View Full Version : Planning New Gaming System - What's The Best Dice?

11-13-2010, 03:35 PM
Ok guy's - here's the deal...
I've had this idea to start my own RPG / Wargaming system floating around in the back of my dented, fat skull for ages, and i've been just toying with the idea off and on - ok - granted. I want to devise a relatively simple game mechanics parlay of unit attributes like Hit points 1 (defensive), Hit Points 2 (to cause damage / attacking), and for modifiers - speed rating (to attack / defend), current health or unit integrity, intelligence, whatever... Anyway - Even though I find all of the different shapes of game dice really cool and pretty and junk, would a simple, run of the mill, Milton Bradley grade "d6" be good to run with - or should I go with something more exotic and a lot less "hum drum" to look at? Inquiring minds wanna know. Also - give me some things you might like to see in a new game system - or at least a system that's simple , yet that you can really build on - and that can be very descriptive. I know that a "d6" is the most statistically unbiased (fair) die type, but it's a little "snoozy" for my taste.

I've got a thousand ideas swimming around in my head for game back stories, modules, you name it. It would be really nice if I could even start a rennescance of the old Pencil and Paper type stuff. Then again - maybe I'm a dreamer. Lol.

P.S. - I have never played D&D - or Advanced D&D - but I use to be nuts for collecting books, Dice, Modules, Maps, Figurines, and other game materials - All original TSR - and the hardbound manuals made for awsome rainy day reading. Used to even have the World Of Greyhawk TSR Boxed Edition - and a hardbound copy of the original D&D Players Manual (The "Sky Blue" colored book / the OLD one!).

11-13-2010, 04:03 PM
D6 are a little boring to look at but they'd be fine with me as the basis of a gaming system. I think you're incorrect, however, in your assumption that they're any more "fair" than any other die (at least those with equally sized faces). The chance of any given number from 1 through 6 coming up on a 6-sided die is 1 in 6. The same is true of any other die (i.e. 1:4 on d4, 1:8 on d8, 1:10 on d10, 1:12 on d12, 1:20 on d20, yadda). Are you talking about rolling multiple dice (e.g. 3d6) to generate a bell curve?

As to the particulars of a new system I can't say that I have any input. I'm a fan of generic systems (and GURPS solves that problem for me nicely) but like the occasional niche system that seems nicely tied to a particular setting (Deadlands use of cards and chips, for instance). I just purchased the 4th edition of Legend of the Five Rings (haven't played it yet, maybe ever) and do sort of like what they've done there with the roll-keep system...sort of makes sense to me. D20 is fine with me for playing over-the-top fantasy games such as D&D but I don't really dig it for much else because of the insane power curve.

11-13-2010, 04:25 PM
When properly rolled, any regular polyhedral die is going to produce a level roll table with a large enough data set. That is, if you roll a d6 6000 times, you should come up with close to 1000 entries in each digit. (You likely won't, but the more you roll, the percent of error will go down.)

Some dice, like your d8, and d10, aren't fully regular, but they're close enough to work fine.

Where the choice of dice really comes into play is in the ranges, and the effect of rolling more than one. When you roll a d20, each digit has a 1 in 20 chance of happening. However, when you roll 2d10, both the lowest and highest value only has a 1 in 100 chance of coming up, where as the middle, 11, has a 10 in 100 chance of being rolled.

Now, what does this mean for a game? It means that using the sum of more dice to increase your range of values also changes the rate at which you can expect a number to come up. The more dice you roll, the closer to the middle of your range you are likely to roll. This means that with a large enough number of dice rolled you can have a 'range' of 1000 from highest to lowest, but are very unlikely to roll something more than +/- 5 of the middle.

Personally I've been playing with 2d12 for a rule set. It gives a greater likely hood of being near the middle, but not so much that the 2s or 24s are going to be overly rare.

Check out a tool like anydice.com to see the different ways different dice combinations come out.

11-13-2010, 05:15 PM
i mostly play d20 games at the moment - D&D 4e being the favorite, the argument against the d20 system is what mearrin and talroth have allready mentioned, that the poorest result and the best result happens as often as the middle result, whereas if you use more dice you get a bell curve and therefor more often get average results. We really don't care that much for the reality of the rool as we just wanne have fun and have a fast-paced game. But with 3d6 or 2d12 you'd get a more balanced series of rolls. Some systems gives you extra dice if you're more skilled, so that is also a possibility. And the new Warhammer has special dice with their own symbols one them. Vampire used "botches" so every 1 you rolled cancelled out one succes. James Bond has a succes rating in addtion to the roll, so if you rolled really good, you skill gave a better/faster result. Rolemaster had open-ended rolls where if you rolled 01-05 or 96-00 on d100 you'd roll again and subtract/add to the roll, making it possible for the poorest skilled person to succeed and the best swordsman in the world to miss.
Remember a scenario I played where my character should catch an arrow in flight as part of a circus act (a ruse to get into a castle) - he fumbled and got struck rather hard by the arrow in front of all the nobles. *lol*
Good luck with you endavour :)

11-13-2010, 05:24 PM
Some other things to consider: d6's, being common to mass market board games, not to mention gambling games, are easy to obtain and cheap. They also have much larger faces in the event that you want to customize them with special symbols. They do, however, only offer 6 possibilities per die, if you're running a single die system.

As mentioned, rolling multiple dice and adding them results in a bell curve, which may be more desirable if you want predictable play with a reduced chance of extreme results. The more dice, the more predictable. Rolling a dice pool and counting results higher than a target number is yet another method (this is used by White Wolf's Storyteller system). I'm not sure how the odds play out in that case; I'd have to do some math to figure it out. There are also other dice rolling methods involving multiple dice, such as the One Roll Engine, which is just plain weird. Look that one up for yourself if you want your brain bent a bit.

Personally, my favorite mechanic so far is the one from Unknown Armies. It uses percentile dice (2d10, rolled as though they were a d100, with one die each representing the 10's and 1's digits), and the objective is to roll as close to your skill level as possible without going over, just like The Price is Right. So a skilled gunman not only hits more often, but he can hit for more damage.

It is not uncommon for a dice system to use two types of dice. The West End Games d6 system designates one differently-colored die as the "Wild Die." If it comes up 6, something good happens, no matter if the action was successful or not. If it comes up 1 something bad happens. Likewise, the aforementioned One Roll Engine has special dice to represent a special degree of skill or a particularly unpredictable result. And D&D, of course, uses a d20 to determine hit or miss in combat and a variety of other dice to determine damage, depending on the weapon used.

There is tons of game design theory over at The Forge (http://www.indie-rpgs.com/forge/index.php). They sometimes come across as a bit snobbish about their own philosophies (or they used to; I have no idea what the atmosphere is like there nowadays), but there's no doubt that you can get plenty of good information from them.

11-13-2010, 05:46 PM
Hmmm - all excellent info folks. So what I gather - thus far - is that I should look into either a 3d6 methodology, or a percentile vs. probability approach with a 3d10 approach (1s / 10s / 100s) or a
2d10 (1s / 10s = n/100). I'm definitely going to have to dig back into my old algebra book to calculate model ranges. Talroth - I am definitely taking notes here, lol.

Ok - I'm going to ask a stupid question now. Lets say that two or three d6 are rolled to determine an outcome whose chances of occurring are 1 to "n", the minimum outcome for a 3d6 roll is "3" if they are rolled simultaniously - so how can you then use that roll when the minimum die roll value is 3>1? How do you determine (justify) a "1" chance in "n" unless you make some kind of adjustments? Apparantly I'ved missed something here. (remember I'm getting senile with age guys - lol).

11-13-2010, 10:08 PM
I'm not a statistics genius but I think if you do a search on 3d6 it should turn up a probability distribution for each value from 3-18...it's been done and posted a thousand times. Besides doing single or multiple added dice you should also look into "dice pools", where you roll a number of dice (with the number generally higher or lower based on a given attribute) and then either look for a success, or count successes, or add them together for a success.

In Deadlands, for instance, each "attribute" has a die type and a base value. Under that, you have skills that derive from it that you can buy up. Say you have Agility at d10 and a value of 3 and, under that, you might have Shootin' at 5. When you try to shoot somebody you'll roll five d10s and take the highest result on a single die and compare it to a target number. I think the target number is generally 6 to hit a human-sized target with no modifiers - so, if you rolled 2, 4, 5, 7, and 8 then you hit, because 7 and 8 beat your target number.

Deadlands also has "exploding dice" where, if you roll the top number on the die you get to pick it up and roll it again, adding the new score to the previous value. If you hit the top number again you roll and add again. It gives additional benefits for every 5 points by which you beat your target number. It also has the concept of "going bust"...if more than half of the dice come up 1s then you fail and something bad happens.

There are many variations on the dice pool concept - worth checking some of them out.

11-14-2010, 12:23 AM
I really wish I still had my Combinatorics textbook. (The drawback of borrowing textbooks to try and save money)

There is a major drawback to "Dice Pools", where you roll many dice, but aren't summing them together. Apparently you can actually get into issues where rolling more dice, decreases your chances. I had a conversation once about it with someone far better at stats than I am, and I think he referenced the White Wolf system as having this flaw as you gain levels. I think it was your chance of total failure or bad botch rose as you gained level/skill, and thus threw more dice, but minor failure decreased or something. Hopefully someone more familiar with the mechanics knows of this issue and can chip in with a better understanding.

The tool at the website anydice.com seems to work very well for showing you the stats.

For 3d6, you get numbers like this: Number rolled followed by its chance of being rolled out of 100.
3 0.46

4 1.39

5 2.78

6 4.63

7 6.94

8 9.72

9 11.57

10 12.50

11 12.50

12 11.57

13 9.72

14 6.94

15 4.63

16 2.78

17 1.39

18 0.46

As you can see, nearly half the time you can expect to roll a 9, 10, 11, or 12. This is a reason why I'm not the biggest fan of rolling 3 or more dice to get a sum: Your results are too predictable. However they can be great if that is what you want in your system; a wide range of possible options, but a more narrow predictable range it is most likely to fall in.

When you compare types of dice rolls you need to consider a few things: What do the numbers actually mean in your system? What do your rolls mean? "Roll X or higher?" or "Roll higher than X", vs "Roll exactly X" makes how you look at your chances a lot different.

If you are rolling a value or better than it, then you look at the target number, and take the sum of the percentages of those values that would beat it. You can then roughly compare that to a dice roll over a similar range. Such as comparing d20 to a 2d12. If you graph the probability you get a triangle for 2d12, and a flat line for the d20. (3 or more dice gets you a proper bell curve.)
If you compare your "Chance to roll X or Higher", you actually get better chances in 2d12 till you get at the upper range. (Past where the d20 can actually roll into) Your chances of rolling a 20 in d20 are a lowly 5%, but over 10% in 2d12. (Of course, you're around 0.7% to roll a 24, far worse to roll absolute max.)

Also remember that while your PCs are more likely to hit at given number in a system, it usually means your NPCs get the same effect. So be careful if you have defenses that creep too high on either side. A game where no one hits anything tends to be boring. As is a game where one side gets the advantage in bonuses that lets them usually hit vs the other side almost never hitting.

11-14-2010, 03:09 AM
In short, there are no "best dice" to use. Different dice and different methods of using those dice, give different results. Depending on what you want out of your game system, different options will do better for you. Replace 3d6 with 2d4+1d10 and you get the same range and average value, but extremes become somewhat more likely. Percentile dice are great if you want to have a fine-grain detail. If you just end up assigning all bonuses and penalties in chunks of +/-5, then you might as well use d20. Whatever you choose based on your initial vision, be prepared to change it as you actually try things out and make adjustments.

11-23-2010, 08:33 AM
Personally I recomend d20 or Percentile since the odds are easier to calculate than multiple d6s. If you must use d6s I recomend a system like Risk or Warhammer where each indivual dice is determined as a sucess or failure on its own. 2d6 is O.K. since the distrubution is well known, 3d6 distrubutions are also not too hard to find. Be wary of systems *cough* tunnels and trolls* which compare sums of 2d6 to 3d6 or 4d6, as 2d6 has very low chance of beating 3d6. 2d6 second highest roll 11 is just barely above the average value 10.5 of 3d6. Increasing the number of dice just makes it worse, since the more dice you roll the more likely you are to come up with a value near the average. If you are just looking for which dice are the coolest looking check out the on-line site of the Worlds Largest Dice collection here (http://www.dicecollector.com/)

12-01-2010, 03:42 AM
Night before last I couldn't sleep. It was a combination of factors, but a major one was that the wheels in my brain wouldn't stop spinning. Which would have been completely awesome if they had been spitting out ideas for my NaNoWriMo entry: I'd have finished that night and been relaxing about it today. Instead, my treacherous brain wanted to work on game design. I got the basic of a system now, and some setting ideas, but it needs a lot of refining still. Anyway, dice: d20s, d10s, d6s. I don't think I'll be adding any more. The d20s are the task resolution dice.

Doesn't really change anything I said above, but I thought I'd mention what I happen to be using on my current design.

Today I decided having a beard with have a game-mechanical effect. That's the kind of thing that should define your game more than your choice of dice.

12-01-2010, 11:29 AM
The character having a beard or the player? 'Cause if I can get some extra damage dice with my goatee, I'm all for it!