Dice in General

[RfC_Pyxis]

Most gamers are probably old-hands at figuring out probability distributions of dice rolling systems (Ye Old 2d6, 2d10, 100d100, etc.), they know that in a roll of 2d6, the most probable outcome of the two summed dice is '7'...

From there it's just a short hop to recognizing that 10d10 is more likely to return higher values than a single fair 1d100 will, and aside from the inconvienience of having to add numbers in your head rapidly, can be used to give 'different' results to rolls than what one usually encounters.

But what thought has been put into using computers for alternate-distribution rolls?  

Being in an area where the gamer population is thin, (Out in the middle of nowhere) and generally tired of playing with the same old bores, I picked up playing MUSHes as my almost exclusive Roleplaying outlet, and in general devoted my spare time to playing one MUSH exclusively, so my experience has been pretty narrowly defined towards the Narrativeist side of things with a liberal splashing of simulationsism.

Getting back to the topic, I spent an hour or three coding into the system of the Mush, a weighted die system, transforming the distrubtion of thier random number genarator from uniform to a (psuedo)gaussian distribution.

For those of you who don't know what I'm talking about, a uniform distribution can be throught of as a single 'fair' die, each roll having the same probability of coming up as any other roll. Or, p(x)=Constant and the definite integral of p() across all possible values equals 1.  

The (psuedo) gaussian distribution basically means that values tend to 'bunch' up around a mean based upon the standard deviation.  All roll values within 1 deviation of the mean occur 34% of the time, 2>d>1 deviations occur 14% of the time, and it gets lower from there, the further out you want to check.   In other words, the distribution graph looks like a bell as opposed to the horizontal line of a fair die.

I call them, 'Hero' Dice.   The player rolls the 'Hero Dice' in instances where a favorable roll would be really useful, say when confronting a life or death situation  -- Our Mu* storytellers/GM's tend towards not liking to kill our players' characters off, despite how annoying they can be sometimes.   The player then chooses the 'mean' and the deviation and 'rolls' the die.  

Now, the system also comes with a 'criticality' modifier, which is just an inverse of the standard deviation and is scaled to the 'die' being rolled.. it helps to limit the amount of math you have to do to set your roll up right...

To limit cheating, it's obvious the die is being used, and each use costs a 'point' from your 'perks point' pool, which are accumulated through Roleplay Scenes submitted to the staff for periodic review (But in general, we plan to give them out without too much fuss).

Why not just use your perks-points to change an outcome?  The idea is that it costs more to definately change an outcome in some direction than it does to take a risk (A calculated one..) and try for it that way.

Unfortunately, the system has yet to be adequately play-tested amongst the players of our MU*, and I'm tempted to see if it or some other similar system of randomization is useful in actual RP.  

What do you folks think?

[montag]

sorry, could you rephrase the question/-s?
do you want to discuss the relative merits of uniform, gaussian and skewed distributions to determine outcomes (outcomes in the sense of ability vs. difficulty or otherwise)?
if so, what would you want to talk about except personal preference?
or do you want do discuss the idea of spending perk points for potential gain or certain gain? (in which case I'd say having both is interesting, though my – inexperienced – guess would be, that players opt for certain outcomes as long as they can afford it, since they'd probably only spend perk points when the result is important to them anyway.)

[Mark Johnson]

I like it.  It must be something in the air.  In fact, my current project allows the player to customize their die curve pretty much however they want.  My game uses fortune in the beginning, and I am pretty sure you are using fortune in the middle, but a very similar idea.

On the other hand, with your idea, if you are simply trying to protect the PCs from an untimely death there may be other ways to achieve it.  For example, burn a perk point for a reroll, which would allow the character to cherry pick a result if they REALLY need a certain result.  Or allow each perk point simply to be a +X bonus to a roll, which eliminates the curvature but still allows the characters to be effective when they wish.  Both of these options are simpler and may achieve the same result (you tell me if they don't).

Talk Soon,
Mark

[Andrew Martin]

I call them, 'Hero' Dice.   The player rolls the 'Hero Dice' in instances where a favorable roll would be really useful, say when confronting a life or death situation  -- Our Mu* storytellers/GM's tend towards not liking to kill our players' characters off, despite how annoying they can be sometimes.

You've mentioned "storytellers/GM's", annoying players, not killing off PCs, and life and death situations. It sounds like your Mu* is becoming disfunctional; not having clear goals of play; a game system that is at odds with the intent of the "storytellers" and with the GMs. Just like a face to face, pencil and paper roleplaying game, Mu* can become disfunctional.

I'd suggest clearly defining what each Mu* is about; it's goal or purpose. Is each Mu* and it's GM about:

[*]which player (or play group) can gain the most from each enounter? Defeating monsters, getting the loot, increasing the PCs powers, defeating the GM's monsters in a fair fight?
[*]storytelling, life & death situations, important PCs, choices that define a character?
[/list:u]

[RfC_Pyxis]

Heh, I got a little rambly there, didn't I?

 The thrust of the post was to discuss the use of 'different' dice in a Roleplaying setting; has it seen any succsessful playtesting?  What were the circumstances under which these different dice were used?  
 Do players think these different dice are 'fair'? (Not fair in the sense of an unloaded die, but fair in the sense that, okay.. that's sufficiently random for my tastes...)

 I attached those extra fiddly bits to describe how the system was set up currently (Burning perks, etc).

Yes, I feel our Mu* has become disfuntional over time, and the dice weren't created explicitly to prevent player deaths, but those are the subjects of another post in the future.

[montag]

The thrust of the post was to discuss the use of 'different' dice in a Roleplaying setting; has it seen any succsessful playtesting?  What were the circumstances under which these different dice were used?  
 Do players think these different dice are 'fair'? (Not fair in the sense of an unloaded die, but fair in the sense that, okay.. that's sufficiently random for my tastes...)

oh, sorry I didn't get that. It's an interesting idea, but I can't think of an application where one goes from uniform to normal (or the other way). Of course any dice pool mechanism does change distributions to a certain degree, and any sort of bonus dice will act in a similar manner, but as I understand it, that's not what you were thinking of. It's an interesting idea, but I think the change is a bit extreme. If you want to have some steps in between, you easily end up with 1d, 2d, 3d and the a dice pool is just around the corner (though you'll rarely use 1d in those IMO).

[M. J. Young]

The thrust of the post was to discuss the use of 'different' dice in a Roleplaying setting; has it seen any succsessful playtesting?

I guess my question is, different from what?

A lot of games use the so-called "straight curve"--a single die rolled against a target number, usually with modifiers.

On the other hand, some (GURPS, I'm assured) use multiple dice summed to create bell curve (gaussian) rolls.

Further, off the top of my head I can think of three distinct applications of the dice pool concept--total successes, total defeats of opposing pool, and highest die wins.

Fudge also creates a sort of gaussian result, with its positive and negative results from multiple dice.

There are some non-numeric dice systems, too, such as the Diverse Lunacy system in Legends of Alyria.

What would constitute "different"?

From there it's just a short hop to recognizing that 10d10 is more likely to return higher values than a single fair 1d100

I'm not sure that's exactly true; at least, what you've stated is not clear enough to mean what it suggests.

The average roll for 10d10 is 55; for d100 it's 50.5. the minimum roll for 10d10 is 10, so if you have to roll at least 9, 10d10 guarantees success, while d100 offers an 8% chance of failure.

But the chance to roll 100 on 10d10 is only one in 10^10, or one in ten billion (I think that's right); and although I'm hesitant to do the math, I would wager that you're more likely to roll anything above 80 on d100 than you are on 10d10, if that's the target. I know you've got a 20% chance to roll at least 81 on d100; I am not sure your chance of success is as good as that on 10d10.

More dice in the mix causes the curve to skew more sharply toward the center.

You can find a lot of this in the dice curves appendix in Multiverser: Referee's Rules, which looks at some of the dice pools which were popular when it was published, as well as various other approaches to die rolling.

--M. J. Young

[Mike Holmes]

Just for reference my spreadsheet tools used a "die" mechanic that was, essentially three dice that went from zero to whatever the rating of the skill was divided by three. So, if you're rating was 30, the range was from 0 to 30 with the bell centered around 15.

Another one that I used was to have a natual log distribution of ratings along the range of the ability. So you'd tend to get results that were near maximum most of the time, with occasional "fumbles". Note that the natural log one was pretty boring. I mean, it was really, really, really, really, really realistic. Which meant that you could well predict outcomes. Which is boring. More random (greater standard deviation) is more fun in a lot of cases.

But, in any case, games already incorporate these ideas in various ways even with dice. With computers, just choose a curve that does what you want, and go with it.

Mike