Have gotten  he game design bug under my skin again.

My last incarnation of a simple game mechanic used 2d6+1dX+N vs. DC, where 1dX is the value of a Stat in dice from 4 to 12 and N is a positive integer from a skill. And DC is an arbitrary integer to calculate success or failure

Current thoughts on simplifying this-

I like 3dX because of the pleasant bell curve it produces. I also have a very very vague skill list, which I believe may be able to allow me to use the skill list as a form of stat, effectively combining the two categories.

The current model goes like this-

You have the four main skills Combat, Athletics, Communication, and Survival all with ratings of 3d4.
You bump these skills up by swapping one die at a time with the next higher die until all three are upgraded. so 3d4, 2d4+1d6, 1d4+2d6, 3d6, 2d6+1d8, so on an so forth.

You can make up any additional skills fitting to the setting and game and spend points upping those as well, and by writing them down have them at 3d4. Someone that doesn't have a special skill (say Knowledge Arcana) would pretty much auto fail any roll in that category.

Probability wise, this stepping will give a character  steady growth of +2 to max roll, and a +1 to their average roll. Everyone can still screw up by rolling all 3's but that chance drops from 2% on 3d4 to 0.05% on 3d12.

Pros-
Dead simple, one tier rolling
Bell curve for graduated success calculations

Cons-
Need multiple dice
Number range is bounded by dice
Will need additional work to bring about character definition, but i find this is true for any system and falls into some form of 'perks' or 'traits' class anyways

Main question, Has this been done already and how did it turn out?

Hiya,

I haven't seen this before! It looks nifty.

The question is whether the fine-grained aspect of the curve - as opposed to the basic 3d6 or 2d10 ones, which have served quite well over the decades - is worth the effort. That lies in the context of the roll, both in terms of setting and more importantly, of the kinds of distinctions you'd like to see among characters.

My only advice is to avoid all modifiers that affect values rolled, like -1 or +1, and so on, and have all modifiers affect the dice themselves. So if you plan to have a difficulty modifier in your system, call it something like "Drop two dice by one level," making my 3d8 roll into d8+2d6, for example.

Best, Ron

Oh wait - I have to ask that you create an external document and link to it. Requirement of the forum.

Best, Ron

i really don't have anything like a coherent doc at this time. Most of my thoughts are largely in that first post.
https://docs.google.com/document/d/1jYSIdP6lqggdYLTfjqU22ka5efXzXSToOXEHpcoygVY/edit?authkey=CKj5q5QL&hl=en_US
there place holder that I'll update as things go along

Some more thoughts. Going with the 3dX as a base, there are many ways to tweak the rolls or their meanings to extract useful data, hopefully with minimal work. One thing I want to bring up in the game is using the bell curve to generate varying degrees of success or failure. It won't be for every roll, some times a simple pass/fail is all that's needed.
Tweak ideas-
-/+ Integers:  This is quick, simple, highly functional, and very boring. By adding we increase the minimum, average, and maximum uniformly.

Extra dice, pick 3:  Something i want to make very core in the game is adding extra dice to a roll and letting the roller take the three highest. Depending on the dice used, this largely leaves the maximum value alone, doe not effect the minimum, but greatly increases the odds of rolling above the normal average of the three dice.

-/+ Dice ranks: I believe this would be a useful tool, mainly in the reduction. Reduced ranks do to injury or impairment come to mind, though I don't want to pigeon-hole a mechanic with story elements yet. Likewise additional ranks seem like a good way to clump up 'team work' type events into one roll.

So far, I'm leaning strongly to having a 'skill focus' that characters will pick after fleshing out their dice ranks which will be a specialization on some smaller facet of a skill. This focus means they do that particular task more reliably than other similar skills and would there for get +1dX where X is the highest die type in the parent skills rank.

If a character had a particular skill at say....1d8+2d6 they would throw a number between 10 and 13 43% of their rolls
If the same character used a focus in said skill they would pick the top 3 from 2d8+2d6 which means they would be throwing 14-18 on about 46% of their rolls.

One thing i'm getting stuck on is how tools will interact with the math. It's easy to say you have tool X or weapon Y so you get a +1 or +2 to your roll, but I think I want to save straight integers for circumstance modifiers so the GM won't have to get tricky in order to impart bonuses or minuses as they see fit. It's possible tools could add their own dice, a kind of insurance, and be tagged with some kind of modifier. By having said tool you can impart a result that would not be permitted without having that tool at hand. The issue arises, that with any tool, un less it is so automated as to take any action from the character (which would mean the player doesn't need to roll) no matter how great it is, if you don't know how to use it, it's junk.

Here, my mind bends to tools not effecting rolls at all, but maybe having some form of modifier to the result curve. Something like 'If you roll higher than the DC by 3, construction takes 1/2 time' or 'Every 2 success past the DC, target takes increased damage'

And after all of that I could still see flipping it all around and having Skill Focus give you a straight +1 or +2 and tools giving you small extra dice to just increase the likelyhood of you doing better than average at something since it won't increase the cap by much.

so for the time being I have a basic roll. Currently I'm going with a roll vs DC. This set up is simple, and easy to compute, but leaves something to be desired in any situation where you have two or more active participants. Combat comes to mind most often. You have the classic d20 set up of roll vs some base + character skill/armour/specials/what not and then the more simulationist roll vs roll.

Translating a character's ability into a stactic DC is very simple with the dice mechanic used. Look out how many ranks they've bumped their dice in a particular skill. Add this to 7. The value generated is on the low side of the median for their bellcurve in that skill. It also means that characters of equal skill level have exactly 50/50 odds against one another.
However the mere structure indicates a flow to the actions taking place, that is one person is acting the other reacting. It also implies a certain granularity into the events taking place. This situation happens most often in combat, so it breaks down to one person swings, or shoots their gun, the next person does what they want to after this action, so on and so forth.
In a roll vs roll arrangement, you remove certain aspects of granularity and event flow. A single roll can be used to determine 1)who wins and 2)by how much they win for a single blow to  an entire encounter. It also breaks the roll free from the focus of the DC style, that is to say, you can have more people than two roll at once and have meaningful results. You do not have to make rolls agaisnt every participant one at a time.

It should be noted that my roll vs roll differs from other types (most notably the roll under d%) in that stalemates are less common. Two equally skilled characters will only roll a 0 on average 13% of the time at 3d4 and 5% of the time at 3d12. This compared to one player rolling 'to hit' vs another 'to not be hit' which becomes an increasingly long stalemate has both characters become better skilled.