Quote from: Cat the Bounty Smuggler on February 16, 2022, 02:37:21 AM
Sure, but note that your approach makes it binary. Berta's Cha 12 could also mean "He's suspicious, but something about Berta makes him want to trust her. He hasn't passed judgment yet." So she has a chance, but she'll have to sell it harder, taking longer.

That is just a matter of granularity. Attributes could still be quantified beyond "bad, normal, good". It's relatively simple: A range of X (=15 for attributes from 3 to 18) is only useful for rulings if you want X possible outcomes. I mean, what if she had Cha 11? Or Cha 13? In my experience, that level of granularity usually doesn't apply, and even a judgment like "she has to try a little harder" makes little difference to "it works" in practice.

I rarely see a need for more than 3-5 outcomes for an in-game task, and mostly just two (failed and succeeded) suffice. According to my own experience, even in systems that actually codify something like "it takes longer" into the system (MegaTraveller's task system, for example), this factor is usually disregarded because it's not important.

Modern D20-based system try to make the relatively miniscule bonuses (in 5E more than ever) seem important and manage to get away with this illusion because of people's desire to get every advantage they can and thus overvalue small numerical increases. But it is a simple matter of fact that for 90% of all rolls, the difference between Cha 10 (+0) and Cha 15 (+2) is irrelevant. It is just obscured by having a lot of possible dice roll outcomes, even if they translate into just two situational outcomes.

Quote from: Wulfhelm on February 16, 2022, 05:28:26 AM
Modern D20-based system try to make the relatively miniscule bonuses (in 5E more than ever) seem important and manage to get away with this illusion because of people's desire to get every advantage they can and thus overvalue small numerical increases. But it is a simple matter of fact that for 90% of all rolls, the difference between Cha 10 (+0) and Cha 15 (+2) is irrelevant. It is just obscured by having a lot of possible dice roll outcomes, even if they translate into just two situational outcomes.

Depends on the range. If you have a 15% chance of failure, a +2 can reduce that to a 5% chance of failure. A 10% difference can triple your odds of survival. This comes up more than you're suggesting, because the difficulties of various rolls aren't evenly spread across the spectrum. Instead, they tend to be clustered. Positive actions like hitting tend to be more likely than not, while rolls to avoid negative effects can be quite unlikely.

In my system characters get -5 to +5 in their talents. While this naturally are things like Strength and Fitness, it also includes things like Wealth and Status (higher wealth gets higher payout on missions when hired for jobs. Using average wealth for a party of adventurers.)

negative numbers mean the characters are disadvantaged while positive advantaged towards the average human of your own sex.

If a talent is 0 (average,) it is simply not recorded on the sheet.

Also matters in which attributes/modifiers are commonly used and which ones are outliers but still meaningfully occur. 

If the raw attributes go from 3 to 18, but every character starts with a minimum of 8, then the 3-7 range isn't meaningful for a player.  It might as well be handled by some kind of "weakness" tag on the monsters.  NPCs are, of course, a vast gray area.  If you want to build in that the frail old sage really can't climb the ladder very fast and might fall, then it is back to how often that arises.  Some games more than others.

Naturally, it also depends on the math of the underlying system.  How well do the attributes/modifiers fit seamlessly into it so that players and even the GM can start to just use them without having to think about it consciously every time?  Which is to say that they should be chosen to fit, not copied from a prior version or another game and then hammered squarely into a round hole.  There's a certain amount of reverse engineering going on with attributes that fit a system well.

Personally, I really appreciate the curve and range of the early D&D -3 to +3, with the relatively rare modifiers and its natural bell curve outliers.  I do like a bit more spread, though.  I find the effective -1 to +5 of later D&D to be deficient in matching the system (in some ways, not all), and it makes the -1 this weird bit in the system.  Plus, those modifiers are on a very different curve that is not to my taste (though OK in its own context).  I also like some room for growth on attributes as characters advance in power, not least because it can make rolling for attributes more viable.

That's why I built my own system to go from -5 to +N, with the same math for frail NPCs and monsters, but with characters sitting in the -4 to +4 range, and the -4, -3, and +4 occupying the same rare space that the early D&D -3 and +3 occupy.  The attributes ranging from 1 to 20+ gives room for simple attribute advancement rules that fits the curve of attributes I wanted to get--i.e. reverse engineered exactly to do that. 

Later D&D "simplifying" the modifiers to +1 every 2 points of attributes is exactly a case of "Chesterton's Fence":  Tearing something down without understanding its purpose.  Though in this case, they've kept the frame of the fence for tradition and removed its meaning.  So yes, would have been better to have dropped the attributes altogether given that first, bad step. 

Quote from: Fighterboy on February 15, 2022, 03:33:56 PM
If an attribute is in the average range, why bother even recording it?

I've often thought that a player should choose a handful of attributes from a longer list than the standard 6. Those are either above or below average, conferring appropriate bonuses, and nothing else matters. May be more flavourful for a character to have bonuses/penalties in strength, dexterity, courage, moral fortitude and leadership?

From an OSR perspective it shouldn't be too hard to implement?

Nuance. Both a 9 and an 11 are average, but the 11 is greater than the 9. There will be no mechanical bonus advantage, but it can provide a better handle in role-playing the character.

Quote from: Pat on February 16, 2022, 05:58:46 AM
Depends on the range. If you have a 15% chance of failure, a +2 can reduce that to a 5% chance of failure. A 10% difference can triple your odds of survival.

In 90% of all actual cases, that won't matter. That's just a mathematical certainty. Your +2 only matters if you roll a 2 or 3. For the other 18 possible rolls, it is irrelevant.
(And of course, for a D20, that is already the extreme end of the spectrum even if you take relative chances of success as a yardstick instead.)

And also of course, if you put the life or death of a character in the hands of a single D20 roll, you want every little improvement you can get. But that doesn't mean that it's any less random. A character with a +2 advantage can not reasonably hope to regularly survive save-or-die-dangers that a character without that bonus wouldn't. (Which is why I'd argue that save-or-die is probably not very good design.)

The relative ranges of die roll results and bonuses are simply such in D20 systems that random chance is much more relevant for success than character bonuses. That is a tangent to the original discussion, but it is nonetheless reality.

  For OSR purposes, a 10 strength was average.  A 15 gave no mechanical advantage regarding combat or bonuses. 
 
   But do you have any idea how much stronger you look military pressing 150 beside a person who can only do 100?   Come on now!!

Depends on the system... If you use roll d20 under attribute, obviously every point counts.

And players often like these small details about their PCs.

For NPCs, however, its +0s all over, unless that particular individual deserves a distinction.

Quote from: Wulfhelm on February 16, 2022, 05:28:26 AM
That is just a matter of granularity. Attributes could still be quantified beyond "bad, normal, good". It's relatively simple: A range of X (=15 for attributes from 3 to 18) is only useful for rulings if you want X possible outcomes. I mean, what if she had Cha 11? Or Cha 13? In my experience, that level of granularity usually doesn't apply, and even a judgment like "she has to try a little harder" makes little difference to "it works" in practice.

The ranges don't have to be the same for every situation, though. This NPC requires a Cha 14 to bluff, that one is more gullible and only needs an 11, and that one over there is a master spy who won't fall for anything less than Cha 17. And as @Eric Diaz said, with a roll-under or even a modified d% task, every point counts.

It's ultimately a matter of taste, of course. My point is that the old style was to use anything and everything at your disposal to make rulings, and the full 3-18 range is one of the things that can be used.

Quote from: Eric Diaz on February 16, 2022, 10:46:35 AM
Depends on the system... If you use roll d20 under attribute, obviously every point counts.

And players often like these small details about their PCs.

For NPCs, however, its +0s all over, unless that particular individual deserves a distinction.

Even better: just stat NPCs as monsters. If something is going to be relevant outside of combat, you can note it in the description.

---

Oh, and I thought of another reason: 0e and B/X allow you to move points around at character creation. So at least there, the actual scores matter.

Quote from: Wulfhelm on February 16, 2022, 09:06:20 AM

Quote from: Pat on February 16, 2022, 05:58:46 AM
Depends on the range. If you have a 15% chance of failure, a +2 can reduce that to a 5% chance of failure. A 10% difference can triple your odds of survival.

In 90% of all actual cases, that won't matter. That's just a mathematical certainty. Your +2 only matters if you roll a 2 or 3. For the other 18 possible rolls, it is irrelevant.
(And of course, for a D20, that is already the extreme end of the spectrum even if you take relative chances of success as a yardstick instead.)

And also of course, if you put the life or death of a character in the hands of a single D20 roll, you want every little improvement you can get. But that doesn't mean that it's any less random. A character with a +2 advantage can not reasonably hope to regularly survive save-or-die-dangers that a character without that bonus wouldn't. (Which is why I'd argue that save-or-die is probably not very good design.)

The relative ranges of die roll results and bonuses are simply such in D20 systems that random chance is much more relevant for success than character bonuses. That is a tangent to the original discussion, but it is nonetheless reality.

Not how it works. If you only fail on a 1-3 on a 20, and a +2 means that you only fail on a 1 in 20, that divides the failure rate by a factor of 3. And if that's a save or die, over the course of a campaign, lacking that +2 means you're three times more likely to die. It doesn't matter what you roll the other 95% of the time, that remains the overall effect.

Quote from: Pat on February 16, 2022, 01:18:08 PMNot how it works.

Yes how it works. For 90% of all actual rolls, a +2 bonus didn't matter. That is a simple fact. And to be sure, something like "your chance is three times as high" is often deceptive and only psychologically relevant IRL too: If your chance to develop certain types of cancer is three times as high when living near Fukushima Daiichi (hypothetical! It isn't!) then that might scare some people, but not rational people who realize that even "three times as high" is still "extremely low."

In event, according to my experience the perception of players if you spell out to them either "you have an 85% chance of failure" or "you have a 95% chance of failure" is going to be pretty much the same.

Or to put it differently: If, in game terms, you had either a 0.01% chance of dying or a 0.05% chance of dying, you'd naturally opt for the former if you had free choice, but your expectation would in both cases be "Nah, I'm gonna be fine." Even if IRL, this would be perceived to be massive differences on which to hinge the functioning of entire societies.

With the range of bonuses and die rolls being what it is in D&D derivatives, luck of the dice is way, way more important than character abilities for nearly all situations, and especially for such "save or die" type situations. That is a plain fact. To be honest it's going to be hard to discuss this with someone who can't accept that.

QuoteAnd if that's a save or die, over the course of a campaign,

If it happens only once or rarely, then it was overwhelmingly just a matter of luck. And if it happens regularly, then... it's still a matter of luck.

Quote from: Wulfhelm on February 16, 2022, 02:15:13 PM

Quote from: Pat on February 16, 2022, 01:18:08 PMNot how it works.

Yes how it works. For 90% of all actual rolls, a +2 bonus didn't matter. That is a simple fact. And to be sure, something like "your chance is three times as high" is often deceptive and only psychologically relevant IRL too: If your chance to develop certain types of cancer is three times as high when living near Fukushima Daiichi (hypothetical! It isn't!) then that might scare some people, but not rational people who realize that even "three times as high" is still "extremely low."

No, it's not how it works, and your choice of an example shows you're completely missing the concept. The chance of an individual getting cancer from Fukushima is very low, and because it's measured over a lifetime, it has a tiny marginal effect on the individual's expected lifespan. In extreme cases, it might shave off a few months. But a PC in a RPG that uses save or dies will have to make save or dies checks over the course of their career. If the chance of failing goes from 5% to 15%, then the character's expected lifespan will drop by a factor of 3. If a character normally lives for 60 sessions, they'll now have an expected lifespan of 20 sessions. That's a huge difference.

Ars Magica uses only integers for stats.  That stat is the modifier.  If you have no modifier then the stat is zero.  When I first saw it I thought, "of course!"  Unfortunately for D&D we're stuck in a weird tradition of tables and charts from Gary's time as an insurance clerk in Chicago.  The way I use "zero" stats in D&D is whenever there's a tie in a contest the higher stat wins.

Quote from: AtomicPope on February 16, 2022, 07:05:27 PM
Ars Magica uses only integers for stats.  That stat is the modifier.  If you have no modifier then the stat is zero.  When I first saw it I thought, "of course!"  Unfortunately for D&D we're stuck in a weird tradition of tables and charts from Gary's time as an insurance clerk in Chicago.  The way I use "zero" stats in D&D is whenever there's a tie in a contest the higher stat wins.

One argument against 0-centered stats is psychological. Shifting a -3 to +3 scale up 4 points to a 1 to 7 scale makes some people happier, because a 4 sounds better than a 0, even if the mechanics are perfectly equivalent in all other ways. This seems to be factor with 0s and negatives, and can lead to stat inflation because there's pressure to avoid the lower half of the scale.

Quote from: Fighterboy on February 15, 2022, 03:33:56 PM
If an attribute is in the average range, why bother even recording it?

I've often thought that a player should choose a handful of attributes from a longer list than the standard 6. Those are either above or below average, conferring appropriate bonuses, and nothing else matters. May be more flavourful for a character to have bonuses/penalties in strength, dexterity, courage, moral fortitude and leadership?

From an OSR perspective it shouldn't be too hard to implement?

For the same reason we have a zero in the number system.