A search for an optimal resolution mechanic

[Ganbare Gincun]

People who want to add "10% of the total" to their chance of success gravitate towards flat RNGs because flat RNGs are good at giving you that output. If you get +4 on a d20, then 20% of the time you roll the die you will succeed instead of failing because of that modifier. People who want to add "one standard deviation of the total" to their chance of success gravitate towards curved RNGs because curved RNGs are good at giving you that output. If you get +4 on 3d8, there is one standard deviation of the data that would have beaten you that now does not.

So if you wanted to switch from a curved range to a flat range, would you want to retain the bonus/penalty modifiers that you normally use within a given game system (like say for example HERO System), or re-calculate them to accommodate for the new standard deviation of the new range?

[Dean]

Dean, I rather like that idea.

Going from a dice pool; to a single dice roll is a lot better imo.

I'm considering using this as a method to show a "change" in tiers. Allowing a character to replacing a pool of 5 dice; with a single d12 on that "hit" spread would keep total dice pools small as creatures get more powerful.

Why thank you. So your saying for your die pool system, abilities would allow you to go up to 5d6 in a given die pool and then afterwards let you use a 12 sided die labeled as follows:

0, 1,1, 2,2,2, 3,3,3, 4,4, 5: Which is 12 sides

To simply make a less obnoxious upgrade than 6d6 and beyond. Is that correct?

[PhoneLobster]

Labelled single "curved" dice are things I still dislike for transparency reasons.

But they DO eliminate one of the other major stumbling blocks of standard curved dice mechanics by removing the requirement to total up several dice on every roll you make, which as small as it is really adds up (or rather multiplies) on resolution times and complexity costs fast.

Having a single "curved" dice brings the resolution times and complexities directly in line with standard single dice mechanics.

So it has that going for it, at least, so I certainly would encourage all those god damn curved RNG adherents to convert to such a mechanic immediately due to it's obvious objective superiority. But then, being who they are they won't even understand why that would be an improvement.

[Judging__Eagle]

Dean, I rather like that idea.

Going from a dice pool; to a single dice roll is a lot better imo.

I'm considering using this as a method to show a "change" in tiers. Allowing a character to replacing a pool of 5 dice; with a single d12 on that "hit" spread would keep total dice pools small as creatures get more powerful.

Why thank you. So your saying for your die pool system, abilities would allow you to go up to 5d6 in a given die pool and then afterwards let you use a 12 sided die labeled as follows:

0, 1,1, 2,2,2, 3,3,3, 4,4, 5: Which is 12 sides

To simply make a less obnoxious upgrade than 6d6 and beyond. Is that correct?

Pretty much.

I've segregated powers that creatures can buy into "tiers"; and each ability only goes to a dice pool of 5; the 6th point simply shows that the creature is now at the "zenith" of that type of training.

Players are allowed/encouraged to add dice pools together equal to [The number of Tiers they have powers in + 1 piece of Equipment].

Previously, I just had dice pools go to 6d6; but that 6th dice doesn't really represent the change I want it to.

Instead of having dice pools that go to 6; players can buy the zenith training, and roll a d12 instead for that pool. Resulting in players combining zenith and non-zenith pools; without having total dice rolled go to 20+; usually more along the lines of 10-ish at the most.

[Username17]

These funky dice with different numbers on them are terribad. First of all: that's incredibly inconvenient. 24 sided dice? Are you fucking kidding me? Secondly, those don't actually have any of the mathematical effects of a curve. Indeed, they are not curved. There is no derivative available between one number and the next, it's just a step function. And finally, they have very little power in generating unique states. I mean, fuck, even 3d3 generates 27 unique states, 24 is actually quite paltry.

-Username17

[CatharzGodfoot]

I mean, fuck, even 3d3 generates 27 unique states, 24 is actually quite paltry.

If you're using base 3. Otherwise it just generates 7 unique states in 27 different ways. But, yeah, the d24 is bullshit. Even if you can get it to stop rolling.

[Judging__Eagle]

I'm thinking of using d12's actually. Everyone that I game with has "one" or "two"; and they could generate results that correlate with a bell curve.

Technically; a single d3 could represent 1, 2 or 3 hits in a pool.

[Manxome]

Secondly, those don't actually have any of the mathematical effects of a curve. Indeed, they are not curved. There is no derivative available between one number and the next, it's just a step function.

WTF? So are dice pools. Dice are discrete, so any possible way interpreting a roll of any finite number of dice is technically a step function.

A dice pool with 3d6 and TN=4 is statistically identical to a d8 labeled 0,1,1,1,2,2,2,3. They both approximate bell curves to exactly the same degree. Similarly, you could relabel a hypothetical d36 to be exactly equivalent to the sum of 2d6. And using relabelings that don't happen to exactly correspond to some multi-die roll you're familiar with are no less valid.

Putting weird labels on a many-sided die may not be an ideal way of approximating a curve for a number of practical reasons, and d24 certainly doesn't seem like the obvious choice even if you did, but complaining about whether the derivative exists is definitely the pot and the kettle unless you're proposing an alternative that involves infinite dice or infinite-sided dice.

[RobbyPants]

All this talk of d3s reminds me of Betrayal at House on the Hill. All of it's dice are d6s with two 0s, two 1s, and two 2s. This makes them effectively each d3-1. So, when you roll X dice, you always have a minimum of 0, and average of X, and a maximum of 2X. This is basically the same as an XdY bell curve, but it's quick for figuring out your average and your bounds.

The whole resolution mechanic is that when you roll a stat check, you roll a number of dice equal to your stat and you try to meet or exceed a particular value. There are also opposed rolls for combat.

It's a really simple system, although if you can get extra dice, things get easy pretty fast. It's hard to gauge how much of that is the system and how much of that is because of all the really weird scenarios.

[wotmaniac]

Well, since there's already a thread here, I use this one ....

So I've been toying with my own version of a die pool mechanic; and I'd like to have you guys rip it apart for me. Here's the basics:
Dice of choice is a d12. Roll is based on skill and related attribute.
Let's say that the entire pool is derived from just your "skill", and you've between 1 and 12 ranks (giving you the same # of dice in your pool). There are no other sources of dice
Attributes range from 1 to 6, and this sets your TN. i.e., a 1 in an attribute means that you only get a hit on a 12; a 2 gives a hit on an 11+; [...]; and a 6 in the attribute hits on a 7.
Then you have some sort of "power level" stat, or some-such, that lets your dice explode at progressively lower numbers. So, level 1 is no explosion; level 2 explodes on a 12; etc., capping at level 5 letting your dice explode on a 9. This is independent of your attribute; so you could, for example, have a case where you have, say, a 2 in the attribute and a 5 power level, meaning that even though a 9 on the die would not be a hit, but you would still get to re-roll it. (that last bit is still a little iffy).

Like I said, this is just the basics.
Go ahead, I'm ready -- let me have it.


(oh, and if I really need to start another thread, just say the word)

[Username17]

So fixed TN 13, but you add bonuses to every die, and you're rolling a fuck tonne of d12s? What the fuck made you think that was a good idea? And the whole variable bonus dice thing is a non-starter because we already saw how shitty that was in nWoD.

-Username17

[souran]

There is a type of dicing game that has not been talked aobut yet.

Silhouette: dice pool of d6 plus a static modifer, take the higest die, additional dice above a theshold number add a fixed (usually 1) value to the result, check against a target number.

Example: you have an "attack" skill that is a pool of 3 dice and a +2 modifier. you roll and get 2 sixes and a two. Your result is 6 (highest rolled) +1 (additional 6 rolled) +2 for a grand total of a 9.

This system has an exponential difficulty scale while combining some of the advantages of systems like shadowrun and d20.

I know frank doesn't care for it but honestly I really think that the silhouette system does'nt get enough discussion this board. Its easily the most robost system mechanically I have ever played.

too bad that the games that they decied to make with it are "canadian mechwarrior" and "canadian mechwarrior IN SPACE!"

Seriously, the game system is quite difficult to break, generates the answers the rulebooks say they are going to generate, and they included all their statistical analysis about how their derived their expected difficulties/modifer values and whatnot in the back of the book.

Hell they even tell you how you can play their system with d8s and d10s instead of d6s and what that does to the whole game. Its actually pretty cool.

[PhoneLobster]

Silhouette: dice pool of d6 plus a static modifer, take the higest die, additional dice above a theshold number add a fixed (usually 1) value to the result, check against a target number.

See this.

See how BAD and STUPID your description was and how by reading that sentence it was almost impossible to understand WTF you were talking about?

This is the FIRST of MANY things wrong with this stupid ass mechanic.

I can't even BEGIN to imagine why you think it is "good" the sheer resolution times alone per basic roll are hideous.

Go hang your head in shame. Dunces hat and a corner for you!

[PoliteNewb]

Silhouette: dice pool of d6 plus a static modifer, take the higest die, additional dice above a theshold number add a fixed (usually 1) value to the result, check against a target number.

See this.

See how BAD and STUPID your description was and how by reading that sentence it was almost impossible to understand WTF you were talking about?

WTF is wrong with you? I've never even heard of Silhouette, but I was easily able to comprehend what he was talking about, even without his example.

Are you just reading-impaired? Do you require the mechanic to be so simple that chimpanzees can use it? Seriously, what?

[PhoneLobster]

WTF is wrong with you? I've never even heard of Silhouette, but I was easily able to comprehend what he was talking about, even without his example.

Oh rly?

Cause when I read it went like this...

dice pool of d6 plus a static modifer

Okaaay... but if it's so "Static" why don't we know the modifier now, but whatever, moving on...

...no wait NOT moving on. Because ACTUALLY that static number isn't "added" now at all, you add it AT THE END. This is actually a terrible attempt at some sort of nested brackets without the nested brackets where "dice pool of d6" is apparently a complex OPERATION that we need to refer to the rest of text to decypher then we need to go BACK to this first bit of informal text and finish the bit where we add a static modifier. Holy crap that is BAD DESCRIPTION.

, take the higest die,

Take it where? Do what with it? Is it a "success"? A +1? Is it worth the value on the roll (this is what he MEANS to have mentioned)? BUT HE DOES NOT FUCKING SAY. Not here, not in the next bit, not ever.

additional dice above a theshold number add a fixed (usually 1) value to the result,

add "unknown, maybe a 1" to... yeah. How many times? Just once IF there are additional die? Or once PER additional die? Or something? We don't know!

And we are checking for being above a "threshold number". Oh great, first mention of that. Is this the same threshold as... wait, the "higest die" part didn't have a "threshold" at all. It's purely for... this entirely retarded part of this retarded mechanic.

I wonder what the threshold is. Is it "static" is it changeable? WTF is it? I can understand the concept of "check against a number" but without any further explanation of the NATURE of this number we have learned little to nothing about this mechanic right here. Especially with a different target number at the end presumably filling our actually changeable difficulty role.... Or is it?

check against a target number.

Well that would seem simple enough only with that threshold business that is the SECOND target number checking against phase in this MESS.

Oh and we aren't told HOW to check I mean what is this? Greater than? Equal or greater? Maybe it's even a ROLL UNDER for all the fuck we know at this stage?


In fact lets step aside from the badness of the description to talk about the complexity for a second. And reveal what the actual nature of the poorly explained mechanic you are defending out of sheer spite actually IS.

1) Roll multiple dice.
2) Compare results to each other and search for "higest die".
3) "Take it" in some manner.
4) Check each remaining die against mysterious "Threshold"
5) Multiply (number of) remaining die "over threshold" by "unknown number maybe 1" (note souran does not in fact clearly explain that EACH dice adds this number)
6) Add result of 5 to result on "higest die" (not actually clearly explained anywhere by souran's description)
7) Add "Static Number" modifier to total (very poorly placed in souran's description)
8) Compare result against target number

With no idea what comparison leads to a success or not. Didn't seem worth mentioning at this point apparently.

So that's an EIGHT STEP operation for your BASIC MECHANIC. Every time you make a check in this game you go through EIGHT DISTINCT STEPS OF OPERATION. Also some of those steps may occur multiple times perhaps MANY multiple times, before you continue on to the next step.

So even once we decipher it how in HELL is this a mechanic of practical and usable complexity for actual people in an actual table top game to actually use?

Do you require the mechanic to be so simple that chimpanzees can use it? Seriously, what?

It is a FORMAL mechanic. The BASE mechanic of a game. It needs to presented in an utterly unambiguous and readable manner. THAT was NOT unambiguous or readable.

I know you have a personal vendetta against me Politenewb, but really, get a fucking grip.

This is an insanely stupid mechanic and it was very poorly described.

Defending it and that description of it makes you look... wait... like you...

[Ice9]

1) Roll multiple dice.
2) Compare results to each other and search for "higest die".
3) "Take it" in some manner.

Man what. At the point when you start trying to claim that "Roll multiple dice and take the highest" is a confusing and byzantine process, you have become a parody of yourself.

[PhoneLobster]

Oh look ANOTHER guy prepared to throw all credibility to the wind because of a personal vendetta against me. And also because he is stupid.

Let's compare and contrast what he read...

"Roll multiple dice and take the highest"

To what was actually said.

dice pool of d6 plus a static modifer, take the higest die

So actually we are dealing with a dice pool mechanic where results of die rolls are often not the rolled numbers but binary successes. And we are adding a static modifier BEFORE taking our "higest" whatever. Only we aren't adding first because he just threw the end of the routine in the beginning because he couldn't describe his mechanic clearly at all.

Yeah. "Roll multiple dice and take the highest" would have been clearer, not in this context clear enough, but clearer. It's a pity that isn't the description we are talking about.

a confusing and byzantine process, you have become a parody

And you know, taking the simplest three steps of an 8 step process, not even including the iteratively repeating steps, and pretending that the 8 step process isn't byzantine because PART of it is relatively simple AFTER being deciphered. Yeah. You aren't a raging moron...

[erik]

What really threw me was the threshold number. I didn't know if it referred to the result of the dice that were not the highest or if it was the quantity of dice that were not the highest.

i.e. each die that rolls over a certain threshold of possible rolls versus each die that you have over a certain threshold of dice to be counted.

If the threshold and threshold bonus are variable, that does sound like things will get a bit more confusing as a base mechanic. Not insurmountably so, but it is more things to track when you also have die counts, modifiers and target numbers that likely will shift on top of it all. Five things to keep track of for any given roll doesn't sound like a friendly mechanic.

[PhoneLobster]

Another thing about the thresholds not made clear in the description but implied by the example is that when you "take" the "higest die" you aren't just recording/remembering the result for your calculations, you are ALSO actually REMOVING the die from the pool PRIOR to making the threshold comparison to get the number of die over the threshold.

And what the threshold number is remains unclear even after the example, where for all we know the "threshold" might as well be "the result on the highest die", or you know, any other number over 2 or something... something... underpants gnomes.

[wotmaniac]

So fixed TN 13, but you add bonuses to every die, and you're rolling a fuck tonne of d12s?

At first, I didn't know what the hell you were talking about here; but then I realized that you were just coming at it from the other direction .... I hadn't thought of it in that light. That seems simpler to explain than "attribute-dependent TN". Huh, nifty.

And the whole variable bonus dice thing

Okay, now I legitimately don't know what you're talking about here. Your dice pool is fixed at however many ranks you have in that skill. And you will never be rolling more than 12 dice, ever.
We picked d12 because, well, we really like d12s.

[Ice9]

And you know, taking the simplest three steps of an 8 step process, not even including the iteratively repeating steps, and pretending that the 8 step process isn't byzantine because PART of it is relatively simple AFTER being deciphered. Yeah. You aren't a raging moron...

You fucking listed:
3) "Take it" in some manner.

As if that were somehow confusing, not to mention acting like "search for the highest die" was an involved process. You're trying to find it confusing, to the point of ridiculousness.

And yeah, the target number is not specified ... in somebody's second-hand mention of the system. I guess that means the system itself is incomprehensible.
In other news, d20 is now an incomprehensible mess because somebody wrote a bad description of it on a message board.

[PhoneLobster]

Recording or remembering your initial value to add other values to is a step. Especially when we don't know precisely how to calculate that initial value AND we are simulatneously removing a corresponding dice from a dice pool that will afterwards undergo further comparisons as part of the complex process.

"Take it" is in fact a TWO OPERATION STEP. In all fairness I should have split it.

But I guess you didn't notice the bit where you TAKE a dice out of the pool... Even though I mentioned it in detail in another post. Before you posted. Once again being a post ahead of you in pointing out exactly what is stupid about what you are ABOUT to say. You have a real talent for that shit.

[Ice9]

Technically correct, but stupid. In practice, an operation like "roll a die pool, set aside the highest" takes trivial time and effort, less than counting hits or adding dice.

Since it's combined with counting hits, the total time is obviously not less. It's not huge, but whether it's worth doing depends on how much information you extract from the roll (it could yield multiple outputs ala ORE, no idea if it does). I'm not going to say it's the best method, but it's nowhere close to as confusing or slow as you try to claim. Obvious hyperbole is obvious.

[PhoneLobster]

Technically correct, but stupid. In practice, an operation like "roll a die pool, set aside the highest" takes trivial time and effort, less than counting hits or adding dice.

That is not how operational complexity works, NOR is it how you clearly and concisely describe a process.

Note that "roll a die pool, set aside the highest". Has a notable comma in it. Thats a separation of steps right there. Further "set aside" is a an attempt at obfuscation on your part to pretend that "record result" and "remove dice from pool" are the same thing. Which unfortunately they are not. And indeed by attempting that observation you leave us with a description that does not explicitly make it clear that the dice has been removed from further calculations involving the dice pool.

Finally you refer to "the highest" without telling us to actually go perform the comparison operation to calculate the highest number. In an attempt to downplay the importance and complexity cost of this operation. The MOST complex part of the FOUR STEP operation you are obliquely referring to.

To be ABSOLUTELY CLEAR AND CONCISE. Your example text should translate to...

1) Roll a pool of dice
2) Determine the highest result
3) Record/Remember highest result
4) Remove any one dice with the highest result from the pool.

At a stretch I would be prepared to throw an "AND" in and roll 3 into 2. But really any time you include an and or a comma in a distinctly stepped process you are basically describing two steps as one and that gets muddled. I am quite serious, even a simple single die roll and comparison mechanic, if described as numbered individual steps will and SHOULD come out as a bunch of steps.

Meanwhile YOUR shortened description leaves too much to assumptions and implications.

You also then describe collecting and rolling a number of dice greater than 1 as "trivial" complexity.

And also include collating and comparing results to determine the highest result as "trivial complexity".

These are NOT TRIVIAL OPERATIONS. Rolling multiple dice is MULTIPLE TIMES more complex than rolling 1. Reading off the numbers from multiple dice AND comparing those numbers through some form of sorting algorithm (and you use one no matter how subconsciously) is MULTIPLE times Multiple Times (yes two multiples!) more complex than just reading off the result from 1 dice.

These operations are significantly more costly in time and computation resources in comparison to alternative single die mechanics.

That complexity MIGHT be worth it. But considering we aren't actually even half way through the full basic mechanic routine at this point THAT is unlikely. And it would remain a SIGNIFICANT burden of proof for any advocate of this mechanic to explain just exactly WHY such multiplicative computation and time costs are required.

[Murtak]

That xd6+y mechanic definitely sucks. I can already tell that it is incredibly easy to have tasks that a party member can not succeed at, namely anything with a TN of 7+. Heck, if you have anyone with, say, 4d6+4 at the table he is very unlikely to ever roll below 15 or so. At the same time someone with 3d6 is unlikely to ever roll that high. I can't imagine this being intended.

Bonus suckage points for having a RNG that curves more and more the more skilled you get. This may actually be intended, but also makes it very hard to figure out what a +1 does to your chance of success.