No whining or complaining about dice rolls in any way

>The difference is when the stats say you're concretely able to do something you don't have a huge chunk of random failure lurking there and when you're not able to do something then you get fucked like you would in real life

The chunk of random failure on a d20 doesn't have to be huge. If you have 75% chance of something, you have 75% chance of something, regardless if you are rolling a d20 or 3d6.

> Little bonuses and maluses make a bigger difference when things are closer to an even match and it minimizes the randumb when they're obviously slanted.

So use larger bonuses/maluses for a d20. You kinda have to, the dice range is pretty big (d10 is more ideal imo).

I'm not saying that, say, 5e dice math is well designed (SW SAGA has math a lot better fitting for the d20), but I'm so fucking tired of the 3d6 meme regurgitated by people who never sat down to think about this shit.

>If you have 75% chance of something, you have 75% chance of something, regardless if you are rolling a d20 or 3d6.
It's mental effect, for the GM who thinks being good at something means you fail 1 out of every 4 times. Very infrequently do you actually have this kind of scenario; most of the time either you succeed consistently or you struggle to succeed at all.
>So use larger bonuses/maluses for a d20.
This doesn't provide the same effect because when your distribution is curved, the effect of the same bonus or malus is different at different parts of the curve, in a very natural way handled by the math. Trying to achieve this with d20 simply feels hamfisted and arbitrary, unless you draw up tables to help you decide, in which case you defeat your own purpose.

> the effect of the same bonus or malus is different at different parts of the curve, in a very natural way handled by the math

Why is "I can't tell at a glance how much a bonus is worth" a good thing somehow suddenly?

>Trying to achieve this with d20 simply feels hamfisted and arbitrary, unless you draw up tables to help you decide, in which case you defeat your own purpose.

You don't have to simulate a curve with a d20. You don't _want_ to simulate a curve with a d20. You just want to use bonus sizes that are meaningful instead of many +1-2.

Also, this is leading away from the whole "2d10 is more realistic" shit to begin with.

>Why is "I can't tell at a glance how much a bonus is worth" a good thing somehow suddenly?
Knowing that a bonus is worth exactly n% at all times is both an unnatural way of playing and also doesn't help as much as you'd think.
>You don't have to simulate a curve with a d20. You don't _want_ to simulate a curve with a d20.
Because that's not what it's suited for; correct. But I hold the curve as more desirable.
>You just want to use bonus sizes that are meaningful instead of many +1-2.
And you have to decide from scratch every time or draw up a table for what makes "meaningful". I think it's better for the human to think simply and the math to make things meaningful.

>Knowing that a bonus is worth exactly n% at all times is both an unnatural way of playing and also doesn't help as much as you'd think.

How is having a good ability to estimate how much a thing helps unnatural?

>And you have to decide from scratch every time or draw up a table for what makes "meaningful".

... No? You have to decide once, when you design the game. Then just keep applying it consistently.

Yeah, but chances of Critical Hits in Pathfinder, for example, would be strongly affected, for example.

And target numbers for Skill Rolls would need some tweaking I imagine, since we'd move from a Median to an Average.

>How is having a good ability to estimate how much a thing helps unnatural?
It takes training to understand probabilities realistically. See the 75% example. 1 out of 4 chance to bungle something is shitty odds but people only see "75% is a big number" whereas "95% is too big". So it's better to think in terms of "it's close but this will be a big help, this will fuck me over" than "oh huh, this bonus makes me 10% more likely to succeed" when you don't know what 10% really means other than vaguely that 25% gut-feels more palatable to you than 35%.
Critical hits were a mistake. Properly implement gradated success or tuck it back into your trousers.

Just gonna leave this here.

Distribution shaped like a tit is no way to play, user.