Healing surges and other such fail.

[Murtak]

I am utterly incapable of asssuming that because if we roll 20 d20s that we will roll a 1 with absolute certainty and that we should make it so that a 1 succeeds for PCs accordingly.

No one is claiming a 1 has to succeed. All I have seen claimed is this: A 5% chance of character death is too high for a regular or even a challenging encounter.

If you want a real chance of failure you will have to look at penalties besides death - such as the heroes having to run away (which I have yet to see you acknowledge as viable).

Its not viable when there's no reason for them to run away, such as being killed if they don't.

Again, no one is saying the heroes have to be actually invulnerable, merely that, if competently played, they should not die. "Competently played" may in fact include "we have to run away or get killed". It is however, important that they do in fact get this chance to run away.

Then roll 20d20s, using any damn d20s you want, and present the results. How long until you get all of 1, 2, 3, 4, 5, etc. 5% of the time each?

Goddamn it, you are either trolling or incapable of computing multiple die probabilities to a degree I can barely imagine. Getting one of each number is rare, yes. Getting at least one of any single number is not.

Until you get that, I will laugh at your math and ignore it. Because hypothetical math with no connection to reality is meaningless. The actual fact that I rolled 2s, 3s, and 5s on 6d6 yesterday is not.

Oh dear. Guess what, I have seen people fail to roll even a single 6 on over 60 dice. According to your math that means the task was way too hard, right? I mean, clearly those 6s are harder to roll than statistics would imply.

And please pretend you are capable of reading and capable of acknowledging that your hypothetical probability is at best capable of predicting what is more likely to happen with a given roll or set of rolls...not something that dictates that my twentith dice roll will be a 1, because there's a 5% chance of a 1 and I haven't rolled one yet.

No one is claiming this. In fact I would vehemently argue against this. Dice have no memory. Your chance of getting a 1 on a d20 is 5%, no matter what you rolled in the past.

Run away to fight another day
- We adjust our system to make sure heroes only die after they are beaten down, smacked around a couple of times and then coup-de-graced.
- Being smacked around penalizes your offense but not your defense.
- We ensure that the heroes can still escape while being smacked silly.
- All of this applies to memorable villains too.
What does this proposal not do that you wish a system to do?

1) It fails to make NPCs the same level as the PCs as capable as they (the PCs) are.

That would be the "all of this applies to memorable villains too"-part.

3) It fails to avoid Conantics, who can slay 10,000 Uruk-hai single handedly. Unless taking on all ten thousand is "not competently played", which isn't supported by making his chance against any one of them 0% unless you have a "mobbed" rule hiding out there.

I haven't even suggested a rule either way - but it is not hard to come up with a viable system. Example: You "die" once (fail a save-or-die, run out of HPs, etc.) - except instead of dying you end up "beaten up", which makes you fail all attacks and non-harmless spells 50% of the time. Die again and you fail 90% of the time. Die again and you are unconscious.

How does that sound to you?

[Tshern]

Are you aware of the fact the chance of a person being named in any given way varies a lot depending on where you live and how old you are? While the people you know (a lot of them are in the same age category as you are, I suppose) might have no Humphreys among them, it might still be a popular name around the USA. Just because you know a certain number of people named in a certain way doesn't mean you can draw conclusions based on it.

Well, apparently you can, but you shouldn't.

[Murtak]

The point is, I know more people with my name than with the name Edward (I can't recall meeting anyone with that name). Despite the fact that statistically there may be more Edwards.

So those statistics are all well and good, but they don't always work in any given occasion.

Of course they don't. Statistics deal in averages. Let enough people roll the dice and some of them will fail and die in a freak accident. And those same averages indicate that out of the trillions of things that happen to each of us some of them will be exceedingly unlikely if viewed on their own (such as not getting a single 6 on over 60 dice).

I'm quite sure everyone of us can present such an event from their own personal life but that does not mean such events are likely.

How long until 5% of your rolls are 1s -by actual dice you are rolling-, not the ideal of a perfectly balanced die? Or 2s? Or 3s? Pick any of the numbers.

There is no perfect balance.
No one is claiming it exists.
Not one bloody person.

The odds however are very, very clear. And the higher the number of tries the closer your actual balance gets to the actual odds.

[MartinHarper]

Martin: Any of 1, 2, 3, 4, 5, 6. Supposedly, there is a 16*% chance of rolling any one of those that actually means anything to what the odds of me beating someone's armor of 5 in Battlefleet Gothic.

Well, that's true. If you roll a six-sided fair dice, it has a 1/6 chance of rolling a 1, by definition. Probability maths does not say that therefore 1/6 of all rolls will be a 1. Probability maths does not say that therefore if you roll six dice, you have a 100% chance of rolling a 1.

How long until 5% of your rolls are 1s - by actual dice you are rolling-, not the ideal of a perfectly balanced die?

That's a modified drunken walk problem in nineteen(?) dimensions, and it has been a long time since I've done one of those. Also, I would need statistics on the variance in bias of actual dice from ideal dice, and I don't have those handy. If you're interested, I suggest you read some papers on the drunken walk problem, gather statistics on the bias of mass-produced dice, and work it out yourself. I'm not interested, because I don't see how it's relevant to the thread.

[Elennsar]

No one is claiming a 1 has to succeed. All I have seen claimed is this: A 5% chance of character death is too high for a regular or even a challenging encounter.

So is a 0.5% chance according to...damnit, forget who. Virgileso? Someone a few pages back. Will drop this if I'm misremembering.

Again, no one is saying the heroes have to be actually invulnerable, merely that, if competently played, they should not die. "Competently played" may in fact include "we have to run away or get killed". It is however, important that they do in fact get this chance to run away.

Which is not being denied anyone other than when things happen that would interfer. Trying to run away with a broken leg isn't recommended, and not having broken legs isn't really appealing.

Goddamn it, you are either trolling or incapable of computing multiple die probabilities to a degree I can barely imagine. Getting one of each number is rare, yes. Getting at least one of any single number is not.

Which is perfectly capable of happening for a check that won't cost you anything. A 1 on an attack roll (barring fumble rules with extreme outcomes, which no one is arguing for) is incapable of killing you. You are quite capable of rolling that 1 there and it won't kill you. Or a 1 on a Search check to find a clue. Isn't the idea behind taking 20 that you can roll a 1 without critical consequences so you "roll' until you get a 20?

Oh dear. Guess what, I have seen people fail to roll even a single 6 on over 60 dice. According to your math that means the task was way too hard, right? I mean, clearly those 6s are harder to roll than statistics would imply.

Thank you for showing how actual dice don't nicely conform to probability and sometimes people have a "streak of bad luck", whatever that means.

No one is claiming this. In fact I would vehemently argue against this. Dice have no memory. Your chance of getting a 1 on a d20 is 5%, no matter what you rolled in the past.

Then do it to the people who are screaming that a 0.5% chance of character death in every encounter equals a 100% chance in 200 encounters.

That would be the "all of this applies to memorable villains too"-part.

You forgot the NPC allies. Assuming your PCs have any.

I haven't even suggested a rule either way - but it is not hard to come up with a viable system. Example: You "die" once (fail a save-or-die, run out of HPs, etc.) - except instead of dying you end up "beaten up", which makes you fail all attacks and non-harmless spells 50% of the time. Die again and you fail 90% of the time. Die again and you are unconscious.

How does that sound to you?

Make this work for the guys who can't hit your AC 40 because they only have +5, or anything else below +20 (so they can't eliminate your hit points) and we have something worth discussing.

I don't want PCs dying right and left. But I do want to have the fact that charging the dragon is capable of getting me killed. If you'd do it like this:

run out of hit points, fail a Fort save, then fail a second Fort save, that might work as well as "run out of hit points".

Or not. But its something to discuss based on "how do we want to handle 'what it takes to kill someone'", since we are doing this for, in your words, the memorable villains, I'd like to make sure we're happy with it for both PCs -and- their enemies.

So long as a PC isn't able to take on infinite lesser opponents ("Many" is fine, but we need a point that caps out) and all major characters (PCs and the NPCs who qualify) are at a reasonable level of risk, we are good.

That's a modified drunken walk problem in nineteen(?) dimensions, and it has been a long time since I've done one of those. Also, I would need statistics on the variance in bias of actual dice from ideal dice, and I don't have those handy. If you're interested, I suggest you read some papers on the drunken walk problem, gather statistics on the bias of mass-produced dice, and work it out yourself. I'm not interested, because I don't see how it's relevant to the thread.

Since we are rolling actual dice, the fact that mathmatically the +25 to save vs. DC 30 "ensures" something is not important unless the actual dice support that.

There is no perfect balance.
No one is claiming it exists.
Not one bloody person.

Then don't complain that the perfect dice will generate this outcome and we can rely on it with more certainy than the sun rising.

[Tshern]

Drunk walk problem? Care to give a very brief explanation?

[violence in the media]

First, for Elennsar's reference:

Statistics

Probability

Those might help.

Now, what you (Elennsar) seem to be saying is that because game rules might dictate that a PC will never lose in combat to a single Orc, that a PC will never lose in combat to 9000 Orcs. Right?

I assume we want to eventually reach a point where a battle against N orcs is winnable, or even an assured victory, but a battle against N+some additional quantity of orcs is not. How would you propose we do that? How do we fix that without resorting to complex mathematical formulas better left to a computer?

[Elennsar]

Assuming you use reasonable (but not good, defined as above average) tactics:

Odds vs. a single orc: 0%.
Odds vs. two orcs: ?
Odds vs. three orcs: ?
Continue until you reach when you want "multiple". Let's say ten orcs.
Odds vs. multiple orcs: ?
Odds vs. a lot of orcs: ?
Odds vs. a huge number of orcs: 50%? 70%?
Odds vs. a thousand orcs (or whatever): 1000%.

"you eventually are overrun and killed."

For instance:
1-9: 0%.
10-20: 20%
21-25: 30%
26-30: 40%
31-35: 50%
36+ etc.

Rough, mind you, and this is "all the orcs at once". Naturally, you should ensure that you have as far from that as possible. And it assumes that fighting lots of orcs will take energy and time and tire you out, so even if 30 orcs are no more capable than 20 orcs, you are less, which means they will have an easier time.

No computer necessary. But just as we have modifiers for flanking as is, being surrounded would be even worse, etc. And at some point, we'd declare that you lose if you try to stand and fight through the whole damn set. (Somewhere around one hundred orcs at once should do).

Naturally, if you play like Drizzt, you can easily avoid having the whole hundred orcs at you at once.

But if you face 300 orcs, you're going to have a much, much harder time of that. Eventually, either your modifiers are so low or the DC is so high that failure occurs unless you say "I run out of there", in which case being chased by over nine thousand orcs really isn't more threatening than by one (unless we like the idea that mobs give each other speed bonuses, but that's for a different style of campaign).

[Murtak]

There is no perfect balance.
No one is claiming it exists.
Not one bloody person.

Then don't complain that the perfect dice will generate this outcome and we can rely on it with more certainy than the sun rising.

No one is claiming that. But the chances of it occurring are damn high and unless you can disprove my (or Franks or Martins) math you might want to to start acknowledging that many small chances of death add up to near-certain-death very very fast.

Example: You "die" once (fail a save-or-die, run out of HPs, etc.) - except instead of dying you end up "beaten up", which makes you fail all attacks and non-harmless spells 50% of the time. Die again and you fail 90% of the time. Die again and you are unconscious.

How does that sound to you?

Make this work for the guys who can't hit your AC 40 because they only have +5, or anything else below +20 (so they can't eliminate your hit points) and we have something worth discussing.

I don't want PCs dying right and left. But I do want to have the fact that charging the dragon is capable of getting me killed.

-snip-

So long as a PC isn't able to take on infinite lesser opponents ("Many" is fine, but we need a point that caps out) and all major characters (PCs and the NPCs who qualify) are at a reasonable level of risk, we are good.

So all that's left is edge cases and picking a cutoff point you are comfortable with?

In DnD numerous foes can already hit near-arbitrary armor classes via aid another. Alternatively use one of these:
- a 20 always succeeds
- a 20 is a 30
- when you roll a 20, roll another die and add it. If you roll a 20 on that die, add it again (and so forth).
None of these are ideal solutions, but they probably work well enough for those few edge cases.

As for the cutoff point, that is a matter of playtesting and adjusting as needed.

[MartinHarper]

Drunk walk problem? Care to give a very brief explanation?

Sure. Essentially you have a drunk who is trying to walk somewhere, but he is drunk, so he's bad at it. In the simplest example, he is on a road outside a pub. Every time he takes a step, he has a 50% chance of taking a step North on the road, and a 50% chance of taking a step South. The maths shows that his mean distance from the pub after N steps is root N, and that as N tends to infinity, the number of times he returns to outside the pub also tends to infinity.

Having reflected further, Ellensar's question is actually a modified drunken walk in one dimension, not nineteen, because he's asking when a specific number will make up 5% of the results, not asking when all twenty numbers will take up 5% of the results each. That's much easier.

... the people who are screaming that a 0.5% chance of character death in every encounter equals a 100% chance in 200 encounters.

It doesn't, and nobody is screaming that. An independent 0.5% chance of character death in every encounter equals approximately a 63.30% chance of character death in 200 encounters.

edit: fixed quotes and maths

[Tshern]

Makes sense, you caught the interest of a drunken amateur mathematician. Thanks for the explanation.

[Elennsar]

No one is claiming that. But the chances of it occurring are damn high and unless you can disprove my (or Franks or Martins) math you might want to to start acknowledging that many small chances of death add up to near-certain-death very very fast.

I have a better idea. You start acknowledging that your math is based on perfect dice and your tolerance for risk is based on "I don't want ANY until the big boss" so the fact it adds up at all is horrifying you.

Or you're posting as if it does.

So all that's left is edge cases and picking a cutoff point you are comfortable with?

In DnD numerous foes can already hit near-arbitrary armor classes via aid another. Alternatively use one of these:
- a 20 always succeeds
- a 20 is a 30
- when you roll a 20, roll another die and add it. If you roll a 20 on that die, add it again (and so forth).
None of these are ideal solutions, but they probably work well enough for those few edge cases.

No, all that is left is making sure there is actual risk without having to set up stupid rules. If we want the possibility of a 30+modifiers, roll a d30. If 20+modifiers is not good enough, then we have a problem with the DC.

Auto success 5% of the time -really- doesn't make sense.

Rolling another die and adding it is also spending more time on dice rolling than is necessarily desirable (I would rather spend my finite paitence on the occasions that multiple rolls are necessary to represent an action).

What I want is the possibility that a group of orcs is capable of being threatening even if an individual orc is not, and that higher level (by "enough") =/= able to ignore that.

In D&D, the maximum bonus you could get (in melee) from add another would be +16 (8 people assisting). The max you can roll is 20. So you cannot ever hit an AC greater than 36+your attack modifiers.

Vs. AC 40 guy, that means on a 19 or 20 (somewhat worse/better if flanking is added), you will take a hit by an orc...which will do pitiful damage you don't care about.

Now if you had a chance of it being serious damage, that might be fine. But 2d4+6 is not serious at the level you have AC 40.

As for 0.5% chance adding up: Well, I forget what page I thought I saw it on, so consider my comment on people claiming it reaches 100% stricken until otherwise noted.

So here is the question. Assuming we want heroic fantasy...do our characters -have- 200 encounters?

D&D's experience system punishes people who have fewer encounters. That doesn't feel right for the genre, regardless of the chance of dying, because 200 fights would be a lot of fighting.

No point claiming we need to represent heroic fantasy's odds of survival without representing what the characters in it are doing.

[Murtak]

Fuck this. You are either to dumb/ignorant to recognize mathematical proof when you see it, or, and I think it is more likely at this point, you do not care for what other people say.

[Elennsar]

Mathmatical proof will be considered as absolutely 100% valid the instant the dice we are rolling match it.

And when they do match it, the 3s are on the "4+ or die".

Until then, whether or not your math is excellent, typoed, or not is not going to dictate my PC's death in three sessions or five or whatever.

[TarkisFlux]

Elennsar, I think this thread is dumb and wouldn't be posting here if you ignorance of probability didn't hurt my brain.

The math is based on non-biased dice. It then looks at all possible sets of rolls that you could have and gives back the set of those that match whatever criteria you want. Since you either don't know or don't understand that, here's a simple example. I'm going list a lot of the possible outcomes you could get from rolling a d4 3 times:

1 1 1
1 1 2
1 1 3
1 1 4
1 2 1
1 2 2
1 2 3
1 2 4
1 3 1
1 3 2
1 3 3
1 3 4
1 4 1
1 4 2
1 4 3
1 4 4
2 1 1
2 1 2
2 1 3
2 1 4
2 2 1
2 2 2
2 2 3
2 2 4
2 3 1
2 3 2
2 3 3
2 3 4
2 4 1
2 4 2
2 4 3
2 4 4
3 1 1
3 1 2
3 1 3
3 1 4
3 2 1
3 2 2
3 2 3
3 2 4
3 3 1
3 3 2
3 3 3
3 3 4
3 4 1
3 4 2
3 4 3
3 4 4
4 1 1
4 1 2
4 1 3
4 1 4
4 2 1
4 2 2
4 2 3
4 2 4
4 3 1
4 3 2
4 3 3
4 3 4
4 4 1
4 4 2
4 4 3
4 4 4

There are only 64 ways that can work out. Count them if you want, I didn't miss any. There are only 27 results in that set of all possible rolls that do not include any 1s. 27 of 64 possible results is about a 42.1875% chance of not rolling a one in any of 3 independent d4 rolls. It is certainly true that you only have a 75% chance of not rolling a 1 on a d4 on any individual roll, but the odds of not doing it over 3 rolls are significantly worse because each of the results are equally possible at the start.

The rest of the math that people are doing here is the same shit, but shorthand because no one wants to count individual state. They figure out the maximum number of possible results, pull out the subset that they care about, and draw inferences from it. No one here is claiming that you will fail one of those chances, just that the odds of it happening are shitty over the number of tries that you get.

[Username17]

I have a better idea.

I sincerely doubt that. But go on.

You start acknowledging that your math is based on perfect dice and your tolerance for risk is based on "I don't want ANY until the big boss" so the fact it adds up at all is horrifying you.

George, did you play a lot of football without a helmet?

What the fuck is wrong with you?

I can't even make a coherent argument against you here, because you have so much math fail going on that it's not even in the same universe as things that make any sense at all. Your arguments are like This Guy. There's a wall of text, but there's so much ignorance going on that I literally don't know where to start.

I think I might have to go back to definitions of words like "and" and "because." Possibly Employ Muppets.

-Username17

[MartinHarper]

You start acknowledging that your math is based on perfect dice.

Actually, the math is based on the probability of death in each encounter being constant and independant from previous encounters. It doesn't depend on using perfect dice. It doesn't depend on using dice at all, let alone perfect dice. The probability holds true in games based on cards, coins, or penis comparisons.

Evidently the probability of death in each encounter is going to be neither constant nor independent. Unfortunately, the probability of you constructing a rational argument based on those things is low.

----

In 4e, I'd estimate that a regular sized encounter has a 0.5% chance of defeat, doubling that leads to a 5% chance of defeat, and doubling that leads to a 50% chance of defeat. Of course that depends very much on the specific monsters and characters involved, but I feel it's a good rule of thumb. I don't think games need to have special rules to ensure that large groups of worthy adversaries slaughter the PCs while small groups are slaughtered by the PCs. It's going to be an emergent property of any sane combat system.

Assuming we want heroic fantasy...do our characters -have- 200 encounters?

For comparison purposes, Buffy has been through 144 episodes, and has died twice.

[Elennsar]

If you stand and fight, you risk dying vs. capable opposition. If you don't want to risk dying, don't stand and fight.


Howling that you are unable to have characters who make a mockery of the odds when the odds actually apply is at most relevant because you want to have protection via sequals and I don't.

I want to have characters who -can- survive, but not ones that will unless I choose for them to die.

And your stubborn insistance that probability will work with the dice we are using is not matched by the actual examples of dice being rolled.

Yes, it is likely that it will. It is also likely that I will not get sick today.

Evidently the probability of death in each encounter is going to be neither constant nor independent. Unfortunately, the probability of you constructing a rational argument based on those things is low.

My problem is not with your math. My problem is with you insisting that your math means that we cannot have character death even -threatened- more than very, very, very occasionally because you don't want to have to deal with it (character death).

I don't think games need to have special rules to ensure that large groups of worthy adversaries slaughter the PCs while small groups are slaughtered by the PCs. It's going to be an emergent property of any sane combat system.

The problem is, games do need rules to ensure that even if a dozen orcs are not a threat, a hundred (which Mooshie and Drizzt being faced with all at once, after their traps are used up, could very well kill them...and in fact Mooshie sets up the traps and calls for his allies because the amount they initially face will probably kill them, despite the fact both are much more capable than the orcs individually) are.

In D&D as written, two tenth level characters vs. a hundred ordinary orcs are not threatened unless the orcs use good tactics. In Salvatore's novel, Mooshie and Drizzt do need to use good tactics to ensure -they- aren't killed. (which they do and which doesn't happen)

For comparison purposes, Buffy has been through 144 episodes, and has died twice.

So in Buffy the Vampire Slayer: The RPG, we need to make it so Buffy dying is pretty damn unlikely (and at times not going to happen if Buffy does what it takes to qualify as "competent"), and that resurrection is an option.

However, it presumably is relevant whether she takes the steps necessary or not, as opposed to just relying on being high(er) level.

That's not supported unless a weak opponent -can kill her-.

George, did you play a lot of football without a helmet?

What the fuck is wrong with you?

That I refuse to accept that Drizzt being in danger at level 10 from being killed by level 1 orcs is a bad thing.

I don't mind your math. I mind your insistance that we have to ensure that he's unkillable by level 1 orcs because the math is certain to match the dice and the outcome is certain to be undesirable under any circumstances.

[Leress]

If you stand and fight, you risk dying vs. capable opposition. If you don't want to risk dying, don't stand and fight.

I said that already

Howling that you are unable to have characters who make a mockery of the odds when the odds actually apply is at most relevant because you want to have protection via sequals and I don't.

I want to have characters who -can- survive, but not ones that will unless I choose for them to die.

I want ones that would most likely survive since it means they are more likely to run into the fire than not.

And your stubborn insistance that probability will work with the dice we are using is not matched by the actual examples of dice being rolled.

Yes, it is likely that it will. It is also likely that I will not get sick today.

Then how would you figure out the percentage of survival and/or success then?

My problem is not with your math. My problem is with you insisting that your math means that we cannot have character death even -threatened- more than very, very, very occasionally because you don't want to have to deal with it (character death).

(He) didn't say that.

So in Buffy the Vampire Slayer: The RPG, we need to make it so Buffy dying is pretty damn unlikely (and at times not going to happen if Buffy does what it takes to qualify as "competent"), and that resurrection is an option.

Hello Drama Points

That I refuse to accept that Drizzt being in danger at level 10 from being killed by level 1 orcs is a bad thing.

Okay, I can agree with that, of course this depends on the number of Orcs (since there needs to be a way to make armies matter in Dnd)

I don't mind your math. I mind your insistance that we have to ensure that he's unkillable by level 1 orcs because the math is certain to match the dice and the outcome is certain to be undesirable under any circumstances.

He didn't say that.

[Elennsar]

"If Graul and Roddy had made their entrance then, the orcs, still numbering more than fifty, might have regrouped. With most of their traps exhausted, Drizzt and Montolio would have been in a sore position indeed!"

In D&D, they'd get...well, no experience, because the encounter is that unchallenging.

"But the orc king had seen another brewing problem to the north, and had decided, despite Roddy's protests, that the old man and the dark elf simply weren't worth the effort."

Not "could not be defeated". "would not worth the effort".

Seems like a reasonable enough challenge. Unfortunately, in D&D, two level 10s or so (Mooshie=Montolio...might be higher, might not, and I'm guessing on Drizzt here) would not regard it as worth the time it takes to roll dice over.

Then how would you figure out the percentage of survival and/or success then?

"I have a 15% chance of failing my roll, according to the math. Based on the fact I tend to roll 7 or more, I think I'm safe enough."

(He) didn't say that.

In those words? No. In effect? That's what those who want minimal chance of death have been saying over and over again.

Hello Drama Points

Fine as long as they don't lead to the situation quoted not being a problem. (Having them there to turn bad rolls into decent rolls at times is a very good thing, having them there to ensure bad rolls can't happen is not a good thing at all)

Okay, I can agree with that, of course this depends on the number of Orcs (since there needs to be a way to make armies matter in Dnd)

Roughly a hundred, plus allies. Mind, Mooshie and Drizzt also have allies, so say forty plus orcs apiece.

He didn't say that.

Again, those who have been shrieking about how even a 0.5% chance of dying in an encounter ensures character death is going to reach "likely" are basically saying that.

Ultimately, here's the problem.

If you want Drizzt to be lethally threatened by forty orcs, you have to make it possible for him to die - including with Drama Points in the system (though a Drama Point may make a given roll vs. death a success and that's fine, having them make you not need to make any is not).

If you want Drizzt to be brave to run into a burning building, it has to able to hurt him (not necessarily kill him, that might take more than just running and out kind of exposure, but hurt him badly enough that it would be reasonable to regard it as possible unpleasant consequences).

Whatever else it is for a Space Marine to run into a burning building to save a small child, it isn't brave. It wouldn't even require a courage check if Space Marines could feel fear.

[norms29]

Then how would you figure out the percentage of survival and/or success then?

"I have a 15% chance of failing my roll, according to the math. Based on the fact I tend to roll 7 or more, I think I'm safe enough."

so... we're back to you pleading "magic" (and by magic I mean defective) dice? you are really actively obstructing everyones attempts to workout acceptable odds based on your belief that the odds can be warded off by making sacrifices to the Random Number God? why the hell are you even here?

[Elennsar]

No, we're back to pointing out that any dice I have had the (mis)fortune to roll have not been perfectly balanced dice and therefore the "15% chance" is only approximate when I actually pick up the d20.

Yes, mathmatically, I have a 15% chance of dying. Assuming the dice were balanced properly when made and nothing has changed that.

But if you told me that drinking a deadly (deadly=capable of causing death, as opposed to say nausea) poison was only that likely to kill me (Elennsar, the person typing this), and that you had put that in my drink, I would think you were not trying very hard to kill me.

I'm not a particularly non risk adverse person, but I don't regard 15% as so likely that I would make my character avoid the activity (if they'd be willing to die any amount of the time over it).

Since the arguement seems to be that "acceptable odds" are "eliminate the chance because any odds we can conjure up will add up to odds we don't want", there's no point having any DC that can kill you. If it only happens one in five encounters, then you have 40 cases instead of 200 of 15% chance of death.

If you only want a 50-50 chance with Final End of Game Anyway encounters, you eliminate the scenario with Drizzt. Which is in the goddamn inspiration material.

[JonSetanta]

Safe? With luck in dice?
Pray to your dice god, zealot.

I take that chance, however small, as "Oh. There's a risk of death? Fuck that, I'm out of there."

[Elennsar]

If I felt safe in my dice cooperating with me and therefore I could face (PC) death as a possibility with a calm and serene mindset, I'd be a different person.

My reason has to do with my faith in something else.

As for your risk of death, however small: So, how do you cross the street?

There's a very small chance that someone might not notice (or might not care, or whatever) that you're crossing at the crosswalk with the light in your favor.

Odds of it happening are pretty poor, odds of it killing you are even poorer, but it is not impossible. So how do you IRL do it?

[JonSetanta]

As for your risk of death, however small: So, how do you cross the street?

There's a very small chance that someone might not notice (or might not care, or whatever) that you're crossing at the crosswalk with the light in your favor.

Odds of it happening are pretty poor, odds of it killing you are even poorer, but it is not impossible. So how do you IRL do it?

Real life? We were discussing RPGs. Dice. Roleplay.

In-game, if a DM decides you'll die crossing the street, you die.
If not, you say "I'm crossing the street" and they say "OK. You crossed safely."
If there's a CHANCE OF DYING without GUARANTEE they do that nifty dice rolling thing and you have a CHANCE of death. The second I learn that there's a 1-in-20 chance of becoming street pizza every time my character crosses a street, that's bad odds. I am immediately and possibly forever discouraged from walking on road, maybe pavement at all.
In fact, one might as well just stay home... but wait! Your own house cat might kill you for 1d3-4 damage. Easily.
Real life doesn't work like that.

Just.. nevermind. I'm done with this derail before I even began.