TheRPGSite

I've really hit a stumbling block coming up with a good core dice mechanic for my homebrew game.

There are four levels of stats/ability scores: Poor, Fair, Good, Exceptional

There are four skill levels: No skill, Basic training, Advanced training, Mastery

The idea behind the game is to make things as simple as possible, and to remove  a lot of the wacky randomness that can occur with too much reliance on dice.

My basic requirements/preferences at this stage are:
-D6 only.
-The fewer dice the better.
-Bell curves are better than linear distributions.
-I don't want to add too many dice together. Adding two dice is fine, three is pushing it and four would be too many.

I'd like most rolls by a player to involve between 2 and 4 dice.

I'd also like to make it scalable so I could do more powerful creatures or supers by ramping up something (like increasing the number of dice or boosting the target numbers).

A few things I was thinking of:
2d6 + stat + skill vs TN: This has a problem of the range of dice results being greater than the range of stats and skills.

Difference Dice: Roll 2d6 and take the difference between the highest and the lowest. I like the bell curve that this generates but I can't quite get how to scale this for more powerful characters.

Dice pool: One die for each point in a skill with either a fixed TN or using the stat as a TN somehow. This option is scalable, but I haven't quite figured out the details yet.

Help me Obi Wan! Throw some ideas my way.

-Jeff

Quote from: GoOrange-Jeff

Hey Jeff:

Dice pool - Roll Skill dice under Stat TN.
Count successes, add situational modifiers like weapon damage.
Vary dice to scale - d10 = standard, d8 = cinematic/Pulp, d6 = Low Supers, d4 = Supers.

That OK?

-clash

You could check out the CODA system used for Star Trek and LoTR, I guess. That system used 2d6, didn't it? I think Classic Taveller used a 2d6 system as well. These might give you ideas for your own system.

Roll Under the square root of Skill (1-3) squared x Stat (1-4) on 1d6

Increasing levels of difficulty could be represented by different dice types or a dice pool roll.

Say: Difficulty 1-5 dice with 3 being average and requiring a success on all dice?

Let's get down to 1d6 shall we?  Make the target number Skill squared times Stat divided by difficulty and devise a bidding system for handling the fractional points.

What?

Quote from: flyingmiceHey Jeff:

Dice pool - Roll Skill dice under Stat TN.
Count successes, add situational modifiers like weapon damage.
Vary dice to scale - d10 = standard, d8 = cinematic/Pulp, d6 = Low Supers, d4 = Supers.

That OK?

-clash

I'd really like to stick with d6, although if I used others it would certainly open up more possibilities.

I've thought about the skill= # of dice and stat=TN. With d6, the target number breaks down if someone has superhuman strength, for example. If the TN is 7 every die will automatically be a success. Scalability becomes tricky if I stick to the single die type.

I'm still considering some type of dice pool, I like the idea of having certain circumstances give bonus dice.

Maybe a roll and keep system like L5R.

Another idea I liked was a roll (skill)# of dice, keep highest result.

Quote from: David JohansenWhat?

You know you're insane, David! :D

-clash

Quote from: Sacrificial LambYou could check out the CODA system used for Star Trek and LoTR, I guess. That system used 2d6, didn't it? I think Classic Taveller used a 2d6 system as well. These might give you ideas for your own system.

I'm not familiar with CODA, can you brief me on the basics of task resolution?

I've thought about Traveller, and I really like the 2d6, but the problem is it produces a spread of numbers between 2 and 12, where the combination of stats and skills could at most provide 2-8. Therefore in this instance the random roll eclipses the game stats. I could just use 1d6, but that's starting to become boring.

Quote from: GoOrangeMy basic requirements/preferences at this stage are:
-D6 only.
-The fewer dice the better.
-Bell curves are better than linear distributions.
-I don't want to add too many dice together. Adding two dice is fine, three is pushing it and four would be too many.

As an FYI, you need three dice or more for a proper bell curve (unless you create the curve with something like a conversion chart or unconventional reading of the dice).

Point taken John, I should have just said I prefer curved rather than linear distributions.

Quote from: flyingmiceYou know you're insane, David! :D

-clash

Do the math and tell me I'm wrong. :D

Quote from: David JohansenDo the math and tell me I'm wrong. :D

I didn't say you were wrong, I said you were crazy! :D

-clash

It's hard to argue with someone who starts with the premise that you're crazy.

I didn't say that was a bad thing either! :P

-clash

Quote from: GoOrange...
I've thought about Traveller, and I really like the 2d6, but the problem is it produces a spread of numbers between 2 and 12, where the combination of stats and skills could at most provide 2-8. Therefore in this instance the random roll eclipses the game stats. I could just use 1d6, but that's starting to become boring.

In my experience playing Classic Traveller a +4 on 2D6 is pretty much a guaranteed success, because the base target number is not 0.  The non-linear nature really makes a +2 or +4 much more determinative than one would think from linear mechanic experience.  I guess my view is in a non-linear mechanic you don't really need the modifier to span the whole range of rolls to get a highly skill determinative system.  I've seen this as a complaint of Classic Traveller not random enough, "high skills" (+4 mod.) make missing near impossible.

I like using the different types of dice for levels of "heroism" you could also do the same with levels of skill.  I can see of course if you want to stick with D6 for design and ease reasons.

Quote from: XantherIn my experience playing Classic Traveller a +4 on 2D6 is pretty much a guaranteed success, because the base target number is not 0.  The non-linear nature really makes a +2 or +4 much more determinative than one would think from linear mechanic experience.  I guess my view is in a non-linear mechanic you don't really need the modifier to span the whole range of rolls to get a highly skill determinative system.  I've seen this as a complaint of Classic Traveller not random enough, "high skills" (+4 mod.) make missing near impossible.

I like using the different types of dice for levels of "heroism" you could also do the same with levels of skill.  I can see of course if you want to stick with D6 for design and ease reasons.

Intersting take on Traveller, I hadn't thought of it that way. The lower the target number, the more important the skill level becomes (with a 2d6 roll).

Still this breaks down again with opposed rolls, where if two people, one with +2 and anothre with +4 square off, their individual skills will be less of a factor than their die rolls.

When I described the various skill levels, I deliberately didn't apply a number because the actual numerical value doesn't matter to me. Just because there are only 4 levels of skill deosn't mean they have to be valued at 1-4. I could increase their numerical value to make that greater compared to the range of dice results. For instance:
No skill = 0
Basic = 4
Advanced = 8
Master = 12

These values added to a 2d6 might produce a better range of results for my purposes, so that someone with skill mastery would have a clear advantage over someone with basic training.

This might work well with abilities too, giving poor=3, fair=6, good=9 and exceptional=12. For supernatural levels (cybered reflexes, a vampire's speed, or super-strength) I could just have increasing #values. So a vampire could have a speed of 15 compared to a human's max of 12. Superman could have ridiculously high strength of 50 or similar. Mostly I would want to target things at human level action, so the exceptions aren't as big of a deal.

I should note that I am undecided how much to tie in stats to skill rolls in a system like this, if I should do a stat + skill or just skill + roll, with skills having different costs based on your stat. If I used a dice pool method, I could consider Clash's idea of using one for the pool and the other for the TN, but I'm still hashing this out.

Different types of dice

I'm not terribly opposed to using different die types when all is said and done, (but I really like limiting things to just a few d6). If different dice are used, however, the system would have to be one where only one of each type is used at a time (other than 2d10, which most sets of dice come with). One thing I don't like about a system where each skill level is represented by a die type (like basic=d6, advanced=d8, mastery=d10) is that the really skilled types can still get hosed with a low roll. Savage worlds fixes this by adding the wild die, but I'd rather not completely rip off another game.

I'm not trying to overly criticize anyone's ideas, just bouncing thoughts around in my head until something really clicks.

Quote from: GoOrangeI'd really like to stick with d6, although if I used others it would certainly open up more possibilities.

I've thought about the skill= # of dice and stat=TN. With d6, the target number breaks down if someone has superhuman strength, for example. If the TN is 7 every die will automatically be a success. Scalability becomes tricky if I stick to the single die type.

I'm still considering some type of dice pool, I like the idea of having certain circumstances give bonus dice.

Maybe a roll and keep system like L5R.

Another idea I liked was a roll (skill)# of dice, keep highest result.

Have task difficulty affect their target number.

Someone with Superstrength (7) trying to lift a heavy item (let's say something 250lbs is -2) makes his target number 5.

You'll have to figure out what's reasonable through playtesting though.

Automatic successes are a good thing...they speed up gameplay!
 So easier tasks would provide a bonus to the roll. Mr. 3 strength trying to pick up a dog (a +5 to the roll we'll say), won't have to roll to lift it...but Mr. 0 strength does.

Quote from: GoOrangeThere are four levels of stats/ability scores: Poor, Fair, Good, Exceptional

There are four skill levels: No skill, Basic training, Advanced training, Mastery

Use stats to determine dice rolled:

Poor - Roll 3d6, drop highest
Fair - Roll 2d6
Good - Roll 3d6, drop lowest
Exceptional - Roll 4d6, drop lowest 2

...for a total between 2-12 (roll xd6, keep 2).

Use skill as bonus to the roll:

No skill - +0
Basic training - +2
Advanced training - +4
Mastery - +6

For Good and Exceptional Stats, add another +1 for each dropped die that rolls equal to or greater than (6 minus Skill modifier).

Set Target Numbers appropriately:
Ex.:
13 - something that someone with Fair stat and skill Mastery would still have a 58% chance of success at
12 - No skill, Fair stat, about 3% chance of success
10 - something that someone with skill Mastery and Exceptional stat would never fail at
8 - something that someone with skill Mastery would never fail at
7 - something that someone with Fair stat and no skill would still have a 58% chance of success at

Quote from: GoOrangeIntersting take on Traveller, I hadn't thought of it that way. The lower the target number, the more important the skill level becomes (with a 2d6 roll).

Still this breaks down again with opposed rolls, where if two people, one with +2 and anothre with +4 square off, their individual skills will be less of a factor than their die rolls.

I guess it depends on how you view the situation / odds.  A +2 vs. +4 (if taking a differential is a 4 point swing) so if the normal target number is 7 say, one guy will have a 5 and another a 9.  Pretty big odds difference on 2D6, maybe not enough if one for what you are thinking.

QuoteWhen I described the various skill levels, I deliberately didn't apply a number because the actual numerical value doesn't matter to me.

I understand but this is where you can start getting the results you want, if say a no skill guy versus a master gives a 6 point odds shift and 7 was the normal target number(TN), one guy would have a TN of 1 and the other 13, so one will always fail and the other always succeed.  

QuoteJust because there are only 4 levels of skill deosn't mean they have to be valued at 1-4. I could increase their numerical value to make that greater compared to the range of dice results. For instance:
No skill = 0
Basic = 4
Advanced = 8
Master = 12
...
This might work well with abilities too, giving poor=3, fair=6, good=9 and exceptional=12. For supernatural levels (cybered reflexes, a vampire's speed, or super-strength) I could just have increasing #values. So a vampire could have a speed of 15 compared to a human's max of 12. Superman could have ridiculously high strength of 50 or similar. Mostly I would want to target things at human level action, so the exceptions aren't as big of a deal.

I agree, no nee to limit them to 1-4, but if you make 1=4, 2=8 for example why not just have the skills gor from 1-12?  I realize your trying to keep it simple, but would it be simpler to just go 1-12 than add in a layer of conversion?

Have you considered a 0-7 range similar to Traveller?  If you make the TN a 7 on 2D6 and take a differenential a 0 skill person will have no chance against a skill 6, a master level individual.  It falls right out of the skill levels and doesn't add much more complexity than skills form 0-4.

QuoteI should note that I am undecided how much to tie in stats to skill rolls in a system like this, if I should do a stat + skill or just skill + roll, with skills having different costs based on your stat. If I used a dice pool method, I could consider Clash's idea of using one for the pool and the other for the TN, but I'm still hashing this out.

On range and numbers for 2D6 I think you might like some of the gearhead work done in Traveller as a place to get ideas.  The range of skills, effect of abilities, difficulty of task (both target number and dice number approaches) have been extensively worked out.  That is, charts generated with actual odds for many different comparisons.  

I'm not trying to sell Traveller (although would be happy to :)) but it seems all the questions, especially given the number ranges you are thinking of, have been raised and analyzed extensively by Traveller gearheads so you can make use of all their analysis to help you decide on the approach you'd like to take.  If you posted a question on the Classic Traveller or General forums over at Citizens of the Imperium I'm sure someone can point you in the right direction or I can try.  There was some extensive work by one Supplement 4 over there last year (before the server crash and change) that went over all this in exacting detail and critique.  I just couldn't find it with  5 minutes of searching this morning.

QuoteI'm not trying to overly criticize anyone's ideas, just bouncing thoughts around in my head until something really clicks.

Cool.  I don't mind.  Being critical is good, especially of the way things have been done that's how we get something that is new and better IMHO.

All the talk about Traveller convinced me to go with dice pools. The last thing I want for this game is a Traveller feel. I really like the Traveller mechanic and implement is practically as a default for most games I write up, but this one I want to have a much different, rules-light feel.

Abilities are going to be rated 1-4, corresponding to the number of dice you roll.

When you roll a pool of dice, you select the single highest result as your final result. Additional sixes are added on as +1.

Skills aren't going to be a central focus of the game, just an area for characters to have a few exceptional abilities (with only 4-5 skills each). I think I'm going to limit skills to only two levels, basic and expert. Skills will either give bonus dice (+1 die pr +2 dice) or static bonuses to the end result (like +1 to total or +2 to total).

I'm thinking target numbers will vary based on the difficulty of the task, 5 difficult, 6 very hard, and 7 for near-impossible. Anything easy, routine or mildly difficult will be a gimme. Rolls should only be for difficult tasks or opposed tasks.

I'm aiming the game a few genres such as urban fantasy (monster hunters), street level superheroes (like Daredevil), and fantasy. Most of a character's interesting qualities will come from special abilities akin to class abilities in D&D, basic powers, spells or other advantages (like advantages in GURPS, qualities in unisystem, edges in Savage Worlds).

These special abilities would be represented as dice pools similar to stats like strength.

The idea is for a light game where character creation largely focuses on a few tricks that your character can do. Basically, like a supers game where you pick out the basic powers/special abilities first, and then fill in the details like stats and skills after.

Right now I'm still mulling over the whole part about how skills work and target numbers.

Quote from: GoOrangeI'm not familiar with CODA, can you brief me on the basics of task resolution?

I've thought about Traveller, and I really like the 2d6, but the problem is it produces a spread of numbers between 2 and 12, where the combination of stats and skills could at most provide 2-8. Therefore in this instance the random roll eclipses the game stats. I could just use 1d6, but that's starting to become boring.

Orange, I'm sorry I took so long to get back to you. We have so many threads on this site now, I get lost sometimes. :deflated:

It's been a long time since I've read it, and I don't have my Lord of the Rings core book handy, so forgive me if what I relay to you is mistaken. Basically, Decipher (the game company) designed what they called the CODA system. I think it was a 2d6 system used for games such as LotR and Star Trek. The LotR book had beautiful graphic design, but terrible editing and organization, plus some rules problems to boot. The core of the game was still good. :)

The game uses d6, has various statistics, like Strength, Wits, etc., as well as skills and 'edges' that function much like feats. You use 2d6 to determine attributes, rather than 3d6, like D&D. Weapons usually inflict 2d6 damage, plus a modifier. A few weapons inflicted 1d6, and others 3d6. Again, I may be misremembering. I can't find my book. The damn book is here somewhere! :mad:

There aren't really levels. LotR has 'advancements' which you use to improve your stats and abilities. The total "hit point" pool is segmented into health levels, plus there's weariness levels. Hardcore physical activity can cause penalties to certain actions.

The game had the following classes: barbarian, craftsman, loremaster, magician, mariner, minstrel, noble, rogue, and warrior. :)

There's lots of stuff I'm forgetting. The game had many problems, but tons of potential. If Decipher had gotten their act together, by playtesting and marketing the game properly, they actually could have given D&D a run for its money....but they blew it. It doesn't help that the Tolkien Estate severely limited what material Decipher could put into the rpg. It was a great opportunity gone to waste. :(

There's a very high-quality fan site for LotR CODA called //www.halloffire.org , but strangely enough, I can't find their message boards any more. I remember some conversation there once about them possibly being forced to take their website down (but don't quote me on that). Maybe there you'll find what you're looking for. :)