How to write no rules

[Grek]

1) I thought that 20d6 was actually the sort of shit that high end specialized dice pools through in Shadowrun already, so I figured 20 is a number people actually use.

Rolling 20d6 and counting hits is a much quicker process than rolling 20d6 and summing the results. I know of no games which require you to do the later regularly.

Also, as I was taught it, a standard deviation (s) is the mean deviation from the sample mean (x). We care about it because 68% of values will be within x±1s, 95% will be within x±2s and 99.7% will be within x±3s.

[Kaelik]

Also, as I was taught it, a standard deviation (s) is the mean deviation from the sample mean (x). We care about it because 68% of values will be within x±1s, 95% will be within x±2s and 99.7% will be within x±3s.

But that's stupid. Because it directly contradicts what people have been saying.

People have claimed that being able to have a second deviation on the RNG or even a third, is somehow better than not having them on.

But that's the exact opposite of "everyone is within 1." I care about the first one, but the first one by definition exists in any RNG, why do I care about the second or third one being on the RNG?

[PhoneLobster]

So actually, one of you is just misleading people who read this shit and aren't into this stuff.

I would like to point out I am entirely happy to discuss the actual function of these numbers WITHOUT throwing around terms like "Standard Deviation" or "Probability Distribution".

I can and HAVE discussed this using fairly simple discussion of nice easily accessible terms like "chances of success" and "percentage" and "average".

Seriously PL, do you want a fucking cost-benefit equation for it?

Why yes. You can start with describing the benefit. Once we actually HAVE a benefit THEN we can start comparing it to costs and draw a line where too many dice are too great a cost.

But until we know that say at least a 2d6 IS actually beneficial we can't actually determine at what value of N a potential Nd6 is too costly.

Without a benefit no matter how small the cost (and at the lowest end a 2d6 is only moderate in cost) is still a cost without benefit. See the 2d6 adds relatively little in complexity and opaque "costs" to an RPG, even in comparison to a 3d6 (since those costs basically progress at a non-linear rate) but still unless I get something for that using a 2d6 instead of a 1d10 or 1d12 the 2d6 is only sending my RPG backwards, however small a distance backwards it may be.

SO anyway lets look and discover your cost benefit scenario starts not with a benefit (much less a measurable comparable benefit) but with a mere assumption of benefit.

If you desire a system where adding the same static number means less to your odds of success the more times you do it for any single given roll

So yes. IF you desire that.

But why do you desire that EVER?

What good function does it perform that makes your RPG better, more playable, more transparent, more efficient?

The whole argument has been about discussing how doing the thing that you simply assume to be a benefit is actually a bad thing to do.

The response has been... because you can do it it is therefore good.

With that sort of reasoning there IS no cost that outweighs benefit.

[Username17]

Also, as I was taught it, a standard deviation (s) is the mean deviation from the sample mean (x). We care about it because 68% of values will be within x±1s, 95% will be within x±2s and 99.7% will be within x±3s.

But that's stupid. Because it directly contradicts what people have been saying.

No, that just contradicts what PL was saying, where he claimed that a data set can't have a second or third standard deviation. A claim so ignorant that it boggles my mind and makes me skim the rest of his posts.

We are talking about action resolution like the action resolution of something like getting into college, or being born really tall. As in something like "to do this action successfully, you'd have to be 2 standard deviations from the mean!" or "you can almost certainly do this, so unless your results are skewed outside 2 standard deviations, you'll succeed." Because when your RNG is "roll some dice and add them together" then that determination is extremely easy. Because the standard deviations in expected data generated by repeated rolls of your dice happen to each fall a static number away from the mid point.

People have claimed that being able to have a second deviation on the RNG or even a third, is somehow better than not having them one.

You always have a second standard deviation. Or something pretty close based on the granularity of your RNG. The question is whether it is simple addition to determine where that point is (as it is on 3d6) or something more complicated (as it is on a d20).

But that's the exact opposite of "everyone is within 1." I care about the first one, but the first one by definition exists in any RNG, why do I care about the second or third one being on the RNG?

No one said this. Go back and read the segment you quoted. What the fuck?

-Username17

[norms29]

I've got my Statistics II text book in my lap, the definition of Standard deviation is in chapter one. for those interested, StDev is defined as the square-root of ((the sum of the squares of the differences of each outcome from the mean) divided by the number of outcomes minus one)

so in a D20's case the Standard deviation is SQRT(665/19) wich comes to approximatly 5.92

Which brings me to my next point which is..

Will the both of you SHUT THE HELL UP!, it's clear that neither one of you know what a standard deviation is. you both seem to have definitions that are close to the real one, but not close enough.

First, frank.

On a d20, that number is largely unknowable. If you're starting at DC 11, you need a +6 bonus to shift one standard deviation, and you'll need a +3 to shift another standard deviation.

this is just wrong, what you are describing is a property that standaard deviations have on a normal distribution curve.

and the fact that a fixed modifier doesn't have the same effect when repeated on a flat distribution that it does on a bell curve (No shit sherlock!) is a wholly seperate issue.

There IS no such thing as a "next standard deviation" anything outside the FIRST standard deviation is not a standard deviation it is, definitionally, a deviation that is non-standard.

This is also wrong, actually much more wrong than Frank. atleast Frank was accuratly conveying an expected property of Standard Deviation, although one he shouldn't have conflated withit given the topic of discussion.

you are repeating a claim that you would know is false just by sleeping through statistics I one of the first things they tell you, and harp on the whole semester, is that on a normal curve 99% of outcomes fall within THREE standard deviations of the mean.

Also, your definition of absolute deviation isn't remotly close either.

http://en.wikipedia.org/wiki/Absolute_deviation

[Username17]

First, frank. this is just wrong, what you are describing is a property that standaard deviations have on a normal distribution curve.

and the fact that a fixed modifier doesn't have the same effect when repeated on a flat distribution that it does on a bell curve (No shit sherlock!) is a wholly seperate issue.

Uh... you just "contradicted" me with my own point. Which makes it not a contradiction. Rolling 3d6 approximates the normal curve very closely out to 2 standard deviations, and thus has the properties that the normal curve has within that region. A flat distribution does not have those properties.

The entire point of this discussion has been that the normal curve has real properties that proponents of rolling 3d6 really know about and really like and really use because they like those properties. That's it. It's nothing more complicated than that.

No exciting magical claims are being made for rolling 3d6. No one here is saying you have to use it. It's just being said that 3d6 RNGs have some of the properties of the normal curve and some people like those properties and use 3d6 as their RNG because they like those properties. In short, while you did just tell me that I was totally ignorant and should shut up, you actually just confirmed exactly what I've been saying the entire time, so I'll take your vigorous angry agreement in good humor.

-Username17

[norms29]

First, frank. this is just wrong, what you are describing is a property that standaard deviations have on a normal distribution curve.

and the fact that a fixed modifier doesn't have the same effect when repeated on a flat distribution that it does on a bell curve (No shit sherlock!) is a wholly seperate issue.

Uh... you just "contradicted" me with my own point.

actually I corrected your terminology, as a I said, you've been using (and in the post I quoted, explicitly defined) the term Standard Deviation to refer to the properties that the standard deviation has under a normal distribution, which was obscuring the point you're trying to make. Especially since this discussion is about whether or not to use a RNG that approximates the normal curve, AND they were the properties that had you arguing for a RNG that approximates the normal curve in the first place, so equating the two was not an acceptable use in this context

The entire point of this discussion has been that the normal curve has real properties that proponents of rolling 3d6 really know about and really like and really use because they like those properties.

yeah, yeah, I 've played Gurps, I know how it works

[TarkisFlux]

But why do you desire that EVER?

Wut?

Look, it's fucking obvious that you don't care about modeling edge cases with a much smaller chance of success than you get with a flat system of similar range, you don't care about a diminishing returns bonus system, or any of the other things that a curved system does and has been explained over the course of this thread. The differences have been described at length, and I'm not really of a mind to attempt to explain them as benefits again when you've already ignored or openly disregarded them several times. The differences are not useful to you personally, and you harbor a dislike of the mechanic because apparently you feel it's less transparent and that should be something we care about or whatever. But if you needed those things to model whatever you had in mind, using a curved system is a better call than a flat system with the obtuse kludges you've suggested over the course of the thread to attempt to mimic properties of a curved system as tehy were brought up.

It's a fucking design call, and it's as much preference as picking what best fits the game you're trying to emulate. No one here has ever said curved rolls were universally better and you should use them all of the time and love them forever, and if someone said something to that effect and I missed it they're fucking wrong. They've said they were different, brought properties to the table that were difficult for a flat system to duplicate efficiently by sacrificing properties of a flat system, and that people who designed around them understood how they worked and weren't being intentionally obtuse. If you care about and can use their properties, they are better than a flat system; if you don't care about them they are worse. That's it.

[souran]

(lets note in passing here that standard deviation as a value has no meaning outside of this first increment, there IS no such thing as "a second standard deviation", that is, oddly non-standard)

Holy Shit! This is fucking vital to my job as an engineer! If there is no such normal thing as 2 standard deviations then all those federal regs that use that to determine acceptance criteria are just gobbledegook!

I need to foward this thread to my bosses becasue our whole "six sigma design" is based around this idea of multiple standard deviations.

[PhoneLobster]

The entire point of this discussion has been that the normal curve has real properties that proponents of rolling 3d6 really know about and really like and really use because they like those properties. That's it. It's nothing more complicated than that.

Would you care to explain any actual REASON why it is a useful or helpful mechanic OTHER than someone THINKS IT IS.

Someone like the idea of the mechanic, that is not good enough. It hasn't been good enough for you with anything else you have ever discussed around here. Why this?

No exciting magical claims are being made for rolling 3d6.

Except that liking it is enough justification to use it even if it does bad things mechanically to your RPG system.

No really give us a REAL REASON why those numbers are good. Explain how they are actually transparent.

And most of all tell us why you freely break your OWN "nothing more complex than addition in your head" complexity limit for RPG mechanics.

In fact tell us why you break your NO HORSES INSIDE rule on this one. People like centaurs, but it doesn't make THAT RPG mechanic OK with you without further justification for inclusion.

you are repeating a claim that you would know is false just by sleeping through statistics I one of the first things they tell you, and harp on the whole semester, is that on a normal curve 99% of outcomes fall within THREE standard deviations of the mean.

A large part of the problem here is that I have been dealing with a discussion leading up to this point where people were claiming to be adding a "significant 30%ish modifier" several times.

Dealing with that is certainly not easy. And when that claim evolves into... "well er its standard deviation, yeah! thats it!" well... And since they claimed to mean some form of standard deviation that occurs without any relation to bell curves or normal distribution...

Remember the trivially correct quibble early that ANY data set can have a standard deviation. Including ones that do not accurately adhere to normal distribution.

One can only assume they thought they were talking about those things. In which case really, the standard deviation ONLY tells you in what range around the mean "most" results fall.

Indeed note they rather specifically criticized the 1d20 because they couldn't add the standard deviation multiple times.

They wanted to talk about standard deviation in total isolation and really that is it. Because in isolation, no, you don't get to make those claims about standard deviations, they DON'T do that for all data sets and these guys just told me I wasn't allowed to talk about bell curves or normal distributions.

Now when they eventually wanted to talk about the three sigma rule I was more than happy to acknowledge it's existence and it's behavior. Go have a look.

[PhoneLobster]

and I'm not really of a mind to attempt to explain them as benefits again

Describing a thing is not the same as telling me why I should use it.

You have told me how a curved RNG works. Which is really dumb because I KNOW how a curved RNG works thank you very much.

But WHY would you use one.

And the answer...

But if you needed those things to model whatever you had in mind

Just begs the question, WHY does it need to model such behavior.

No really, why?

We already know that even slightly complex math is bad just look at everything ever where "Role Master" comes up.

We already now that incredibly rare success chances are incredibly stupid. Just look up everything Elysnar (or whatever his dumb name was) ever got involved in.

So why would you do that? Why are they suddenly good. What are you actually doing with these numbers that is good.

The example we have so far is being completely blind more than once.

With the added proviso of a completely arbitrary definition of the mechanical impact of being blind and how it should apply.

That isn't a good answer. Even as far as completely manufactured favorable examples go it is in fact an incredibly poor example.

It's a fucking design call, and it's as much preference as picking what best fits the game you're trying to emulate. No one here has ever said curved rolls were universally better and you should use them all of the time and love them forever

No they haven't. But interestingly they have had this argument specifically here, despite being asked not to, in a thread specifically about high transparency low complexity "lite" RPG mechanics.

Which is really, hilarious especially with all the concern trolling "you are being unreasonable because we never tried to inflict our ideas on you, waaah waaah waaah"

But really, I'm cool, I've long since given up on so much as one of you meeting MY requirements for justifying the inclusion of such an RNG in the sorts of mechanics that are actually on topic for MY thread.

At this point I would be happy if you can actually JUST point out a strong example that meets the well known requirements for "good mechanics" held up by the curvy RNGs original proponent on this thread, Frank.

How does it meet his "no multiplying or dividing in your head" rule?

Why is it allowed to break the "no horses inside" rule?

In short why is it allowed to wave away the basic question "Why is this a good thing?"

[NineInchNall]

A large part of the problem here is that I have been dealing with a discussion leading up to this point where people were claiming to be adding a "significant 30%ish modifier" several times.

No.

The claim is that on a dice set approximating a bell curve, adding a small modifier (such as +2) has a significant effect on its own and a lesser effect in combination with other modifiers.

Jesus Christ, man. That's trivially true.

[TarkisFlux]

It isn't allowed to wave that question. No one has indicated that it has. They have been attempting to answer a throw away comment you put in after your big post, which makes this annoyingly on topic despite your protestations to the contrary. And I've been bringing it all back to your original post before I went to find something productive to do, it just got so far afield with the semantics bullshit that it took me like 4 posts to be happy doing it...

You get multiple bonuses or penalties to matter less with a curved RNG. You might desire this behavior tf you wanted to allow multiple bonuses to a roll, and have later ones matter less to decrease the desire and benefit of players bonus hunting. You could achieve this same result on a flat system, but requires division in your head. However, given your stated goal in the Formal + Arbitrary = Arbitrary section where you indicate that there should be 1 or less bonuses in a rules light situation (though this stance in unclear when paired with the Attributes section) this is not a net benefit of using a curved system because it is unlikely to come up.

The clustering of results in the center and corresponding reduction of edge cases would also impact the Taking Turns section of your design goals, most likely by increasing the chances that people went at the same time if you were using the "whoever succeeds best goes first" system. I believe this is actually a net benefit when combined with the Influencing results with resources point, in that having your results closer together is more likely to mean that a sufficient boost from this hypothetical resource will push you over the opposition's roll, thus placing more direct control of the narrative into the player's hands. Unfortunately, as you are silent on the value of player control (or I missed it), I am uncertain whether this should be seen as a net benefit against the design specs you put out.

You have a non-static percent chance of achieving the next higher result on your die. This makes edge cases vastly less likely than central cases. This may or may not be a better fit for your stated Degrees of Success goal, depending on how you want your degrees of success structured. The non-static chances of success may also better fit your stated goal in the Co-operation and contagious failure of simply increasing the TN instead of the number of rolls. You have indicated elsewhere that you don't consider needing to model tiny edge cases a worthwhile thing and I'm not sure how small you even want to be concerned with. Still, I'm going to call it a possible benefit but not necessarily one, assuming we can address your transparency concerns anyway, which I'm getting to.

A curved RNG is neither a benefit nor a hindrance to most of the rest of areas you specified, as they are mechanics independent areas and just require a more clear hand at the rules and a better designer at the character sheet.

And all of the above is apparently irrelevant if it fails to meet your transparency requirements. I do not believe that it is a significant decrease in transparency however, especially when combined with with the Good Advice tool from your design goals. That tool exists to tell GMs and players how hard something is so that they can more efficiently go about achieving their intended goals. If you write it well, any single roll resolution system is going to be similarly transparent, be it a flat roll, a curved roll, or a dice pool roll. If you lack that tool, then this is perhaps less obvious than a flat system, but if you lack that tool you've already failed a different aspect of your design goals more completely. Multiple successful rolls of any type are actually not transparent and bullshit and should be written out of the system entirely as a result, regardless of resolution die mechanic selected.

So why is it a good thing? It's no more good or bad than any other single roll resolution mechanic. Why might it be a good thing in relation to your specific design goals? If you were going with a roll + attribute + maybe arbitrary bonus, and you wanted to decrease the value of the arbitrary bonus unless it was larger than the attribute bonus, this would be a good thing. If you wanted to decrease edge cases so that player controlled action boosters were more important, this would be a good thing. If you wanted to fit more extreme edge cases into your degrees of success system and were willing to sacrifice some results success percentage granularity around the central cases, this would be a good thing.

[PhoneLobster]

It isn't allowed to wave that question. No one has indicated that it has.

They have. After all the answer is STILL "but you might like curves".

Even now your answer is "You might like curves".

Sure you name dropped some of my goals but basically what you did after each was "and you might like a curve with that!".

You might desire this behavior tf you wanted to allow multiple bonuses to a roll, and have later ones matter less

Why would I want that? that is a crazy thing to want.

And doesn't your suggested reason...

to decrease the desire and benefit of players bonus hunting.

Contradict not only my own advice but also YOUR own suggestion later that the curve would be cool because you seem to think...

thus placing more direct control of the narrative into the player's hands

But regardless why would I go for that particular "benefit" when the "benefit" of

The clustering of results in the center and corresponding reduction of edge cases

is to make the basic taking turns rule not function more often while simultaneously significantly reducing the granularity of the system in the all important mid range.

This makes edge cases vastly less likely than central cases. This may or may not be a better fit for your stated Degrees of Success goal

Why the hell would I want to make higher degrees of success significantly less likely to happen? Do I want to devalue the entire mechanic all of a sudden? But again basically the "benefit" given is "but you might LIKE it!!!"

But still what you do is..

Still, I'm going to call it a possible benefit but not necessarily one, assuming we can address your transparency concerns

Only then...

And all of the above is apparently irrelevant if it fails to meet your transparency requirements. I do not believe that it is a significant decrease in transparency

So story over. It is a mechanic that has going on one side of it "you might want a curve, maybe" and on the other against it "fails at transparency". That makes it bad. Plain and simple, and even you don't provide adequate excuse making for that in the rest of the ramble.

If you were going with a roll + attribute + maybe arbitrary bonus, and you wanted to decrease the value of the arbitrary bonus unless it was larger than the attribute bonus

Again, if for some reason you wanted to.

Also No it only does that at certain points in the range and considering I suggest you DO use ALL your range that means sometimes using a curved mechanic does the opposite of that and all the other things you discussed.

[TarkisFlux]

It isn't allowed to wave that question. No one has indicated that it has.

They have. After all the answer is STILL "but you might like curves".

Even now your answer is "You might like curves".

Sure you name dropped some of my goals but basically what you did after each was "and you might like a curve with that!".

No. What I did was explain how a curve supports your fairly vague goals and why you might choose it over a flat RNG because of behaviorial changes which you have not previously said you did not want. You sketched a fucking general diagram and I went through and said how a curve met those goals. So yes, "you might like a curve with that". You might also like a flat with that. I don't know because your points are generic enough to preclude neither. Why you would want one or the depends on what you want the game to fucking look like and do. Do I need to tell you what you want in your rules light games now too?

Fuck your goal post shifting. I can't give you a reason to choose one or the other without more details of what a specific rules lite system is supposed to do. My intent was never to tell you that you should take or even like a curved RNG, cause I don't fucking care which one you use. I just wanted to address your bullshit complaint that curved RNGs fail to do the things that you think rules lite systems need to do.

The rest of your response was a pile of confused shit that leaves me wondering if you really need everything spelled out step by step. Like you're incapable of making the same logic steps that I am or missing one or something. But then I get to this shit,

Only then...

And all of the above is apparently irrelevant if it fails to meet your transparency requirements. I do not believe that it is a significant decrease in transparency

So story over. It is a mechanic that has going on one side of it "you might want a curve, maybe" and on the other against it "fails at transparency".

where you take what I say, quote it even, and the proceed to say that it the exact fucking opposite. "Not a significant decrease in transparency" does not equal "fails at transparency". Seriously, what the fuck? You don't even attempt to explain how you use that to justify your opposite position. That goes beyond misunderstanding and right into deliberately obtuse land, which kills any desire I had to converse further with you on the subject. Your original post was interesting, but trying to talk past your hatred of this mechanic and your defensive "why should I like it" crap isn't, and it isn't going anywhere.

[PhoneLobster]

Oh sorry, I assumed "I believe" wasn't you know, evidence or indeed an argument in support of a belief. You know, just a belief.

Those things aren't actually self justifying you know.

[TarkisFlux]

Choke on a barrel of cocks and die PL, and take pics so that we might derive some amusement from you in death. Of course it's not self justifying. That could be why I followed that statement with an argument in support of that position. You ignored it of course, like you've ignored or distorted everything else that doesn't fit your retarded narrow view of the proper way to do rules lite (not that you've even properly justified that, but hey, that's off topic now). Which either makes you an idiot, severely obtuse, or a really successful troll since you got me back here after all.

Here, since you apparently didn't read it the first time, I'll quote myself

especially when combined with with the Good Advice tool from your design goals. That tool exists to tell GMs and players how hard something is so that they can more efficiently go about achieving their intended goals. If you write it well, any single roll resolution system is going to be similarly transparent, be it a flat roll, a curved roll, or a dice pool roll. If you lack that tool, then this is perhaps less obvious than a flat system, but if you lack that tool you've already failed a different aspect of your design goals more completely.

Hey, look at that, support for a statement of belief! And since you seem to need sentence parsing, logic, and processing help, I'll even explain that passage to you and expand on it. Any single roll mechanic can be made equivalently transparent with any other as long as you give an explanation of what to expect. People do not need to know how something works as long as they understand what it does under the various circumstances it might be applied in. If that's insufficient for some reason, which you haven't made a case for, you can go extra transparent on it and add a table with percentages and odds or an explanation of how it works. And this should go in that advice section, so that it is clear how your chosen die mechanic works and what people are actually doing when they set TNs or give bonuses.

Which is no more or less than you would have to do for a flat or any other mechanic for a desired level of transparency. And is also somehow transparency fail. GIANT FROG much?

[mean_liar]

On a side note, PL, you have an incredibly obtuse writing style. Parsing meaning out of your posts is comparatively a lot more effort than just about anyone else here who's also literate.

[PhoneLobster]

On a side note, PL, you have an incredibly obtuse writing style. Parsing meaning out of your posts is comparatively a lot more effort than just about anyone else here who's also literate.

I know.

I rarely manage to write anything short and punchy.

But really, with the current devolution into a massive pointless "Jargon-off" don't expect me to be at the top of even my mediocre version of short and punchy.

Aside from that I try to reserve my limited short an punchy energies for RPG material I write for my own uses. Rambling discussions with raving lunatics can go hang.

PS Tarkis Flux
Ignoring your apparent explosion of inanity did you notice the little bit where I pointed out that for half the RNG the behavior you called beneficial for various factors is in fact directly inverted.

You should think on that a little.

[TarkisFlux]

Yes, and would have given you credit for that point (which is still somewhat retarded unless you actually expect people to have TNs on the bottom of the RNG, which you have given no indication of) had I felt it was worth any further attempt at conversation with you. Speaking of which, there's a button for that around here somewhere....

[PhoneLobster]

(which is still somewhat retarded unless you actually expect people to have TNs on the bottom of the RNG, which you have given no indication of)

Well since basically every RPG system in existance uses TNs accross its entire RNG if not BEYOND I didn't think I especially HAD to point out that no, every TN should not be a "coin flip or worse" scenario.

Declaring that the best TN PCs should face is effectively a coin flip seems like the retarded thing to me.

But heylets quote the title of one of the sections of my example "improved" lite RPG system from this thread.

TNs lower than 1 or greater than 10

That was a sample system which ran off a 1d10 RNG by the way.

had I felt it was worth any further attempt at conversation with you. Speaking of which, there's a button for that around here somewhere....

I see. So my point undermining every single benefit you claimed curves had is moot because you found my effrontery in questioning those benefits to be insulting.

Well with debating skills like that there is certainly no arguing with you now is there!

[MGuy]

PL: Can you sum up what exactly your point actually is? Currently I don't see how it undermines anything at all.

[TarkisFlux]

You are already on ignore PL. You are not on ignore because I found your willingness to question my posts insulting. I have previously admitted fault on this board and moved on, and would have done the same in this case had I not been over attempting to communicate with you. There is no arguing with me on this matter, or any other in the foreseeable future, because we're not having an argument. We are having a failure to communicate entirely. And when that happens with people on the internet I put them on ignore; I prefer no conversation with them to a failed one. I'll disagree with people all day, but when I can't actually communicate with them I walk. This is the last direct response you will get from me until I take you off of it. As I am the one walking away I offer you the last word in our exchange.

---

If you wrote the TN range quote in this thread, I missed it and apologize. I expected TNs above the top of the RNG, but not below the bottom of it.

Your point does not undermine every benefit curves have, it only undermines their benefits with respect to bonuses in areas where you are already very likely to succeed. Which isn't a concern, as keeping people away from auto-success of easy tasks was never specified as a design goal. So it's a small thing, that you are attempting to make a big one for some reason.

It is also easy to incorporate into the advice section. "If you need to roll better than 11 to reach your TN, any bonuses you can get will significantly increase your chances of success. In fact you will need bonuses to make rolls with a TN over 11 consistently. Once you need to roll less than 11 to reach your TN, bonuses start to matter less. Each additional point of bonus beyond this point increases your odds by a smaller amount. So, while it might be tempting to get every bonus you can swing, you really don't need to worry about piling them all up."

Yes, I have just changed the wording, but it gets the same idea across. I have done so because it is easier to incorporate your concern in this wording while still expressing the diminishing returns benefit of a curved system, which is not a particularly strong one in a game with very few modifiers.

Regarding your concern that breaking turn ordering by making simultaneous actions more common does not increase player control, I would point out that as it stands you have a random system of determining turn order in which the only player control is the decision to spend or not spend their special resources. Increasing the likelihood that things happen at the same time means that there are more opportunities for a player to spend their special resources and take their turn first, regardless of the size of the bonus gained from that resource. Clustering events also increases these opportunities (for any given modifier size) over a flat RNG. Breaking a system so that it requires active player choice more often, even when that choice is to not use the resource, is increasing player control of the narrative.

[PhoneLobster]

PL: Can you sum up what exactly your point actually is? Currently I don't see how it undermines anything at all.

Well basically Tarkis is all like "But if you WANTED bonuses to shrink as you went along curves would be totally cool!"

But if you starting TN is lower than the peak of the curve then bonuses GROW as you go along. Which means if you wanted them to shrink then half the RNG is actually doing directly the opposite to what you want.

You are already on ignore PL.

It is OK to put me on ignore.

However it is not good form to put someone on Ignore AND continue to argue with them in a petty attempt to get some sort of public last word while screaming "lalalalalala I CAN'T HEAR YOU!!!!"

What are you? Four years old? Get off my thread.

[Murtak]

Growing or shrinking does not matter. Getting pushed off the RNG is the important thing. So far all examples have been in the upper portion of the RNG, that is why the numbers have been shrinking. But the situation is quite similar when the task gets so easy you can not fail.

In a curved RNG system you are basically in one of the following positions: In the middle of the RNG (at the center, or right next to it), significantly below the middle, or significantly above it. If you are not in the middle part of the RNG you do not care much about bonuses, especially small ones, as they do not alter your chance of success much. If you are in the middle however even a +1 can make a big difference.

This means that for level-appropriate tasks (or skill-appropriate in levelless systems) you care about getting that last small bonus point. It also means that is damn hard to actually push someone off the RNG entirely. So in DnD terms, two level 10 fighters dueling would care about taking the higher ground, swinging on chandeliers and the like. A 10th level fighter fighting some gnolls wouldn't care about higher ground, or being flanked. In DnD as-is however the dueling fighters don't much care about getting a +1 or even +2. Sure, they will take the bonus if they can get it, but hitting on 11+ and hitting on 13+ is barely distinguishable. The gnolls on the other hand will go from hitting on a 19 (or 20 or not being able to hit at all) to hitting on a 15 and might suddenly be a threat.

I prefer the results of the curved RNG for most situations I can think of. Your mileage may vary.