Let's Talk About Core Game Mechanics

[AnotherGuy]

...i like bell curves with degrees of success; you can simulate it with flat rolls + tables, but three-dice pools handle it automatically if you're willing to step through exotic dice chains for skill progression...

(i.e. 3d4 -> 3d6 -> 3d8 -> 3d10 -> 3d12 etcet)

I have designed something similar but not exactly...
A benefit of system like this is, is you can add things (proficiency, ADV, magical weapons and modifiers) without breaking the system.

 

[Mustrum_Ridcully]

I kinda like bucket of dice systems, even though I don't much like needing to roll dozens of dice.
But such a system gets you scaled levels of success, and you can also get some reliablity in your skill.
D20+modifiers often feels too random, but smaller dice tend to have insufficient variance.

I kinda get that some people like "roll under" systems, but I think it makes modifiers often a bit more complicated than I like, and once you want to have things like success levels and the like, it has no real benefit anymore.

Personally, I also dislike penalties. I think you should only ever add. Subtraction is hard. Positive modifiers add to your skill roll, negatives add to your difficulty.


But two more outlandish systems I thought about playing with once:

"Everything goes to 12" (or 20?):
A rating like your attribute or skill training level is translated into dice:
The dice system uses something like d12, 1d1d0+2, 1d8+4, 1d6+6, 1d4+8, 1d2+10. (or d20, d12+8, d10+10, d8+12, d6+14 etc.) , So each of these dice can yield a maximum of 12, but the smaller the die, the higher your fixed value. A high rating means you roll a "small" die with a large fixed value.
The idea is that being good at something makes you very reliable, but you might still succeed on blind luck. The fixed value also gives you an idea of what might genuinely be a "routine" task for someone trained, but could still be a real risk for someone without that level of training and talent.


You might usually be called to roll two or three dice, for example attribute die, skill die and maybe specialization/gear die. Favorable additional modifiers come as extra die you can sub into the ones you already have (before rolling or after rolling is still to be decided?)

A expert climber with Body 1d8+4 and 1d6+6 without any special gear would roll 1d8+1d6+10, meaning anything of 12 or less is a routine task, while someone without the training but similar fitness would roll 1d12+1d8+4, meaning his chance is like 16 % or something to fail a routine task.

DCs for "doable" things would range from 4 to 32, but there are basically fundamental limits characters can reach. But I guess.. is that really useful to have? Do we need to have upper limits, does this in any way prevent power creep or power gaming?

Bucket of Dice, but you add all the dice up
You roll a pool of d6, and add the total, compare against a fixed DC. For every 10 points above the DC, you get an extra success level.
In a typical pool system, any die not a sucess simply doesn't count, in this system, they do. Making the success stage up at 10 is mostly for ease of use - Figuring out if you're 10, 20, or 30 above the DC is trivial (maybe even easier than adding the values of a bucket of dice in your head, so it might not be all sunshine, too.)

 

[Koloth]

KISS - Keep It Stupidly Simple. Roll 1 d20. Roll 3d6. Compare to target number. Maybe have one or two modifiers, stat, buff, terrain, opposition banging on door, etc. But keep these very limited. IMO, where D&D 3.x and PF1 become hard is all the buff, debuff and other things that goes into a roll. Many times we had a list on a whiteboard that was 8+ items long. This is in addition to what folks had on character sheets. Then figuring out what does and doesn't stack. Then changing it all when someone casts a spell.

One advantage to 3d6 is it tends to limit skill growth. A +1 on a d20 system always adds a +5% increase. A +1 on a 3d6 adds a variable number. About 12% for 10->11. About 3% for 16-17. None over 18. At some point, most min-max players will decide that increasing a 10 is better then spending that increase on a very minor improvement.

 

[Reynard]

In Sundered Isles (Ironsworn: Starforged expansion), there is an optional rule that (I think) does exactly what you are talking about. The optional rule assigns to the d10stwo moons of the world, Cinder and Wrath.

From the free sampler pdf (pg 7):

Cinder is hot: aggressive, passionate, resolute, physical. Wraith is cool: careful, mysterious, cunning, unearthly. On a match, choose which has the most influence.

Oh, cool. I had not looked at Sundered Isles.

 

[zarionofarabel]

Put me in the d100 roll under camp. It's absolutely perfect having the percentage chance of success written right on the character sheet. Tell me the odds! I also like the d20 roll under as well, though I think the lower granularity can be an issue over very long campaigns if constant growth is something one is after. The roll under mechanics also have lots of variations of how to do critical/fumble options, as well as ways to implement modifiers, and different ways to do opposed rolls. d100 roll under is a very diverse core mechanic.

 

[Reynard]

Another mechanic I really liked from the 90s was Fading Suns "roll high without going over" system. It was technically roll under, but the higher you rolled the more successes you got, as long as you still rolled under the TN.

 

[Bill Zebub]

Who wants to fit story events and outcomes to a normal distribution?

Me.

If all you care about is binary success/failure than a linear distribution (e.g. d20 vs. DC) is fine.

But if you want degrees of success, then a normal distribution gives you the ability to have extreme outcomes that are much less likely than 5% (which isn't that unlikely). For example, with 3d6 the two extreme ends (3 and 18) are a little less than 1/10th as likely as 1 or 20 on a d20.

One way that sometimes gets replicated with d20 is "exploding dice". E.g., roll a crit and roll again. Or roll max damage and roll again. But that's really just a dice pool with a normal distribution...in multiple steps.

That's why.

 

[Thomas Shey]

Another mechanic I really liked from the 90s was Fading Suns "roll high without going over" system. It was technically roll under, but the higher you rolled the more successes you got, as long as you still rolled under the TN.

I've seen this referred to as a "blackjack" system in more than one place. Both Mythras and Eclipse Phase use D100 versions of it.

 

[Reynard]

I've seen this referred to as a "blackjack" system in more than one place. Both Mythras and Eclipse Phase use D100 versions of it.

We used to refer to it as the Price Is Right system.

 

[payn]

Another mechanic I really liked from the 90s was Fading Suns "roll high without going over" system. It was technically roll under, but the higher you rolled the more successes you got, as long as you still rolled under the TN.

Im playing this right now. There is also a focus point pool you can adjust slightly your chances of success. So, the degree of success is kinda baked into the meta point pool.