A search for an optimal resolution mechanic

[Murtak]

So... that thing with the piles of d6s +X and you pick the highest, +1 for each 6 left over after picking the highest...

Would it work at all well with d20s? (or d12, or d10, etc) It seems like adding more sides to the type of dice rolled would keep numbers in the "average" range a lot, with occasional chances of rather low rolls and a very very small chance of getting the extra high rolls.

Using bigger dice lessens the impact of the +1 bonus and additionally makes that bonus less likely to happen. So using d20 I can guarantee that entire campaigns will happen without anyone seeing one of those +1 bonuses. At that point, why have it at all?

As for "average ranges", I am not even sure what you mean. What is an average range for a number?

[spaceLem]

My own system (I haven't seen anything similar mentioned yet).

Roll 2d6. For every advantage you have, add a bonus die. For every disadvantage, add a penalty die. Bonus dice and penalty dice neutralise each other, so you'll never have both, and never roll fewer than 2 dice.

If you have penalty dice, add the worst 2 dice, if you have bonus dice AFF the best 2 dice. Check the result.

2: critical fumble
3-5: fail
6-8: no thrills success
9-11: good success
12: critical

The results might need to be tweaked, but you get the idea. Apart from the quick resolution (adding 2d6 is easy, finding the highest or lowest isn't too difficult), the major advantage of this system is the variance. The highest variance is when opponents are equally matched, or a task is moderately challenging, while the variance decreases the more that goes for or against you. Also, the chances of a critical depend on your relative abilities, reducing the glass ninja problem.

A slight downside is that you might be throwing a lot of dice if the power levels are significantly different, but above about 5 dice the outcome really becomes "success or critical?" (or equivalent for penalty rice), but no results are ever ruled out. You could easily add a dice cap to suit your hand size.

Other than that, the best systems are 2d6 roll under, 1d20 roll under, or d% roll under (never over with a d%). If you want degrees of success with the last system, then if you pass and the units are a 0, you have a good success, and if the units are a 5, you have a critical success.

[Murtak]

So how does that system handle opposed rolls?

[spaceLem]

So how does that system handle opposed rolls?

Everyone rolls, neutralise any common bonuses (if there are multiple people, maybe adjust so the median bonus is around 0), highest score wins. If you've got something like a jump, then the distance will set the penalty dice.

If it's a PC vs an NPC, and they're arm wrestling, then just give the PC the roll plus the bonuses, and see if they succeed or not. Success on a 6+ means you're generally likely to win if you're closely matched, but you could easily adjust the results table somewhat to make things less easy (e.g. success on a 7+, or maybe keep 3d6 to ensure there's a straight 50% point).

The system is probably better for giving a qualitative result for handwavy situations, than for simulating things too closely, but I think there's some mileage in the system. Plus, I didn't see anyone mention any roll and keep systems in the thread.

[tenuki]

I've been tinkering with a 2d10 roll under system.

Basic success at roll <= TN
Good success at roll <= TN/2
Critical success at roll <= TN/3

This has the advantage of scaling meaningfully beyond the basic RNG range.

A skill rating of 15 signifies a reasonable level of competence. It gives you 85% for a basic success, 28% for a good success and 10% for a crit.

In case of opposed rolls, success levels are subtracted from each other, i.e. critical vs. good = basic, critical vs. basic = good, good vs. basic = basic.

[PhoneLobster]

Roll under systems? Really now?

And variable target numbers determined by DIVISION and one of those is a half and the other is a THIRD?

I'm sorry. Those are all bad things. Go use positive numbers that go in the easy direction and stop dividing things at random holy crap.

[spaceLem]

Roll under systems? Really now?

And variable target numbers determined by DIVISION and one of those is a half and the other is a THIRD?

I'm sorry. Those are all bad things. Go use positive numbers that go in the easy direction and stop dividing things at random holy crap.

Roll under is definitely a good thing. Division by a third, not so much.

My Sunday afternoon game uses a d20 roll under system (the home brew of a guy named Phil), and uses how much you passed by as a degree of success (although only 1s and 20s are crits or fumbles). He's used the same system, (and mechanic) to run games exclusively for the last 15 years, and considers it tried and tested There is a large queue to be in Phil's games.

[Username17]

He's used the same system, (and mechanic) to run games exclusively for the last 15 years, and considers it tried and tested There is a large queue to be in Phil's games.

This is a fucking retarded argument, and you should feel ashamed of yourself.

People can be personally entertaining and I will want to be in games they run. This is largely independent of whether or not the overall system they are using is shitty or not. I'll play fucking d20 Star Wars or New World of Darkness with people that I like. And we're not even talking about whether the overall system is shitty, but merely whether a single mechanic is more cumbersome than it needs to be.

Roll Under handles opposed rolls and situational modifiers worse than roll over does. Period. That's the only difference. It's exactly the same in terms of what modifiers it can accept before the RNG breaks and it is exactly the same in terms of your chances to succeed at any particular task. There is no reason to use Roll Under instead of Roll Over. The only difference is that opposed rolls require more math to calculate and adding modifiers is clunkier and more time sensitive.

Advocating Roll Under is basically insane. It has no advantages. And while the drawbacks are not game breaking, they nonetheless exist.

-Username17

[spaceLem]

This is a fucking retarded argument, and you should feel ashamed of yourself.

No it's not, and I don't, thanks.

"It works" is how D&D started in the first place. IME, people don't tend to point out problems with dice unless they're really bad (so when everyone was bitching about THAC0 in our 2e game, I feel justified in saying that was a bad mechanic).

You can go on about how good something is on paper, but ultimately what matters is how it works in practice. Empirical data and all that. Phil has had a huge number of players over the years, and that is what works. I don't know how large a sample size would satisfy you, probably more than one GM could ever achieve, but he derived his system originally from Call of Cthulhu, and that's been been pretty successful too.

Roll Under handles opposed rolls and situational modifiers worse than roll over does. Period. That's the only difference. It's exactly the same in terms of what modifiers it can accept before the RNG breaks and it is exactly the same in terms of your chances to succeed at any particular task. There is no reason to use Roll Under instead of Roll Over. The only difference is that opposed rolls require more math to calculate and adding modifiers is clunkier and more time sensitive.

Advocating Roll Under is basically insane. It has no advantages. And while the drawbacks are not game breaking, they nonetheless exist.

Sorry, incorrect. Roll under handles opposed rolls slightly less well than roll over, but that is more than made up for by how it's so much easier to perform unopposed actions, as it a trivial comparison that cannot be done with roll over. If you want degree of success, you'll have to do subtraction either way, it might as well be with smaller numbers.

d% roll over is a crock of shit.

[virgil]

You can go on about how good something is on paper, but ultimately what matters is how it works in practice. Empirical data and all that. Phil has had a huge number of players over the years, and that is what works. I don't know how large a sample size would satisfy you, probably more than one GM could ever achieve, but he derived his system originally from Call of Cthulhu, and that's been been pretty successful too.

How is that proof of the system being good as opposed to the GM being good? Empirical data would require numerous different GMs using it successfully.

[spaceLem]

How is that proof of the system being good as opposed to the GM being good? Empirical data would require numerous different GMs using it successfully.

I'm suggesting that it's Phil checking what works with his many players, rather than the players experiencing the same mechanic under different GMs. I know Phil isn't dogmatic about his system, and I'm pretty sure a problematic mechanic would quickly have revealed itself, as we gamers are prone to whinging about stuff we don't like

[Username17]

You can go on about how good something is on paper, but ultimately what matters is how it works in practice. Empirical data and all that. Phil has had a huge number of players over the years, and that is what works. I don't know how large a sample size would satisfy you, probably more than one GM could ever achieve, but he derived his system originally from Call of Cthulhu, and that's been been pretty successful too.

Wow. You're saying it works better than Call of Cthulhu?


CoC is one of the worst systems ever put together. Fortunately the game is based largely around mood, magical teaparty, and the expectations that your characters will die horribly sooner rather than later. So the system doesn't get in the way that much. But that doesn't mean that the game mechanics aren't a war crime. They totally are. Making something that is game mechanically superior to CoC is not even enough information to tell us whether you've met the bare minimum requirements to not get laughed at.

Sorry, incorrect. Roll under handles opposed rolls slightly less well than roll over, but that is more than made up for by how it's so much easier to perform unopposed actions, as it a trivial comparison that cannot be done with roll over. If you want degree of success, you'll have to do subtraction either way, it might as well be with smaller numbers.

d% roll over is a crock of shit.

Wow. You're horrendously wrong here. Degree of Success is incredibly easier with d% roll over. Your Target Number is 100, you roll your dice and add your modifiers. Your degree of success is just the last two digits of the final number.

This isn't even up for debate. If you want to roll percentile dice and you want to do things with degrees of success or opposed rolls, you should use roll over. Mathematical steps can be counted. Roll Over requires less of them. Also, it provides exactly the same actual chances as Roll Under because it is literally the same system which has been simplified to have all the necessary numbers on the right side of the equals sign.

The only way that Roll Under ever has an advantage with anything is if you are generating binary success/failure results and you are not applying circumstantial modifiers of any kind. In that very limited circumstance, Roll Under is superior because it resolves with a simple comparison operation instead of requiring addition or subtraction of any kind. If you are doing anything else (opposing tests, modified tests, degree of success, secret difficulties, etc.), then Roll Over is somewhere between "slightly easier" and "massively easier".

If you are advocating Roll Under for an RPG today, you are a relic. This is 2012, and science has progressed. You might as well be advocating THAC0. It's a pointless overcomplication of the rules system. Like THAC0, it simply reverses the signs on some things with no realized or even potential benefit.

-Username17

[spaceLem]

Wow. You're saying it works better than Call of Cthulhu?


CoC is one of the worst systems ever put together. Fortunately the game is based largely around mood, magical teaparty, and the expectations that your characters will die horribly sooner rather than later. So the system doesn't get in the way that much. But that doesn't mean that the game mechanics aren't a war crime. They totally are. Making something that is game mechanically superior to CoC is not even enough information to tell us whether you've met the bare minimum requirements to not get laughed at.

I have no idea what "magical teaparty" is, but he took the d% roll under, made it into a d20 roll under, and gave skills ranked 1-20, which you check when you pass, and then at some point try to fail in order to improve. Any further similarity to CoC is unknown to me, as I haven't played it. On the other hand, the most negative comment I've ever heard about it came from you, just now.

Wow. You're horrendously wrong here. Degree of Success is incredibly easier with d% roll over. Your Target Number is 100, you roll your dice and add your modifiers. Your degree of success is just the last two digits of the final number.

You've still got to add two double digit numbers it to get that number. Subtraction is marginally more difficult than addition, but dealing with smaller numbers is easier than dealing with larger numbers, and that includes when the smaller number is the result. Also the simple comparison counts for a lot. Consequently, roll over is not objectively easier; either way it's not massively easier (as you assert later on).

This isn't even up for debate. If you want to roll percentile dice and you want to do things with degrees of success or opposed rolls, you should use roll over. Mathematical steps can be counted. Roll Over requires less of them. Also, it provides exactly the same actual chances as Roll Under because it is literally the same system which has been simplified to have all the necessary numbers on the right side of the equals sign.

Well actually it is up for debate, because I'm saying that's not true. I'm fully aware that it's mathematically identical (I'm in the final year of my maths PhD, and I've taught stats to 3rd year undergraduates), however that does not make it identical in practice. You've got two systems with different maxima, you'll get different ideals depending on how you weight them.

The only way that Roll Under ever has an advantage with anything is if you are generating binary success/failure results and you are not applying circumstantial modifiers of any kind. In that very limited circumstance, Roll Under is superior because it resolves with a simple comparison operation instead of requiring addition or subtraction of any kind. If you are doing anything else (opposing tests, modified tests, degree of success, secret difficulties, etc.), then Roll Over is somewhere between "slightly easier" and "massively easier".

How do circumstantial modifiers affect anything? You add the modifier to your skill, then try to roll under it. Easy. If your GM decides to add a hidden modifier, all they've got to do is add the modifier to your score, and then determine the result by comparison when you read the dice out.

For critical successes, in roll under you say "if you passed, and the last digit of the dice is a 0 or a 5, you've got a critical success". This scales with your ability. I can't think offhand of an equivalent system for roll over, except for "did you pass by 50 (or X) or more?" That might be easy, but it doesn't scale the same way, and you've still got that addition to take care of. Plus instead of a gradual chance of critical going from 1-10 as your score goes from 1-100, it stays 0 until 50, then swings up to 50% at 100, which seems rather less elegant to me. And do you seriously need Degree of Success all the time? I don't think so. I'd say that limited circumstance is more the norm.

d% in general is not a particularly good mechanic, unless you use the one thing it's good for, and that's giving percentage chance of success. Not percentage chance of degree of success, just plain and simple binary success, and roll under is where it shines. You can do a lot with that.

If you are advocating Roll Under for an RPG today, you are a relic. This is 2012, and science has progressed. You might as well be advocating THAC0. It's a pointless overcomplication of the rules system. Like THAC0, it simply reverses the signs on some things with no realized or even potential benefit

I'm not advocating THAC0. When we had to use THAC0, we replaced it with the 1e to-hit tables. THAC0 involves scores that decrease as you get better, that's where the problem lies, not in the subtraction. It's an unintuitive system, which leads to confusion as it's difficult to get your head round what you're supposed to do with it. As a point of interest though, THAC0 can be rearranged to be the easiest possible system, provided your GM is willing to share the opponent's AC with you.

Yes, it's 2012, that doesn't mean we need to keep reinventing the wheel just to say "it's better because it's new". People had roll over in the 70s too. I don't know when the first dice pool came out, but I'll bet that's been around for a while. We do, however, have 40 years of experience of what works. I'm aware that roll under is not as popular as it used to be, but with the exception of Runequest or something, I can't think of any games out right now that use d% roll over.

[John Magnum]

For critical successes, in roll under you say "if you passed, and the last digit of the dice is a 0 or a 5, you've got a critical success". This scales with your ability.

Your chance of getting a critical success period scales up, but your chance of getting a critical success over your chance of getting a success stays nearly constant except for a bit of wiggling as your target number goes 13 14 15 16 17. That "scales with your ability" even worse than D&D's system of confirming criticals does.

[spaceLem]

For critical successes, in roll under you say "if you passed, and the last digit of the dice is a 0 or a 5, you've got a critical success". This scales with your ability.

Your chance of getting a critical success period scales up, but your chance of getting a critical success over your chance of getting a success stays nearly constant except for a bit of wiggling as your target number goes 13 14 15 16 17. That "scales with your ability" even worse than D&D's system of confirming criticals does.

It's not that bad. 1 in 10 hits (barring jagginess) will be a crit. That's pretty much the same as D&D's crit on a 19-20/x2. There's no easy way to improve that, except to say "crit if last digit is a 0,3 or 6", or something, which I'll give you isn't very pretty.

D&D's system of confirming criticals isn't wrong, in fact it's reasonably elegant (mathematically speaking); it's just annoying because of the work in the extra roll, and there's that chance of a potential success becoming a failure, which is doubly annoying. There's no let down moment with the d% system.

[Username17]

How do circumstantial modifiers affect anything? You add the modifier to your skill, then try to roll under it. Easy. If your GM decides to add a hidden modifier, all they've got to do is add the modifier to your score, and then determine the result by comparison when you read the dice out.

So let me get this straight: "all" you have to do in order to figure out whether you succeeded and by how much is to have the player add double digit numbers together on their side of the table and then report their modified skill and their literal die roll and then have the MC add or subtract hidden modifiers from the reported skill and then subtract the reported literal die roll from the re-modified modified skill. That's "all" you need. On every die roll.

How do you not see how fucking terrible that is?!

Let's compare to Roll over: the player adds double digit numbers together on their side of the table and reports one number, which the MC then applies hidden modifiers to and gets a final result whose "degree of success" (if applicable) is literally the same number as the tally already achieved. That's less math steps and less reporting steps!

-Username17

[John Magnum]

One in five hits, right, since it's 0 or 5? Anyway, the point is just that your crit-per-hit chance is constant, rather than scaling with ability. I'm not saying it needs to scale with ability, just that in actual fact it doesn't.

[tenuki]

You're describing Lem's method in an unnecessarily complicated way. In case of hidden modifiers, the GM could just ask the stat being tested (if he doesn't have it anyway), apply the hidden modifier himself and let the player roll.

No need to do it ass backwards.

[Username17]

You're describing Lem's method in an unnecessarily complicated way. In case of hidden modifiers, the GM could just ask the stat being tested (if he doesn't have it anyway), apply the hidden modifier himself and let the player roll.

No need to do it ass backwards.

The player submits two numbers. One of them is his character's ability with all the modifiers he knows about. The other is his literal die roll. Then the MC is tasked with doing all the calculations on his own after that.

That is bullshit!

You can get all that information by having the player report one number to the MC. We know this because from 3rd edition on, Dungeons and Dragons already did that. If you are behind the development curve to Dungeons and Dragons, then you are an embarrassing dinosaur. Being behind the development curve is sort of D&D's thing.

-Username17

[spaceLem]

So let me get this straight: "all" you have to do in order to figure out whether you succeeded and by how much is to have the player add double digit numbers together on their side of the table and then report their modified skill and their literal die roll and then have the MC add or subtract hidden modifiers from the reported skill and then subtract the reported literal die roll from the re-modified modified skill. That's "all" you need. On every die roll.

How do you not see how fucking terrible that is?!

Let's compare to Roll over: the player adds double digit numbers together on their side of the table and reports one number, which the MC then applies hidden modifiers to and gets a final result whose "degree of success" (if applicable) is literally the same number as the tally already achieved. That's less math steps and less reporting steps!

I'm not sure where you're getting all that from. No, the player doesn't have to do anything except roll the dice, and tell the GM the result.
Player: "I rolled 42, my skill is 57".
GM: "there was a -20 modifier, so the skill is 37" (calculation #1, apply modifier)
GM: "37 is less than 42, so it's a failure" (calculation #2, simple comparison)
OR if the GM wants to determine degree of success, they go
GM: "37 is 5 less than 42, so that's a fail by 5" (calculation #2, always subtraction)
GM: "here's what happens"

The roll over version goes like this:
Player: "I rolled 58, my skill is 57, so that's 115" (calculation #1, always addition)
GM: "there was a -20 modifier, so that's 95 (calculation #2, apply modifier)
GM: "95 is less than 100 (calculation #3, simple comparison)
OR if the GM wants to determine degree of success, they go
GM: "95 is 5 less than 100, so that's a fail by 5" (calculation #3, simple subtraction)
GM: "here's what happens"

It's really not that much different.

[tenuki]

You're describing Lem's method in an unnecessarily complicated way. In case of hidden modifiers, the GM could just ask the stat being tested (if he doesn't have it anyway), apply the hidden modifier himself and let the player roll.

No need to do it ass backwards.

The player submits two numbers. One of them is his character's ability with all the modifiers he knows about. The other is his literal die roll. Then the MC is tasked with doing all the calculations on his own after that.

That is bullshit!

Your argument is. What's to prevent the player from reporting "success by X before hidden mods"?

[Username17]

You're describing Lem's method in an unnecessarily complicated way. In case of hidden modifiers, the GM could just ask the stat being tested (if he doesn't have it anyway), apply the hidden modifier himself and let the player roll.

No need to do it ass backwards.

The player submits two numbers. One of them is his character's ability with all the modifiers he knows about. The other is his literal die roll. Then the MC is tasked with doing all the calculations on his own after that.

That is bullshit!

Your argument is. What's to prevent the player from reporting "success by X before hidden mods"?

1. That is not how anyone plays it, including Space Lem who specifically called it out being played the other way.
2. That's still shittier than reporting a number rather than a difference. See 2nd edition THAC0 vs. 3rd edition DCs.

Saying "it's not much worse than roll-over" is actually exactly the point. It's worse. Not a lot worse, but nonetheless measurably so. There is no reason to do it the shitty way instead of the better way.

-Username17

[tenuki]

That is bullshit!

Your argument is. What's to prevent the player from reporting "success by X before hidden mods"?

1. That is not how anyone plays it, <snip>

Nobody? Nowhere in the whole wide world? I admire your confidence, if nothing else.

2. That's still shittier than reporting a number rather than a difference.

How is a difference less of a number than a sum? Dude!

There is no reason to do it the shitty way instead of the better way.

There are no objective differences between roll-over and roll-under systems, as long as you stick to linear mods and constant intervals between degrees of success/failure. Roll-over systems are popular because kids groove on "BIG NUMBER = GOOOOD".

[Username17]

How is a difference less of a number than a sum? Dude!

So you're saying that THAC0 is fine because the player can report that "they hit AC 3"?

-Username17

[virgil]

I'm trying to understand the advantage in keeping Roll Under without it being anything other than an Appeal to Tradition. From a detached PoV, as I don't really play in d% resolution systems, I can only see the logic in Roll Over.