What is your ideal RNG range?

[Previn]

Pretty much what the subject says. I'm curious how much of an RNG range people actually feel comfortable using.

What's the point/% chance where you as a GM just say 'don't roll, you're going to succeed anyways' and what's the point where you as a player looks at the odds and say 'That's just not going to happen, we need to find another way or do something different.'

For me anything above 90% just doesn't seem with bothering to check for, and anything less than 20% is pretty much no go territory unless there is no other option.

[Whiysper]

I use a 3d6 RNG, so personally anything that needs a 5 or more is a given, and a 15 or more is the sort of 'only option left' long-shot.

DnD, I actually don't worry about it. Flat-curve D20 is so damn swingy that I don't consider anything a foregone conclusion any more .

[Iduno]

I roll it out, unless it's an opposed roll and someone does entertainingly well or poorly. The randomness keeps it interesting, and 5% chances still happen 1 time in 20 on average.

[OgreBattle]

1d6 with 2-3 steps

Like roll to hit
Roll to deal damage
defender rolls a saving throw

[Previn]

1d6 with 2-3 steps

Like roll to hit
Roll to deal damage
defender rolls a saving throw

So 40k-ish? Is this guard needs to roll a 6 to hit, a 6 to wound with his lazgun and the demon needs a 2+ save?

d6 is a range, and using a d6 three times provides a range of results, but it doesn't really set the expectation of where you stop caring about successes, or stop trying because the change of success is too low (.46% in my above example). I don't think anyone would find a .46% chance of doing something acceptable unless the other option/result was nothing.

People go with 3d6 for a lot of things, because it has that nice bellcurve and easy to calculate results and it has that nice taper at the end. But that same taper at the ends means that the range of reasonable results for 3d6 is really about 7 or 8 results in the middle. Either end is really a waste, and any +/- modifiers are really about moving those 7-8 meaningful middle results. I could drop down to a 2d6 and end up with basically the same usable RNG range.

By that same took, lots of Shadowrun dice pools really only care about 5 results, with everything else being effectively auto-pass or effectively impossible and hitting the impossible side that 1 in 1000 chances can just lead to dumb results.

I guess I'm trying to gauge if it's worth while to even have those tails as part of the RNG, and where the cut-off for people lies.

[DrPraetor]

I use a 3d6 RNG, so personally anything that needs a 5 or more is a given, and a 15 or more is the sort of 'only option left' long-shot.

DnD, I actually don't worry about it. Flat-curve D20 is so damn swingy that I don't consider anything a foregone conclusion any more .

I'm a big fan of 3D6. 3D6 and 4D6 are pretty similar - on 3D6, a +1 bonus can be as big as a 12.5% difference, vs. ~11% on 4D6. The black swan events on 4D6 come up about .1% of the time each, so you might play an entire campaign and never see either a 4 or a 24; on 3D6 you get a 3 or an 18 .5% each, so you probably will see them. People may be mistaken in thinking they want black swan die rolls depending on what outputs the game system actually delivers, but the thing about 3D6 or 4D6 is you'd have to play a long time before you noticed any difference at all in the swinginess EVEN IF THE RELATIVE BONUSES AREN'T EVEN SCALED - therefore 3D6 is better because it's easier and faster to roll. If, for some reason, you wanted appreciably smaller bonuses to hand out...

My personal heartbreaker is essentially 3rd edition with 3D6 instead of a D20 as the core RNG. You don't even change the average output noticeably (which means that much of the math with running and jumping and swimming and shit still holds), but the reduce in swinginess for 3D6 vs a D20 is, as you say, quite noticeable.

[merxa]

My heart breaker also 3d6 with 4 main stats.

I'm still trying to figure out if hit points or hit boxes are the way to go.

To more directly answer the question I typically handwave results if my ad hoc judgement of probability hits the 1/1000 to 1/10000 or greater range.

[Blade]

For me the ideal RNG range will depend on the game.
If the game has you play professionals, I'd rather have more auto-successes.
If the game has you play a ragtag team of amateurs, I prefer to keep the door open to unexpected failures and "1 in a million" (probably more of a 1 in 20 in reality) success.

Generally speaking, I (and cognitive psychology) expect players to prefer auto-success when they have about 80% chance and to still want to roll the dice when have a chance, no matter how small it is.

[Orca]

The number range isn't the issue usually. I do want something like 4 degrees of success (Pathfinder 2's crit success/success/fail/crit fail isn't the problem with their system) without an absolutely fixed chance of the more extreme results, like occurring on a 20 on d20 every time.

[Username17]

There's a good deal of research that people don't treat chances and risks equally when they are presented differently. A chance to win is viewed more favorably than a chance to lose - even if the involved chances are the same. But also people have very little intuitive grasp of probabilities other than 0%, 100%, and 50% and describing similar chances with different words gives large differences in how they are perceived.

Let's consider the difference between 3d6 and d20. Imagine someone said in a 3d6 system "You need to roll a 17+." Now imagine instead that for a d20 system they said "You need to roll a natural 20 and then roll again and confirm with a 13+." Most people are going to find the latter more daunting. But of course the d20 event is very slightly more common (2% versus 1.85%).

-Username17

[Ferret]

percentile roll-over target 100 and degrees of success. I like my granularity.

[OgreBattle]

Yeah 40kish, the 6 6 6 roll is for a 10-20 goon squad blasting at a space daemon before the protagonists show up

[Thaluikhain]

Been thinking about d6 and 40k (or rather, old Necromunda) style shooting.

If you roll d6 and add your BS or whatever and hit on a 7+ or whatever, you've only got a very limited range of numbers. Also, if you're under some effect that halves your BS, BS3 and BS4 (assuming you round up) are the same.

Was thinking perhaps about increasing your BS number, the two hit number and rolling 2d6. Would make rolling multiple shots fiddly, though, you'd need to keep dice in pairs. And my idea to have the number of hits be the number the shooting roll exceeding the target number would mean a lot more hits that way.

[OgreBattle]

2d6 is the warmachine play with a pair method yeah?
With warhams style I figure increasing average attack/shot number can make up for the small rng

I figure a to hit roll should be compared vs a target number representing range visibility and so on, rather than a flat ‘hit on X+’

I also figure every step has a result. So rolling a wound means they’re going to be hurt, their saving throw then determines if they’re KO’d or just wobbled

Rolling a 6 for a critical that gives an additional shot/attack/wound roll is also fun.