, primarily an aesthetic choice, as the difficulties in either case are far more dynamic than I care to write about.  The point to this rambling, I guess, is introducing a third option, wherein both skill and difficulty are randomized.

To use a very simple example:

Frodo's Accuracy is 10. He's given a difficulty of 6 to hit something with a thrown rock.

In the cortex system, his chance of success if 50% (where his dice rolling 1-5 would be a failure, and 6-10 a success).
With DSS, his chance of success is 100% (no matter which number between 1 and 6 is chosen, his skill of 10 will always trump it).
With a mixed system, his chance of success is 75%. And the math here is too annoyingly long to type.


tl;dr: If you had a fantasy skill system, and you were making some sort of action, would you prefer your skill "rolled" against a difficulty, difficulty against a skill, or a mix of the two?

[Ganymede]

If you're going to use a difficulty system, go with one where you roll to equal or overcome the difficulty. Simply, this is the most common form of the difficulty system, so players will understand it better.

[Whirlwind]

Although the randomized difficulty angle (which I've never actually used before, so its shiny new game-mechanic smell might be putting stars in my eyes) does have the two handy features of giving an idiot an outside chance of success, and having a "Gimmie" factor intrinsic to the system (as opposed to the Storyteller system, where "If your stat-total is greater than the difficulty..." is added as an optional footnote).

[Thenomain]

I've been reading but not yet playing a number of variable-die-size systems, and I have never had a problem understanding the basic mechanics or guess the consequences of the various dice pools or rare but occasional extra dice.

Using a set die type and a changing target number is more traditional, but you did miss one: Set die type but variable number of dice.  Or is this part what you meant by "variable skill"?

[Evalyn]

I've been reading but not yet playing a number of variable-die-size systems, and I have never had a problem understanding the basic mechanics or guess the consequences of the various dice pools or rare but occasional extra dice.

Using a set die type and a changing target number is more traditional, but you did miss one: Set die type but variable number of dice.  Or is this part what you meant by "variable skill"?

Not quite-- if we remove the die altogether and just use a rand() function, the ultimate question would be whether to randomize the difficulty or the players skill, before comparing it to the other; ie, is [rand(<skill>)] >= <difficulty>, or is <skill> >= [rand(<difficulty>)]? Or, for the third option, is [rand(<skill>)] >= [rand(<difficulty>)]?  Even your example of "variable number of dice" would fall under the former most category, where the final skill number is dynamic (to some extent, I know with WoD it's how many passes/fails, but bear with me here).

If you're going to use a difficulty system, go with one where you roll to equal or overcome the difficulty. Simply, this is the most common form of the difficulty system, so players will understand it better.

While I can sort of agree with the argument that the most common system is the one players will understand, I'm not quite as worried about the comprehension of the system. Mind you, I'm saying this right after clarifying a semi-confusing aspect, so take this with a grain of salt. To be honest, I think that the most confusing system in the world can be explained in a way that even the semi-literate can understand-- and anyone below that level of reading comprehension probably shouldn't be approaching a game based on the flowery writings of a dozen bored telnet'ers.

Although the randomized difficulty angle [snip] does have the two handy features of giving an idiot an outside chance of success, and having a "Gimmie" factor intrinsic to the system (as opposed to the Storyteller system, where "If your stat-total is greater than the difficulty..." is added as an optional footnote).

Note that there is a big difference between "chance of success" and "guaranteed success."  To clarify something I hadn't elaborated on, each system would have to have a different set of difficulty numbers-- for the example used in the OP, the DSS difficulty would be 20, to have a similar success rate of 50%.

The advantages of each depend purely on your style of play and/or staffing; would you prefer there be difficulties that no player can pass, under any circumstance? Do you like the idea that there are certain tasks so menial that, no matter the handicap, you will always succeed? Or would you prefer to approach tasks with the mindset that there's a possibility things can fuck up, no matter how unlikely-- or that, if you're 99% sure you're going to die, there's a 1% chance that god will smile upon you?

I hope that makes sense-- just got home from work, and I'm a little bit too frazzled to articulate things.

[Thenomain]

In that case, consider your audience.  If you're going to expose the system, rand the skill.  It's closer to what people expect  (I don't see a difference between 1d6 and rand(1, 6), but sure, why not.)

I'd be against rand vs. rand because I prefer to have a solid idea of the expectations of my character stats.

nWoD no longer has a straight threshold system; the dice rolled is modified and only one success is needed to be considered a success.  There is a threshold system, but it's for tasks that are suitably complex to be considered non-binary.  Not that this changes anything, but I don't think oWoD's dice system was very good.

[miss adventure]

I like the idea scaling of skill by die size, starting at d2 or whatever, then rolling over to a new die at a d10 or d12. Then you take the highest die roll. So d12 becomes the highest skill, and sure a d12 + d2 is sucking for any real gain, but you do reduce the chance of ever getting a 1. Note that this approach fully intends for the system to wrap values several times over, so people might easily have 3d12+d4.

You can also take the top 2 or 3 dice if you want a more curved base expectation.

For instance, You MIGHT do all this as a simple nurmeric scale, so skill 55 reduces to 4d12+d6 (or a d7), and you might take the highest single die. Or the highest 2 dice.

You could even follow the Silhouette idea and say that any extra max die rolls (probably only for d12s) gives you a +1 to the total.

On a computer, you could choose any die sides as the max, such as going up to a d20, or even again using pairs of dice as individual results, eg the above example would become 4 sets of 2d6, and a d7, add up the highest N values. Then you have a real curve, instead of total compression.


What does DSS stand for?

[NaN]

Don't Rest Your Head.

[Ganymede]

While I can sort of agree with the argument that the most common system is the one players will understand, I'm not quite as worried about the comprehension of the system.

Evalyn:

When it comes to developing a trust relationship with players, what you worry about or think is irrelevant. What matters to a player is what the player understands about the system: it helps them figure out how to create their PC, if you're using a stat system. One of the attractive features of White Wolf's ST system is that the game itself provides a simple measure of describing various power levels -- 1 is the average pool for unskilled Physical and Social rolls, and 4 is the average for someone considered "practiced" or "experienced" in a particular area. This general key helps players figure out how to put together his or her PC, even if you don't really know the percentages of success.

(Of course, this hypothesizes that White Wolf carefully figured out the percentages of success with regard to pools, a point that's easily challenged.)

So, think of the players first: that's why you should go with a simpler system.

[panic]

Have you thought about having both being randomized? Using a system that uses opposed rolls rather than one fixed and one variable?  Say an easy task has a difficulty of 1d6, a routine 1d10 and a difficult a 1d20, which is rolled against someone's skill level of 1d6 for a noob, 1d10 for someone adept, and 1d20 for someone who's an expert? 

[Whirlwind]

If I had any say in it, I'd want to tweak that so there's a gimmie-level.

I detest the systems where I can have a PhD in Whatsitology and be a professional Whatsitologist for the last 20 years, yet I screw up a Whatsit-related task so simple, most 12 year olds would get it right.

[Evalyn]

If I had any say in it, I'd want to tweak that so there's a gimmie-level.

I detest the systems where I can have a PhD in Whatsitology and be a professional Whatsitologist for the last 20 years, yet I screw up a Whatsit-related task so simple, most 12 year olds would get it right.

But at the same time, I don't like the idea of auto-passes in any situation. I'd consider myself pretty damned good at Modern Warfare, can knock down bitches like none other-- but there are days when I log on, and I'm just not into it, and I end up little more than a splatter on the wall.

When you roll to test your skill in PhD Whatsitology, you've got probability on your side. With your massive Whatsitologist brain, the likelihood of your failing is one in five hundred: but there's still that slight chance.

Have you thought about having both being randomized? Using a system that uses opposed rolls rather than one fixed and one variable?  Say an easy task has a difficulty of 1d6, a routine 1d10 and a difficult a 1d20, which is rolled against someone's skill level of 1d6 for a noob, 1d10 for someone adept, and 1d20 for someone who's an expert? 

[snip]
tl;dr: If you had a fantasy skill system, and you were making some sort of action, would you prefer your skill "rolled" against a difficulty, difficulty against a skill, or a mix of the two?

[snip] The ultimate question would be whether to randomize the difficulty or the players skill, before comparing it to the other; ie, is [rand(<skill>)] >= <difficulty>, or is <skill> >= [rand(<difficulty>)]? Or, for the third option, is [rand(<skill>)] >= [rand(<difficulty>)]?  Even your example of "variable number of dice" would fall under the former most category, where the final skill number is dynamic (to some extent, I know with WoD it's how many passes/fails, but bear with me here).

Admittedly, I can't quite figure out how to stress what I'm thinking, without somehow saying something wildly different-- but yes, I have thought about both skill and difficulty being randomized. The way you phrased it makes a helluvalot more sense than I did, at least.

Evalyn:

When it comes to developing a trust relationship with players, what you worry about or think is irrelevant. What matters to a player is what the player understands about the system: it helps them figure out how to create their PC, if you're using a stat system. One of the attractive features of White Wolf's ST system is that the game itself provides a simple measure of describing various power levels -- 1 is the average pool for unskilled Physical and Social rolls, and 4 is the average for someone considered "practiced" or "experienced" in a particular area. This general key helps players figure out how to put together his or her PC, even if you don't really know the percentages of success.

(Of course, this hypothesizes that White Wolf carefully figured out the percentages of success with regard to pools, a point that's easily challenged.)

So, think of the players first: that's why you should go with a simpler system.

Man, the coding part is easy, but now I have to worry about players?!

Out of sheer curiosity, if the math were explained in an easy-to-understand format, would the "mixed system" be a more viable option? Ex: "d6 is an average difficulty roll, and if you have a skill of 10, your chance of success is 50%*"?


*Note, the math here is likely totally and completely false, but I have to leave for work -now-, and can't take the time to actually figure out what the percentage is.

[Thenomain]

The rule of thumb in computer game design: Each element that builds upon a previous element increases the difficulty of learning exponentially.  (For instance, wall-jump is exponentially more difficult to learn than jump.  Mind you, jump is just mashing the button, but it's an easy to understand example.)

If you can minimize the need to know the math, you can make a more complex system without increasing the learning curve.

[Evalyn]

The rule of thumb in computer game design: Each element that builds upon a previous element increases the difficulty of learning exponentially.  (For instance, wall-jump is exponentially more difficult to learn than jump.  Mind you, jump is just mashing the button, but it's an easy to understand example.)

If you can minimize the need to know the math, you can make a more complex system without increasing the learning curve.

Note: I'm not planning on teaching a math lesson here. When put into practice, players would only need to know that a higher skill is more likely to succeed than a lower one-- no more a new concept than the other two (more common) ideas present.

I put the math there to highlight the innate differences between the three systems, and how the latter most fundamentally changes how "threshold"* rolls are handled by the former two. (And most certainly not because math turns me on, nuh-uh)




---
*Where, after a certain level, rolls will automatically succeed (Skill >= Difficulty, w/ Difficulty being "rolled") or fail (Skill < Difficulty, w/ Skill being "rolled")


ETA:
ALSO.

What does DSS stand for?

I am a retard-face and didn't notice miss adventure's question, 'til she PM'd me asking why I was such a douche. The answer, of course, lies in whether you believe in nature or nurture. A thousand apologies, please forgive my scatterbrain-ed-ness.

DSS stands for "Dahan's Skills System;" from what I recall, it was originally written with Star Wars in mind. The basic idea is that your skills range from one to a billion (with 50 being "Average," I think?), and are "static," to be compared to rolled difficulties. In essence, you're given a difficulty of dN, where N is the difficulty. You roll dN and, if it's lower than the skill in question, you pass.

Typical output looks like this:
You test Maryl's Agility against a 60 difficulty. The result is successful (42).

(Maryl's Agiliy is 50. I rolled a d60, and the result was 18. 50>18; roll is successful by a margin of 42)