New Game Dice Convention : Could you review it please?

[joao.mariano]

I'm working in a dice pool system with what I think is a fairly original dice convention:

Random action resolution requires the usage of a variable number of 6-sided dice (d6). This is called a dice-pool. Dice-pools include two components: A Basic Die (BD) and modifier dice. The players always roll the BD, even if no modifiers apply. The number of modifier dice is the same as the value of the modifier that applies to the situation. The dice-pool is notated as:
-Bn if there is a bonus modifier, where n is the BD plus the value of the bonus modifier. This is a bonus dice-pool.
-Pn if there is a penalty modifier, where n is the sum of the BD with the value of the penalty modifier. This is a penalty dice-pool.
-The BD if no penalties apply to the situation.
Once you have sorted out the number of dice in your dice-pool all you have to do is roll it and:
-In the case of a bonus dice-pool, roll the dice and retain the highest value rolled.
-In the case of a penalty dice-pool, roll the dice and retain the lowest value rolled.

Could you please tell me if this dice pool convention is original or if coicindentaly there is anything like this somewhere?

[Jeph]

First off, Welcome to the Forge! (tm)

Secondly, dice conventions barely matter. In my experience, there are two types of dice mechanics: stupid, and easy. As long as you've got an easy mechanic (which you have), you're good to go.

(An example of a stupid mechanic would be, say, roll 2dx, multiply them, divide by Obstacle Number, and try to meet/beat 10, with your Degree of Success being your final result over ten, rounding down. Horribly complicated and way to time consuming.)

Thirdly, originality barely matters, either. Nobody will give a flying glob of monkey's dung whether your rolling mechanism is similar to 500 other games out there if the more interesting parts of the game are exciting and fun. A great example is Jake Norwood's The Riddle of Steel, which uses a rolling mechanism quite similar to the World of Darkness method. However, TROS's combat system, sorcery system, Spiritual Attributes, and other coolnesses make that point irrelevant.

But to answer your question, no, I've never seen a published game with exactly the mechanic you describe. SILCORE could be said to only use the Bn half of your system (roll take high), so you might want to give that a look-see for inspiration. Also, I've used a roll take high/low combined with rollunder on various occasions. But, as I said before, originality in dice conventions is hardly what makes or breaks a game.

Good luck,
--Jeff

[Valamir]

I don't know that there is a specific published game with this idea as the central one to its die mechanic.

But the idea of handling die pools that go negative by rolling a pool and keeping the worst is about as old as die pools themselves.

[Sindyr]

I am afraid to ask, but into what category does the following fall:

4 six sided dice with the faces numbered 0 - 5, 4 of these generate a 0 - 20 spread with 10 being the average. (and median, and mean)

Roll = 4d(special6) + stat + skill +/- mods vs a target number of 20.  Amount under or over 20 = relative degree of success/failure.

[average stat + average skill = 10]

[MachMoth]

Bah!
Done or not, if you came up with it on your own, be proud of it.  If you find out it has been done before, well, great minds think alike.

The general agreement here is that dice mechanics are a maintenance factor.  They can ruin a game, but do little to make it more fun.  While I can't say I fully agree with that, it does stand that dice mechanics aren't worth losing sleep over, when other things are still left to do.  So, on a good day, the best compliment you'll get is a "I don't hate it."

As a forewarning, the forge is a difficult place to strive for innovation.  Not to say you shouldn't try, just know what to expect.  Between everyone here, they've probably seen 99.99% of the most obscure work on the planet (and there is A LOT out there).  Even if they haven't seen it, at some point someone (being designers and all) has probably said to them "hey wouldn't this be interesting."  So, trying to find something they haven't seen is generally an effort in frustration.  The general goal is to produce a quality work, with game that hasn't been done before.

So overall, good job (I mean, I don't hate it :P) and keep up the good work.

[Sindyr]

Was that (in part) a reply to me, or only the original poster?

[MachMoth]

Sorry,
I took my time making the post, so I didn't see yours prior it.

[Darcy Burgess]

They can ruin a game, but do little to make it more fun.

I've noticed this sentiment as well.  And much like you, I'm not 100% in agreement.

A case in point is Greg Stolze's One Roll Engine (Godlike et al.)  I've never seen anything else quite like it before, and I believe that it is a large part of the reason that Godlike is as fun a game as it is.

The "one roll to rule them" ethic of the system speeds and simplifies task resolution, and also carries echos of TSR's cool saga cards -- there are so many options available to the clever GM in interpreting a single die roll.

This is not to dismiss the excellent game design outside of the die mechanic, but this is a case in point where the right die mechanic makes a big difference.  Task resolution and timing are nicely handled, and with elegance.

[Walt Freitag]

Hi João, and welcome to the Forge!

See this thread for a discussion and critique of a very similar mechanism.

Over The Edge uses a mechanism like this, as discussed in that thread.

In the mechanism you describe, the first extra point of bonus or penalty modifier makes an enormous difference in the odds of success for most target numbers. This may or may not be desirable. Also, if the target number that you're comparing the result of the roll (highest or lowest die) to also varies, then the effects of a bonus or penalty are wildly different depending on the target number (see the probability table and discussion in my post near the end of the above-linked thread), so deciding whether a given modifying factor in play should change the target number or add bonus/penalty dice to the pool can become a complex decision. (This is not an issue if the target number is fixed, or if all rolls are opposed -- that is, the result of a roll is always being compared against another roll. However, because it gives each side only six possible results --a number from 1 to 6 -- your system will produce a lot of ties with opposed rolls.)

A different but related mechanism that has a similar structure (base die with a pool of modifier dice, read differently if the overall modifier is positive or negative) is described here. One variant goes like this: the base die is a d10, all bonus/penalty dice are d6s. BD: 6 or higher succeeds. Bn: success if any die shows 6 or higher. Pn: success if no die shows 6 or higher. This gives a base chance of 50% with a very smooth and gradual change in the success chance as bonus or penalty dice are added. To make each point of bonus or penalty stronger, use a d8 as the base die and read 5 or higher as the indicator of success/failure.

- Walt

[Walt Freitag]

Sindyr, that sounds like a fairly conventional roll-under mechanism. You're using multiple dice added together, which gives you a bell curve (results near the middle of the range are more likely than results at the extremes). But it's still a single roll, not too different from rolling one larger die. (Some people call it a dice pool when multiple dice are rolled and added together, but the most common usage is to call it a dice pool only when each die is read separately in some way, and usually only when the number of dice can vary.)

There are many well-known systems that use 3d6 added together, including Hero System and GURPS. Interestingly, there's not much difference mathematically between the behavior of 3d6 summed and 4d6 summed, except at the rarely-occurring very highest and very lowest results -- and that with 4d6 there's no target number that can give you an exactly 50% success chance.

Also, you really don't need the special dice. Nothing changes in your mechanism if you use normal d6s and set the target number at 24.

- Walt

[Mike Holmes]

Aw, Walt beat me to all the good stuff.

(Some people call it a dice pool when multiple dice are rolled and added together, but the most common usage is to call it a dice pool only when each die is read separately in some way, and usually only when the number of dice can vary.)

The reason for the terminological problem is that d6 AKA West End's Star Wars, used a pool of added d6s. The size of the pool was variable, but they were read as one sum. They called this a dice pool. So, you've got set numbers of dice with sums (this example, GURPS, Hero, etc), variable numbers of dice with sums (D6), and variable numbers read separately (WOD, TROS), and even set numbers of dice read individually (that game of Vincent's and the other one like it :-) ). Just as broad categories. It's best not to use the term die pool without a lot of clarification.

and that with 4d6 there's no target number that can give you an exactly 50% success chance.

In general, that's true with any even number of dice, wheras odd numbers give you that 50% - but have no median score. Hence why 10-11 is "average" in D&D (10.5). With 4d6, the median is 14.

Also, you really don't need the special dice. Nothing changes in your mechanism if you use normal d6s and set the target number at 24.

Assuming all else is equal. There are potentially some good reasons to base off of zero if the system has secondary requirements. Zero based attributes, for instance can demand this. I like dice that start at zero. I also like FUDGE dice (d3-2). Probably the programmer in me thinking in terms of zero based arrays and such. Also, it's easier to make these dice using a Rand statement than ones that start at one.

Mike

[Andrew Martin]

I am afraid to ask, but into what category does the following fall:

4 six sided dice with the faces numbered 0 - 5, 4 of these generate a 0 - 20 spread with 10 being the average. (and median, and mean)

Roll = 4d(special6) + stat + skill +/- mods vs a target number of 20.  Amount under or over 20 = relative degree of success/failure.

[average stat + average skill = 10]

It's equivalent to:

Code:

4D6 - 4 + stat + skill

which was used in a TSR SF game, IIRC.

[joao.mariano]

First, thanks for all your welcome feedback and sugestions. It's great to have such an input from the best indie rpg designers or theorists.
Second, I must give credit to the author of this particular version of dice mechanic. It was created by Sergio Mascarenhas who gave me the permission to work it out for a new indie game i'm developing. Sorry but it wasn't my intention to seem like I had designed it.
And you're right, I should have posted the full concept of the game i'm designing because it doesn't make sense to only post the dice mechanic. Let me flesh it out more and I will do it!

[joao.mariano]

Also, if the target number that you're comparing the result of the roll (highest or lowest die) to also varies, then the effects of a bonus or penalty are wildly different depending on the target number

In fact I was thinking of a target number between 1 and 6 for static actions (scaling a wall or something like that). 1 would be a fumble. 2 would be failure. 3 a marginal sucess. 4 trough 6 would be a gradual level of sucess. I would use called rolls or read the results in the dice.
I would used contested rolls for dynamic actions (pc vs pc, pc vs npc). Whoever rolls highest wins if he has sucess.
The negative modifiers would run from 3p to 1p and positive modifiers would run from 1b to 3 b.
How do I calculate probabilities for this? How do you do it yourselves?
And should I get rid of the fumble mechanic?

[Andrew Martin]

And should I get rid of the fumble mechanic?

The answer to this question depends upon the type of game play you want. Does the game involve the PCs (and their players) looking like fools and idiots? If so, then keep the fumble rules, and perhaps increase their frequency. If not, then remove the fumble and failure rules.

How do you do it yourselves?

Some people write out every possible combination of the dice and sum the individual results. Other people write computer programs or spreadsheets to do this for them. Other use a computer program to roll simulated dice 10,000 times and record the frequency of each outcome. Some people use statistical theory and calculate the probabilities of each outcome.

How do I calculate probabilities for this?

Choose one of the above methods. :)