2d6 vs. 3d6 vs. 1d20 systems, etc

So Veeky Forums, lets talk about dice, probability, and all that good stuff. Lets try to talk about the positives and negatives of each type of mechanic, of which there are of course many.

>2d6
Fast and loose, you see a lot of 2d6 systems for "rules-light" games. Seems to have just enough variability that random chance plays a role, but the extra d6 adds some consistency, not to mention of you get to add an extra dice for whatever reason (bonuses, etc.)

>3d6
that nice bell-curve is always good if you like consistency. Crits on 3/18 offer a rare, powerful critical. Not too many downsides that I'm aware of.

>1d20
We've all probably played D&D, and we all know it. Bigger number = better, add any modifiers, etc. 5e's advantage/disadvantage system adds an extra layer, but at the end of the day 1d20 means you have a 5% chance of critical success or failure in combat, and a lot of people, for some reason, translate that elsewhere.

>xd6
Common in D6-based games, like D6 Space or D6 Star Wars. Fine overall, but exploding dice mechanics are controversial to say the least. Often requires you to have an obscene number of dice for the table, but at least D6 can be purchased 10/$3 or so at a grocery store, unlike other types of dice where you have to hunt a store.

What is everyones' favorite? what are the strengths of roll-over and roll-under? Why should people who use those clear crystal dice sets be banished to the lowest levels of Hell, and those with high-contrast dice be put on the highest clouds of Heaven?

3d6 is the only acceptable choice

why? 2d6 and xD6 games aren't that different in terms of bell curve, albeit 2d6 games are more of a triangle.

I'm starting to like the 2d6 and now making a GURP 2d6 system. So far I made Magic Girl and Slasher Game edition.

xd10 is objectively the best

It's not easily ported to other systems, but I really like Blades in the Dark's mechanic:

Roll xd6, keep the single highest die (except for multiple 6s)

Compare this result to chart:
1-3: bad
4-5: success, with consequences
6: success
Multiple 6s: crit

The only problem is that it only really works for a couple of dice, so you can't have many different modifiers, ratings, bonuses etc, since anything beyond a single bonus die makes a huge deal.

0d6 is done by rolling 2 dice and keeping the lowest one, you can't crit on 0d6.

2d10 exploding up and down provides a very nice probability curve.

If I had to choose one in a vacuum, 2d10 is all I need.
>MUH BELLCURVE REEEEEEE
There is a bellcurve in opposed rolls.

>using physical dice
>not using a computer to generate random numbers

>Adding multiple results
Slows down the game, even if it's only a fraction of a second for players that are good at mental arithmetics. Yes, it's easy to add three numbers together, but it's one more thing to think about, which is espescially taxing on the dm. It also makes it difficult to roll multiple checks/saves/attacks in one go.

>D20
Is a poor mechanical design for gaming dice. It's too round so it ends up rolling instead of bouncing, which increases the interaction between the table and any imperfections on the dice. This increased interactions means that imperfections in a rolling die have a much greater effect on the probability distribution of the different possible results, compared to a bouncing die. The bottom line is that d20s are more likely to be severely flawed in terms of not having equal probability of rolling each result, compared to more angular dice. In my experience d20s from regular brands (i.e. not precision dice) are often so rounded and bulging on some of the sides that they're not just unlikely to land on that side, they're can be hard to just balance on that side. I've never seen that problem in other gaming dice.

my favorite is star wars ffg custom dice

That girl looks like a bitch.
I hope she gets choked to death on a beach somewhere.

I played with so many shit DMs that it ruin d20 for me. No matter how easy the task or how high is your bonus, if you don't roll a 15+, you automatically fail. The amount of DMs I meet with this mentality is mind blowing.

Middle dice of 3d20

I like rolling under and I'm not sure why, It's just more satisfying than stacked up runaway numbers. In general it has a lot more to do with the system's particulars than the dice used by the system to determine whether I like it though.
Except d10s. Fuck nonplatonic solids.

Making a 3d6 game just feels like copying GURPS, though.

>my favorite is star wars ffg custom dice
Fuck you.

I'm just going to say it, the Cortex system and Savage Worlds style system (add skill dice + attribute dice) sucks dick

>savage worlds sucks dick
This is news to no one.

>hey guys, what are the merits of various dice mechanics?
>almost every post is a seething preclusion of conversation
You're the most joyless fucks on this board right now, congratulations.

All dice rolls can be broken down into percentages ,and if you're quibbling at inaccuracies of less than 1% you're autistic.

What else did you expect? Veeky Forums doesn't play games, they just bitch about them.

What? GURPS is 3d6, roll under. We can do 3d6, roll over, 3d6, pairs, 3d6, 5 or 6 as success, 3d6 opposed roll to 3d6 and many other ways.

no u

Personally I've liked dice pool systems the most. It's got the smaller degree of variance like 3d6 but I think it allows more granularity in modifiers.

too soon

Play Hero. It's older than GURPS.

4d3 - 8

makes me think of the Standard normal distribution

I usually go for 2d6 in games I make, because a lot of people probably have 2 six-sided dice in their household. At the same time, the range of results in 3d6 is so appealing. Someone, please, drag me in to the light side!

Im in the same boat user

AGE is 3d6 but one d6 is SPECIAL. It gives you STUNT POINTS to do COOL SHIT with your one roll.

I run D&D 3.X homebrew, and almost all of my tables are done in an exploding d6 format. I just find that the mundane is usually able to be boiled down to (four if one explodes negatively)/(five if not) points, and then beyond that is anyone's guess and shit gets weird on the tables. I also usually have my players roll table rolls for me, so that they feel involved and invested.

2d6 master race.

It's the best middle ground between beer and pretzels and in depth mechanics. You get a simple to interact with curve, ways to play with the variables, opportunity to add or subtract from the die rolls or from the number of dice, you can roll under or over or based on odds, and almost every home in the world has a pair of dice laying around. Just enough dice in hand to feel like your roll matters, but not an unwieldy handful or a weird mix of dangerous shapes. Plus two dice inherently feels a bit classic if you romanticize dice.

Which is better feeling/more exciting for your average Joe:

1. Flat modifiers,
2. Roll X, keep Y (e.g. 5e advantage), maybe given a fancy ass name?

I feel like RPG design is about as much about the psychology and expectations everything else.

*as much about the psychology and expectations of the target audience

Fuck me and no sleep.

Might as well just go with a d% system then.

I'm just going to say it, you have shit taste.

Someone rate my homebrew system's dice system

>xd6, ranging from 1 to 5 with an average of 3 in a stat for humans and monsterous creatures going higher depending on size and strength
>skill bonuses start small but increase as levels are gained
>dice are usually opposing rolls, looking something like 3d6+3 vs 4d6+1 or whatever
>situational advantages and disadvantages are +1d6 or -1d6 so using environmental set pieces to gain an advantage is encouraged

It seems to be popular with my group when we playtested

Bell curves satisfy people’s autism. But they really don’t handle modifiers. The good thing about a flat distribution is that you can more or less judge a +1 to be equivalent in value no matter who you give it to.

This. I really don't see how
> that nice bell-curve is always good if you like consistency
is a valid argument

Someone in another thread mentioned 4d6-4 in lieu of a d20 and that seems interesting since it adds a 0 option and has a much less 'LOL GOT A 20' level to using highest or lowest rolls for extreme outcomes.

I like anything that uses d10 and d100

So instead of zero mathematical operations (rolling a d20 and just seeing what it is), we go with four (die + die + die + die - 4)? That's just stupid. If you want fewer "crits", roll a d6 when you get a 20. On a 6, you crit, otherwise you don't. So instead of rolling three extra dice all the time (and subtracting 4), you just roll one extra die 5% of the time. And if you want a zero option, roll a d6 when you get a 1, and on a 6 (or maybe a 1, if you prefer), it indicates a critical failure, or 0.

>adds a 0 option
>0.000625% chance of actually rolling that option
For all practical purposes, it doesn't really add a 0 option: it will never be rolled.

Yet another system without granularity to characters. No unique selling points. Wouldn't play.

Or just play a ruleset where a 20 is only an autohit, not double damage...

>double damage on crit
>not using the classic max damage style crit
the fuck do you play?

not soon enough

Diceless is bestest.
Come at me.

My melanin-enriched companion.

D6 POOLS

>as much about the psychology and expectations of the target audience
I like GURPS and all, but rooting to get the lowest result feels weird.

>Bell curves satisfy people’s autism
REEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
Shit, that makes sense.
Would you kindly try to sell me OUT of 3d6? Currently my system is:

a. 3d6 roll over TN.
b. Doubles (3-3-4, for example) trigger special effects, even if it is a miss (in the example, even if the 10 rolled is a miss, it procs an effect)

The idea of 3d6 is having a less swingy roll, but %-wise I could do it with d20 or d% for the same effect, since 60% in 3d6 is the same as 60% in 1d20.
Now I don't know what to do.

It’s fine if to use for things that are meant to be reliable (skills not combat) and you’re cognizant of never letting the modifiers or target numbers get too far away from the mean

4d6 assign, faggots. One to succeeding at your intention, two to potential dangers (negotiated upfront), one dropped.

>3d6 roll over TN
This on its own is pointless but...
>Doubles (3-3-4, for example) trigger special effects, even if it is a miss (in the example, even if the 10 rolled is a miss, it procs an effect)
this is quite a nice add.

The problem with 3d6 is that a bonus to the roll effectively alters your chances to succeed depending on the target number.

For example: a +1 bonus on 3d6 to beat TN 11 increases your chances of succeeding by 12.5% while that very same +1 bonus on 3d6 to beat TN 18 increases your chances of succeeding by a mere 1.41%. Note: even a bonus of +3 would have increased the chance to succeed on 3d6 to beat TN 18 with only 8.8%: how's that for intuition?

A bonus on a flat distribution (for instance 1d20) simply means the same for all TN: a +n bonus results in an increased chance of succeeding of n*5%, regardless of the TN: now that's something we can easily work with.

Xd10 master race

I like the odds on the attack, but what's the purpose of choosing the low roll for damage? (Mathematically/system-wise, I mean.)

are those dice from a bag of STACK or something else?

xd10

I can think of two ways to "fix" that:

1. Drop the 3d6, go for 1d20, proc special effect on odds/evens.
The chance of doubles on 3d6 are roughly 40%. Yes, odds/evens on a 1d20 is 50%, but that can be worked out on the intensity of the triggered effect.

This works since people are more familiar with 1d20 and working on 5% modifiers, but waiting for odds/evens doesn't pack the psychological punch of rolling dice wishing for doubles (with the added possibility of an extra mechanical layer of triggering some abilities only on even doubles, or only with doubles of 3).

2. Drastically reduce numerical modifiers, keep 3d6.
Since +/- on 3d6 is a little wonky as you said (and as I already knew), the idea if to reduce numerical modifiers. One way of doing it would adopt D&D 5es Advantage system, in which (adapted to 3d6) players roll 4d6, drop lowest.

Don't know how this would interact with the doubles mechanic (does it increase or decrease the odds?) and it diminishes tactical thinking (players only need to search for one reliable and easy way to get advantage, unless I break the modify the stacking rule over advantage, keeping it at 4d6 max, but allowing for example 2 adv vs 1 disadv to not be canceled as is on 5e).

1d20 roll under is the only true way.
Fight me, faggots

So the higher results for damage are less likely, and loosely correlate with results for to-hit, with 20 being a perfect roll so it deals the most damage, while anything [1, X] is going to be a glancing hit at best, and a middling result of 11 is never above 5 damage.

Is 1d6 workable?

Yeah, I get that. I was thinking more, why worry about keeping the numbers so low in the first place. I'm guessing it was conceived for systems with low HP numbers, otherwise combat would become a slog of just chipping away at each other's hits.

Absolutely. Look up the One Dice system. I've used it and it works quite well. Basic mechanic is 1D6+skill+stat bonus vs target number/opposed roll. It's not an *exciting* system, but it's very solid, and using just a D6 greatly reduces the swinginess you get from using larger ones

Making a lowest common denominator IQ Joe understand a roll-under system is fucking terrible.

Also, other things like if you have a fixed target number, it better be some special number. If your system is 3d6 roll over against TN with the default TN=12, it's gonna raise questions. Inevitably. What's a 12? Why 12? Can I have 10 because that's what my dad told me to use? But I like 13 better because I'm edgy. Like little children, hoomans will prod and poke at anything and throw a tantrum if not satisfied.

But enough with the thinly-veiled normie ridiculing. RPGs follow pretty much the same design rules as engineering projects do, albeit a bit more lax.
And engineering is all about trade-offs and the psychology of the end user. That is, if you aim for the breadth of audience, but we assume that.

Call me whatever, but I really think an arbitrary default number smells inelegance. I clearly see the math behind it, and it doesn't bother me to do whatever math. But it still feels much like an ugly truck mirror bodged onto an otherwise sleek and dem curves sport car.

Dice pools are fine automegalophallophilically, as in, the die-penis scales with your character, but they give me a wanker's cramp-like syndrome after a few hours of crunchy play. It's fine in and by itself, but almost the same math can be modelled in a more elegant way, and I'd like my hand be fresh for the actual wank.

Summing over a dice pool is even worse than counting successes. I sometimes have to do accounting work, and no, thanks, I'd like to not have do more of that when I want to relax.

I can continue with these half-objective complaints half-pet peeves ad nauseam, but the point is that instances of inelegance stack and the effect is rather exponential.

Of course, it can be argued that the overall game concept dictates mechanical choices on a per-game basis, but most of the games are bog-standard in most of their aspects. A bog standard setting, a bog standard premise, what have you. Such games don't need to invent a square-shaped wheel out of wanting to be unique snowflakes or a desperate need for a selling point. No, square-shaped wheels are not going to sell well.

Heck, most games would fly better using d20 roll over, because that's eponymous with TRPGs for like 80% of the target audience. But any ramblings about those 80% have to do with mental inertia and comfort zones, which again is psychology.

I like dice systems for deadly combat, like ASOIAFRP, where armor is negative dodge but adds damage reduction. But probably that could be done with a single die and exploding damage.

ITT: people who can't into statistics

*dice pool systems

yeah. like I roll dice for the sound and tacticle sensations, then input the command into a dicebot.

>The chance of doubles on 3d6 are roughly 40%.
44.44%, actually, so it really is pretty close to 50%

Nah man, with normies you just go "in this game you want to roll the lowest possible number". It's the retarded DnD grogs, that can't get rid of the 'muh high roll' mentality.

How does user feel about a dice pool system where you roll various number and sizes of dice and any number 4 or greater counts as 1 success?
Then you have a different color of dice that function similar to how genesys threat dice work that also cause complications on a roll of 4 or greater?

That sounds incredibly clunky. All these degrees of freedom must have some vital in-play variables attached to them. And a game with so many orthogonal variables is likely to be an insufferable grind.

Not even mentioning that modelling, let alone bug-testing such a monstrosity would be a fucking nightmare.

Rolled 3 + 1 (1d10 + 1)

Doesn't seem that complicated when I think in my head
>I am plaything thiefrat stealfast, so I get a base 1d10 to stealing things
>The guy I'm trying to pickpocket is drunk, so I get another 1d6 to stealing his shit
>But his friends are not so drunk, so there is a black 1d4
>And one of his friends is on the lookout because of someone tried to pick a fight with them earlier, add a black 1d6.

Whoops, look like I failed my base rate but I still have that other 1d6 chance for succeeding

Rolled 6 (1d6)

Forgot to roll.

Okay so I rolled a 6, so I successfully pickpocket the drunk guy.
Now to see if the group collectively notices me

I like ORE.

Rolled 1 (1d4)

I keep forgetting the how the dice function works.

Rolled 4 (1d6)

So the group doesn't all notice me, time for the individual lookout
This would be all one roll at the same time, so you can see how quick this would be to resolve all these different factors.

Yes, it wold be real quick
>So, I got... uhhh... five and... two
>but that five is black
>no, it isn't the red is the black
>which ones were the good ones again?

I mean sure, if everyone in the group is a neckbeard and has millions of dice it could work, but my players only have one set of dice and half of them leaves it at home every fucking session.

Okay so the lookout noticed me.
From there maybe on of the party members decides to knock him out or do something to distract him so the alarm isn't raised. I don't know.
Seems quick and simple to me. Only problem is the lack of granularity, I know autists like their situational +1's and might prefer that to a situational +1d4.

And that's not even getting into fun stuff like how spellcastering can work. You can craft your own spells on the fly and it's success, failures, effects, and repercussions are all represented in one roll.
Want to make a big fireball? Roll 1dX, you choose for the size. A bigger die size means more targets can fit into the area of effect. Then roll, say, xd4 black dice to determine how many allies, if any, are also caught in the blast radius. X is how many original targets could fit in the area effect. The enemies would roll their agility die or whatever to avoid it, then roll a black die to see if they fall down or some shit.
I don't know, but it's flexible.

If your players are having trouble just finding numbers greater than 3, they must be having a lot of trouble doing 1d20+mods or 2d6+mods.

d100 is superior
Curvelets BTFO

Nah man, just looking through five or more dice takes a noticeable amount of time if you roll them properly. Sums are easy. Rolling percentile is even easier.

1d20 roll under does everything you want.
If you care about muh 1% you are retarded.

>granularity is bad
>even when it takes no more effort than my clunky-ass method
t. contrarian

>rolling two dice is less clunky than rolling one
Wew lad

>rolling a sphere is less clunky than rolling two d10s with surface area for days
Wew lad

Roll&Keep is an amazing system, it uses a nonplatonic solid but it being arithimetical makes it so much more satisfying than the shitty "counting sucesses" of Storyller.
Non arithimetic systems are just dumb in that they don't provide you with guaranteed results, rolling 1d20+15 is more reliable than rolling 3d20 since it guarantees you will always hit an enemy with AC 15 or less, while the 3d20 will roll higher on average, it's less reliable in the sense that you will always depend 100% in the result of the die itself.
For me, it's always been about using the dice correctly, though I appreciate the dice variety present in D&D and the very clever usage of d12 coupled with d6+special effects on certain results that The One Ring uses.

d10s have like 5% more surface area, but nice try :^)

>being this deep in denial

>using dicelets

OP here, literally first result of Dice on google (from wikipedia)

except common denominator Joe understands that Golf works on the principle of "lowest score = best score"

D100 is only better than d20 if you make use of the roll in creative ways, like switching the numbers for hit location (so a roll of 47 uses the hit location of 74).

I once wanted to do this, quite fun to design.

Even if the system doesn't take advantage of its strengths - of which there are many - d100 is just inherently better because acquiring your % chance of success doesn't require multiplication or (in the case of curvelets) the consultation of a fucking normal curve

3d6 handles modifiers fine, but apparently brainlets like you don't.