Percentiles, what's it good for?

[Aryxbez]

And shrieking, hurling insults, and behaviour unbecoming of a 9 year old isn't being an asshole? Dude, calm down before you give yourself a hernia, this is just a discussion about how to roll some dice, not a holy war.

Most people don't seem to realize this, but in text medium like this, just because one is using curse words to emphasize, doesn't actually mean they're "yelling", insulting someone, or otherwise in an aggressive state. In the case of someone calling you names, it's moreso calling out to your ideas/what you said, not yourself personally. Making multiple "calm down" assumptions to how someone is acting, is a misunderstanding in perspective, that oftentimes is found insulting for being presumptuous. Also add that, ye had misused terms confusing the conversation, these insults can also very well be out of frustration for you making a mess of things by not researching the correct term beforehand (or, don't use the term at all, if unsure what it means).

That said, I'm pretty sure this is the notion of the den, as they have laxer rules on what you can say here, people are free to voice their thoughts. Even if it is quite aggressive, as least it is being honest, and not hiding behind passive-aggression, apologist and other BS on wordplay that harms discussion (as you'll see in most other RPG forums). So, I wouldn't take it personally, and continuing to do so, is a misunderstanding of this forum's culture, as well possibly a conflict in mindset that may harm future discussion.

[Username17]

d% roll under, and roll under systems in general, are defended by people like SpaceLem and Shadzar. People who literally refuse to acknowledge that a constant proportion is not and never will be a margin of success. People who refuse to acknowledge that iterated division is more cumbersome than addition.

If someone says that they like d% roll-under, that is a warning sign. Pretty much exactly the same warning sign as if someone says that their favorite edition of D&D is 2nd edition. It means that their brains were calcified in the 80s and they have refuses to learn or grow since then.

-Username17

[Mask_De_H]

Shit, Frank: in the most basic, roll-and-check sense roll-under is marginally easier to grok. It's just roll dice, look at TN, look at dice.

Roll-over TN 100 does require an "extra step" in that you either need to a) add your CoS to your roll and see if it's 100 or more or b) subtract your CoS from 100 and see if your roll beat that number. Granted, if you work in degrees of five or ten, checking success is pre-school math but people are dumb and "is this number bigger than that number"? is easy.

Once you do anything more complex than that it falls apart and roll-over TN 100 is better, but roll-under does win in that single field.

[RadiantPhoenix]

Shit, Frank: in the most basic, roll-and-check sense roll-under is marginally easier to grok. It's just roll dice, look at TN, look at dice.

"The die must show under this number" is about as hard as, "the die must show over this number"

Once you do anything more than that, though, things start getting harder.

[codeGlaze]

Because THAT IS WHAT THE WORD FUCKING MEANS, ASSHOLE!

And shrieking, hurling insults, and behaviour unbecoming of a 9 year old isn't being an asshole? Dude, calm down before you give yourself a hernia, this is just a discussion about how to roll some dice, not a holy war.

I see that you're new here. Welcome to the Den.

Game On,
fbmf

Hahaha. Priceless. XD

[vagrant]

I've been to other RPG forums. Every time I read a thread and see the uninformed, bullshit self-congratulatory idiocy that's evidenced in every single post I think 'Fuck this, I'm going back to the Den.'

Yeah, this place is full of complete and unrepentant dicks, but they're fucking honest about and some of then even fucking admit mistakes and apologise from time to time!

[TarkisFlux]

Yeah, this place is full of complete and unrepentant dicks, but they're fucking honest about and some of then even fucking admit mistakes and apologise from time to time!

[fbmf]

And shrieking, hurling insults, and behaviour unbecoming of a 9 year old isn't being an asshole? Dude, calm down before you give yourself a hernia, this is just a discussion about how to roll some dice, not a holy war.

I see that you're new here. Welcome to the Den.

Game On,
fbmf

Hahaha. Priceless. XD

I was actually being sincere. I get that we're not like other forums.

Game On,
fbmf

[codeGlaze]

I was actually being sincere. I get that we're not like other forums.

Game On,
fbmf

That's why I found it amusing.

[Mask_De_H]

Shit, Frank: in the most basic, roll-and-check sense roll-under is marginally easier to grok. It's just roll dice, look at TN, look at dice.

"The die must show under this number" is about as hard as, "the die must show over this number"

Once you do anything more than that, though, things start getting harder.

This is true. However, "the die must show under the chance of success" and "the die must show over (the target number minus the chance of success)" or "the die plus the chance of success must be equal or greater to 100" require a different amount of work.

That's a reason you have people think roll-over is weird or too hard. It requires an extra step on the player's side for its most basic check than the most basic roll-under check. This is a function gamers make in every other goddamned game but d% ossified into roll-under.

[Username17]

I admit that a comparison step is faster than an addition step. If you literally don't have any modifiers at all, and thus never have to do an addition step, you are legitimately faster with roll under. This is, however, a ridiculous niche case. No RPG worthy of the name is going to run an RNG that is unmodified by circumstances or opposition.

Which gets us to the brain calcification thing. SpaceLem's argument was literally that you could avoid the hateful addition (until you had a modifier that affected success chances in any way), and "all" you had to do was remember twenty different special numbers in three different categories that came in two different frequencies. But even that was easy, because ten of the special numbers were also all the numbers with a single last digit, so it was really like you only had to memorize eleven super special numbers in three categories that came in two distinct frequencies. And anyway, who really cares about whether your game can give different outputs for success rates across tasks of varying difficulty anyway?

That's seriously his argument for why we don't need to add numbers together. But of course, addition is the second fastest operation there is, so by the time you've done all that crap to maintain the purity of using a roll under for the speed of doing everything in a single comparison step... you don't even have that anymore.

-Username17

[jadagul]

Frank: I think it might hold up if you have static large modifiers and frequently-changing small (single-digit) modifiers. Because I suspect the way most people would calculate this is:

Roll-under: you have a target difficulty written down on your sheet, that's pre-figured, with all the large modifiers factored in. When you want to make a roll, you add or subtract the small temporary modifiers to this difficulty (a bunch of calculations with single-digit numbers), and then you roll your dice and do a comparison operation.

Roll-over: You have a static bonus written down on your sheet. When you make a roll, you roll your dice (to get a double-digit number), add your static bonus (another double-digit number), add or subtract all your temporary bonuses (single-digit numbers), and then compare to 100 (pretty trivial).

What roll-under gets you (in such a system) is never having to do double-digit addition at the table.

As an aside: while "largest successful roll wins" isn't technically a degree-of-success calculation, I'm pretty sure it gives you the same probability distribution as actually taking degrees of success.

[rasmuswagner]

A lot of modern games use very few rolls, use ante-up mechanics, negotiate the stakes and so on.

Setting those aside, and focusing on games with frequent and simple rolls for success (and quality of sucess), here's my maxim: If I can't roll some fucking dice and immediately tell a good roll from a bad roll, before I know the exact difficulty and situational modifiers, then fuck you.

That's one of the reasons "roll-under, highest passed roll wins" gan go eat a dick.

[tussock]

I admit that a comparison step is faster than an addition step. If you literally don't have any modifiers at all, and thus never have to do an addition step, you are legitimately faster with roll under. This is, however, a ridiculous niche case. No RPG worthy of the name is going to run an RNG that is unmodified by circumstances or opposition.

See, the ultimate idea is to get as much of the math that needs done finished up front. That's what math education ends up doing, teaching you to do as little math as possible. We use multiplication because it's easier than repeated edition.

And ...

(1d20 <= Skill+mods) >= DC+mods.

is the ultimate simplification of every possible permutation of known (player-side) and unknown (GM-side) modifiers, assuming you do anything sensible when your mods grow larger than the RNG. If the non-standard mods on the player side are ever zero you win and even if they're not you're adding smaller numbers and still only ever doing less or equal operations on them, and comparing smaller numbers too. Win all around.

1d20 + Skill+mods >= DC+mods.

or ye olde

1d20 + AC-mods >= THAC0-mods

(which is used as THAC0-mods - 1d20 <= AC-mods) are simply wasting everyone's time. Factually. The very same reasons that THAC0 was rightly replaced with BAB apply and should see BAB replaced with the superior option.

That someone can make a bad argument for it does not make it less true. What one calls out for the GM to do their comparison with is the number on the die. One often doesn't need to work out mods at all, in a way that's more intuitive than the default with d20.

Obviously you could add stupid things to that like old designers did in the 80's or whatever, but it doesn't need them at all. I have no idea why anyone would want to add subtraction to it, other than to emulate AD&D.

If I can't roll some fucking dice and immediately tell a good roll from a bad roll, before I know the exact difficulty and situational modifiers, then fuck you.

That happens in every system. There's always a set of numbers that usually win no matter how convoluted your outcomes are, and always a set of numbers that are going to be close and make you do a touch of math (even if it is just a comparison or two).

[hyzmarca]

Frank: I think it might hold up if you have static large modifiers and frequently-changing small (single-digit) modifiers. Because I suspect the way most people would calculate this is:

Roll-under: you have a target difficulty written down on your sheet, that's pre-figured, with all the large modifiers factored in. When you want to make a roll, you add or subtract the small temporary modifiers to this difficulty (a bunch of calculations with single-digit numbers), and then you roll your dice and do a comparison operation.

Roll-over: You have a static bonus written down on your sheet. When you make a roll, you roll your dice (to get a double-digit number), add your static bonus (another double-digit number), add or subtract all your temporary bonuses (single-digit numbers), and then compare to 100 (pretty trivial).

Well no. Because with roll over you can subtract your bonus from your TN of 100 and write the modified TN on your character sheet so that it's pre-figured.

You'd only worry about adding or subtracting large numbers during rolls if you have a variable TN, which isn't a terrible idea but adds a fuckton of complexity.

In other words, you can have a roll over system that's effectively identical to roll under your skill by having your skills start at 95 (or 100 for 0 chance of success) and become smaller as you level up. This is a bit counter-intuitive in a 2E AC way, since it means that low skills are good and high skills are bad, but it's just as simple mathematically and it had more space for additional modifiers.

You can get rid of the counter-intuitiveness by adding an extra additional subtraction step that will only be used when skills are leveled up.

[spaceLem]

I admit that a comparison step is faster than an addition step. If you literally don't have any modifiers at all, and thus never have to do an addition step, you are legitimately faster with roll under. This is, however, a ridiculous niche case. No RPG worthy of the name is going to run an RNG that is unmodified by circumstances or opposition.

I hardly ever see modifiers in play. If there is, the GM says there's a -20 penalty, then I'm rolling under 40 instead of 60.

Which gets us to the brain calcification thing. SpaceLem's argument was literally that you could avoid the hateful addition (until you had a modifier that affected success chances in any way), and "all" you had to do was remember twenty different special numbers in three different categories that came in two different frequencies. But even that was easy, because ten of the special numbers were also all the numbers with a single last digit, so it was really like you only had to memorize eleven super special numbers in three categories that came in two distinct frequencies. And anyway, who really cares about whether your game can give different outputs for success rates across tasks of varying difficulty anyway?

Checking if the units die rolled a 0 is trivial. No need to recalculate anything if your numbers change, or write down different values, the division is pretty much free.

That's seriously his argument for why we don't need to add numbers together. But of course, addition is the second fastest operation there is, so by the time you've done all that crap to maintain the purity of using a roll under for the speed of doing everything in a single comparison step... you don't even have that anymore.

Did you remember that the bigger the numbers, the slower it gets to add them? It's also easier to work with large round numbers.

Roll under requires no mental effort, or any sort of calculations (except adding bonuses or penalties if the GM offers them before you roll).

[fbmf]

I admit that a comparison step is faster than an addition step. If you literally don't have any modifiers at all, and thus never have to do an addition step, you are legitimately faster with roll under. This is, however, a ridiculous niche case. No RPG worthy of the name is going to run an RNG that is unmodified by circumstances or opposition.

I hardly ever see modifiers in play. If there is, the GM says there's a -20 penalty, then I'm rolling under 40 instead of 60.

Seriously?!

Game On,
fbmf

[spaceLem]

Seriously?!

Yes, I'm a little surprised by that actually. Currently playing BSG (add polyhedrals, lookup on success table), Advanced Fighting Fantasy (2d6 roll under or roll over, modifiers for stuff like disguise or stealth spelt out before we roll, ... actually they do come up fairly often now that I think about it), Amber (god knows what the GM's up to), and just finished a homebrew (d20 roll under, margin of success involved subtraction, but there were never any modifiers before you rolled, so a pass never became a failure, or vice versa -- also sheer MTP, but probably one of the fastest paced and best games I've ever been in). About to join a Basic D&D game, which will use 2d10 roll under for stat checks, don't know if modifiers will be applied before the roll. The last time I played D&D 3.5, it seemed like there were modifiers all the time, everywhere.

I'll give Frank that I'd never accept multiplying your skill prior to a roll for a hard/easy check (higher skill characters should not get a bigger penalty for a given difficulty -- I've seen that system described in some games like... OpenQuest I think?), and I suppose it's consistent to consider crits and fumbles in the same way, and the best way to get that is to use some arithmetic. It's just that after success and failure, crits and fumbles are just interesting side effects, and having them as a flat proportion of successful rolls is good enough.

As for the games with d20 and 2d6 roll under, they don't have the handy way of generating crits and fumbles that you have on d%, so we tend to rely on all dice rolling a 1, or margin of success. I'm reasonably okay with subtraction on a d20, wouldn't want to do it with a d% though.

[hogarth]

I hardly ever see modifiers in play. If there is, the GM says there's a -20 penalty, then I'm rolling under 40 instead of 60.

Seriously?!

Before 3E came out, I didn't see many modifiers used in play (maybe range modifiers, but that's about it). That was a bad thing, and 3E's multitude of possible target numbers was a giant leap forward, as I've noted in the past.

[Desdan_Mervolam]

I rarely if ever saw modifiers enter play prior to third edition as well. Especially in percentile roll-under systems.

[spaceLem]

Before 3E came out, I didn't see many modifiers used in play (maybe range modifiers, but that's about it). That was a bad thing, and 3E's multitude of possible target numbers was a giant leap forward, as I've noted in the past.

I don't want to be an armchair theorist -- a lot of the games I play in just seem to not care that much about having many modifiers, and so it seems sensible to create a tool that reflects that sort of usage. There is a trade-off of correctness vs speed, and if you value the former, then obviously roll under is going to suck for you.

The major advantage of using a d% over say, a d20, is that we're trained to think about chances of success in terms of percentages, and the d% offers a very natural way to present that. Otherwise, apart from a few dice tricks, it's a pretty rubbish die to use, and offers very little over a d20 (how often do you really need that extra precision?). If you're concerned with modelling reality better, there are definitely much better dice mechanics out there.

If you have a lot of tactical play (which I tend not to bother with because I suck at it), then you need the modifiers, because otherwise every option ends up the same, and you might as well just not bother. Personally I like being upfront about modifiers, as I like to know what my chances are before I roll the die, rather than roll and stare dumbly at the GM, while they consider, hoping they'll be satisfied enough with the result to grant me my success.

[Ancient History]

The major advantage of d% is that allowed oD&D to fill books with 100-entry tables.

[fbmf]

I hardly ever see modifiers in play. If there is, the GM says there's a -20 penalty, then I'm rolling under 40 instead of 60.

Seriously?!

Before 3E came out, I didn't see many modifiers used in play (maybe range modifiers, but that's about it). That was a bad thing, and 3E's multitude of possible target numbers was a giant leap forward, as I've noted in the past.

That is my experience as well, but that was thirteen years ago.

Game on,
fbmf

[jadagul]

Frank: I think it might hold up if you have static large modifiers and frequently-changing small (single-digit) modifiers. Because I suspect the way most people would calculate this is:

Roll-under: you have a target difficulty written down on your sheet, that's pre-figured, with all the large modifiers factored in. When you want to make a roll, you add or subtract the small temporary modifiers to this difficulty (a bunch of calculations with single-digit numbers), and then you roll your dice and do a comparison operation.

Roll-over: You have a static bonus written down on your sheet. When you make a roll, you roll your dice (to get a double-digit number), add your static bonus (another double-digit number), add or subtract all your temporary bonuses (single-digit numbers), and then compare to 100 (pretty trivial).

Well no. Because with roll over you can subtract your bonus from your TN of 100 and write the modified TN on your character sheet so that it's pre-figured.

You'd only worry about adding or subtracting large numbers during rolls if you have a variable TN, which isn't a terrible idea but adds a fuckton of complexity.

In other words, you can have a roll over system that's effectively identical to roll under your skill by having your skills start at 95 (or 100 for 0 chance of success) and become smaller as you level up. This is a bit counter-intuitive in a 2E AC way, since it means that low skills are good and high skills are bad, but it's just as simple mathematically and it had more space for additional modifiers.

You can get rid of the counter-intuitiveness by adding an extra additional subtraction step that will only be used when skills are leveled up.

With roll-over you could subtract your bonus from 100 and write that down on your sheet. This is true. I'm skeptical, however, that this is what people in real life would actually do.

[Username17]

My experience with 2nd edition under several different DMs was that some DMs never handed out check modifiers, but other DMs really liked handing out modifiers to ability checks. Those modifiers were almost always exceedingly large (like -4 or even -8), and the results were that people failed stuff all the time and also the players responded by inflating their attributes in a vicious cycle that prompted the DMs to impose even harsher modifiers to attribute checks.

Certainly, the actual books are written to accommodate the latter playstyle. For example, the official rules for taming a Pegasus are that you have to make a Ride Airborne check at -10 while it's asleep. Ride Airborne is itself a Wisdom check at -2, so the rules are in fact asking you to roll twelve points lower than your Wisdom on a d20 to get your flying horse upgrade. No retries! And that's even assuming the DM let you spend two proficiency slots on getting Ride Airborne when you didn't have a flying mount yet. If not, I hope you like an additional -4 penalty for non-proficiency. It's no wonder that the official characters were covered with high stats to the point that a 15 or 16 might be their "low attribute". The attribute rolls being asked for basic tasks were nothing short of ridiculous and you might seriously need a 17 just to have a 5% chance of success right out of the book.

-Username17