I'm thinking of using a homebrew system for a campaign I want to run, using an extremely open ended character generation method similar to Risus or Over the Edge.

I can't decide on a dice mechanic though.

I definitely want to use only d6s, but don't want to use a simple Traveler type 2d6 + mods type of roll, or a JAGS 4-20 roll. I think I'm going to have skills/abilities/powers etc. all represented by a number of dice or at least a number ranging between 2 and 6 (possibly even up to 10).

A few ideas I was considering:

1. Ratings expressed as #d6. Roll your rating and add the results, compare to target number. This is simple, but makes it impossible for something with a low rating like 2d6 to hit a target number of, say 20, which would be possible with a 5d6 rating.

I like the idea of getting lucky and having chances at near impossible tasks.

This could be fixed with adding in exploding 6s

2. Using successes (ala Shadowrun). Any result of 4 or more on a die is a success. Different tasks require more successes than others. Rerolling 6s is again an option.

3. Roll and keep variation. Roll rating #d6, but only keep the X highest results. I was thinking of making X a static number, like 4. No matter how high your rating is, you only keep four. If your rating is less than 4, you keep all the dice.


Modifiers to rolls could be extra dice (most likely candidate), extra successes, or bonuses to the total roll (+2 etc.).

Any thoughts or suggestions?

Ever played a game that used a similar system and liked or disliked it?

Right now I'm just throwing these things out there to the sounding board to see what strikes everyones fancy (including my own.)

Thanks.

you roll 2d6, one black, one white (or whatever colors you like) subtract the result of the white die from the black die. This will give you a range of +5 to -5 sort of like a fudge scale.

Option one with exploding sixes at d6-1 each makes every task possible for all pool sizes and still has easy math. It isn't intuitive, but exploding like that just adds one to the average result of the die while still theoretically allowing an infinite result. It also means determinging defaul difficulties for actions is pretty easy. If a one die schuck should be able to do it half the time the difficulty is 4. Difficulty 9 for a 2 die average Joe, etc.

In case what I'm saying isn't clear (it wasn't to me when I first used it) what I mean is that if you roll a 6 on a die you roll again, but with d6-1 (for a possible bump of 0-5). Another 6 means another roll of d6-1, and so on. So you could roll a 6, 6, and 2 with each die for a total of 12.

Quote from: James McMurrayIn case what I'm saying isn't clear (it wasn't to me when I first used it) what I mean is that if you roll a 6 on a die you roll again, but with d6-1 (for a possible bump of 0-5). Another 6 means another roll of d6-1, and so on. So you could roll a 6, 6, and 2 with each die for a total of 12.

That was an alternate die method for SR3, due to the amusing trend that difficult 6 and 7 were the same thing.  The issue though is that at each "step" the odds drop down.  If you combine that with multiple dice in a pool, and you get the same issue SR3 had: absolute chaos in the probabilities.

How do you make things tougher: take dice away, require more successes, or raise the target number?  Each of those changes the odds, but in a different and "fun" to calculate way.

Caveat: A lot of folks really like SR3, so that may be exactly the route you want to go.  Myself, I prefer a system that's easier to understand.  I'd probably go with Add Up Traits, roll that many d6.  If you want to be consistant, have penalities take dice away and bonuses add dice.  Any dice at 5 or 6 is a success.   You could also do penalties (and/or difficulty) as required successes, and if you want "long shot" victories then allow for a reroll on 6's, but only when the odds are against you (like you've got 2 dice and need 5 successes).  Otherwise, people with big dice pools will chain off huge numbers of successes.

I'm not talking about having a target number and then counting successes at that number. I'm talking about having a target number and totalling all the dice in order to reach that TN. So if you've got 4 dice you roll them all, add them together, and hope you hit TN 17.

The probability is really easy, because the average die roll under the system is 4.5 per die. Thus each pool size gives a bell curve centered around an average roll of 4.5 * size. Strict eyeballing, which is what most players do, is even easier: if your dice * 4 is higher than the TN, you're in pretty good shape.