2e to 3e ability score changes were a bad decision.

[Psychic Robot]

Not that 2e's versions were perfect, of course, but turning every +2 to a stat to a +1 bonus was a poor design choice that is now a permanent part of the game. Even if the designers had decided to change it so that every +3 equated to a +1, the RNG would be in significantly better condition.

[icyshadowlord]

...why does this remind me of the nerf Pathfinder gave to feats?

"No, your Improved Trip only gives a +2, not a +4. Now go out there and gather the damage, you meat-shield."

Anyway, are you suggesting a solution to this problem you mentioned, or are you just pointing this out for the hell of it?

Edits, because edits.

[Psychic Robot]

just pointing it out

[Lokathor]

If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

[hogarth]

If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

Right; all you care about is the bonus. Concealing that by requiring you to divide by 3 instead of by 2 is just ridiculously pointless.

[DrPraetor]

If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

Well, no, because "normal" stats are in the 3-18 range.

So, there's a difference between:

12-13	+1	21%
14-15	+2	11%
16-17	+3	4%
18-19	+4	0.5%

and

13-15	+1	21%
16-18	+2	5%
19-22	+3	0%

and what was actually in 2nd edition, which varied by stat but tended to be:

15	+1	5%
16	+2	3%
17	+3	1%
18	+4	0.5%

So in D&D 3rd, there are noticeable differences in the general population, while in D&D 2nd, only 5% of people have any positive modifiers at all (and another 5% have penalties.) I'm not sure why you think that makes the RNG "better off", but okay.

[Username17]

I rather like 3e's bonus spells paradigm. Each incremental bonus adds the same +1 to a new thing and only when you get to the end of the list does it go back and start becoming a +2. With enough things in each category, you could differentiate between a lot of different attribute numbers without actually breaking the RNG.

-Username17

[RobbyPants]

How tight do you think the RNG should be for a d20 on any given roll for any given level? 10 points?

[hogarth]

What I do not want to see is a system so stingy with bonuses that you end up with garbage like in 4E, where you're totally supposed to get excited about a +2 bonus that lasts for one turn, even though 90% of the time it doesn't do anything at all.

[RobbyPants]

What I do not want to see is a system so stingy with bonuses that you end up with garbage like in 4E, where you're totally supposed to get excited about a +2 bonus that lasts for one turn, even though 90% of the time it doesn't do anything at all.

What are you looking for, exactly?

[hogarth]

What I do not want to see is a system so stingy with bonuses that you end up with garbage like in 4E, where you're totally supposed to get excited about a +2 bonus that lasts for one turn, even though 90% of the time it doesn't do anything at all.

What are you looking for, exactly?

For D&D? Frankly, I don't have a huge problem with 3E, although I might cut down on the sheer variety of different bonuses.

On the other hand, I think you could do a fine version of D&D without any attributes at all.

[echoVanguard]

I rather like 3e's bonus spells paradigm. Each incremental bonus adds the same +1 to a new thing and only when you get to the end of the list does it go back and start becoming a +2. With enough things in each category, you could differentiate between a lot of different attribute numbers without actually breaking the RNG.

-Username17

The problem was generally that while this mechanic works well if you have enough categories, it also requires that you consult the chart every time you want to configure your bonuses (usually only once per level, but every time you have to make an ad-hoc calculation).

[quote="Lokathor]If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.[/quote]
I don't think this is true.

echo

[hogarth]

[quote="Lokathor]If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

I don't think this is true.[/quote]
You think it's sheer coincidence that, in 3E, belts of giant strength come in flavours of +2, +4 and +6, but not +1, +3 and +5?

On a side note, the "basic" version of Pathfinder is apparently going with the "modifier as ability score" system, instead of (ability score - 10)/2.

[echoVanguard]

You think it's sheer coincidence that, in 3E, belts of giant strength come in flavours of +2, +4 and +6, but not +1, +3 and +5?

No, but I also don't think that's the only way to distribute stat upgrades.

echo

[RadiantPhoenix]

How tight do you think the RNG should be for a d20 on any given roll for any given level? 10 points?

I believe there was some mention of 5 points somewhere, but that was only for one character's bonus.

e.g.: if sneaking is a 2n skill, a 5th level character trained in sneaking might have anywhere from a +10 to a +15 bonus to sneak checks under 'normal' conditions (i.e.: expected level 5 equipment, standard environment), which would be opposed by the opponent's perception (which also needs to be 2n), which could be anywhere from 20 to 25 if you wanted to keep a default 55% success rate, which goes from 30% to 80%.

There would probably be circumstances that could modify it beyond that, but those would be unusual.

[RobbyPants]

How tight do you think the RNG should be for a d20 on any given roll for any given level? 10 points?

I believe there was some mention of 5 points somewhere, but that was only for one character's bonus.

e.g.: if sneaking is a 2n skill, a 5th level character trained in sneaking might have anywhere from a +10 to a +15 bonus to sneak checks under 'normal' conditions (i.e.: expected level 5 equipment, standard environment), which would be opposed by the opponent's perception (which also needs to be 2n), which could be anywhere from 20 to 25 if you wanted to keep a default 55% success rate, which goes from 30% to 80%.

There would probably be circumstances that could modify it beyond that, but those would be unusual.

Five points on each side for a total swing of 10? That makes sense. Although, you'd have to do a lot of pruning to get all the possible modifiers down that low.

[RadiantPhoenix]

Five points on each side for a total swing of 10? That makes sense. Although, you'd have to do a lot of pruning to get all the possible modifiers down that low.

Alternatively, you could require people to take a certain number of the bonus items and prohibit taking more than a certain number of penalties.

[fectin]

The problem was generally that while this mechanic works well if you have enough categories, it also requires that you consult the chart every time you want to configure your bonuses (usually only once per level, but every time you have to make an ad-hoc calculation).

Floor((ability-2*(spell level + 1))/8) or ceiling((1+bonus-level)/4)

And you build it into an automagic character sheet anyway.

[Stubbazubba]

The problem was generally that while this mechanic works well if you have enough categories, it also requires that you consult the chart every time you want to configure your bonuses (usually only once per level, but every time you have to make an ad-hoc calculation).

Shh! Careful, or shadzar will smell the blood in the water and talk about how 'back in the good ol' days' looking things up on a chart was how you played the game, and getting away from that is not D&D.

[Swordslinger]

If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

Yes. The 3-18 stat system needs to die anyway, and just get replaced with something similar to Mutants and Masterminds 3E where your stat is your bonus. There's no point at all having odd ability scores that do nothing. The only reason it's stayed there is because they were afraid to kill a sacred cow.

[Josh_Kablack]

PR, this isn't your usual trolling, instead you are talking like a certifiable loony here - has shadzar or condorDM or darwinism hijacked your account?

Lemme refresh your memory:

In 2e an 8 Strength and a 15 Strength only differ by encumberance numbers and changing BBLG from 1% to 7% (both of which are too low to matter)

In 2e a 7 Dex and a 14 Dex are +0 to everything

In 2e a 7 Con and a 14 Con only differ by System Shock and Rez survival rolls

In 2e, a 1 int and an 18 int only differ in languages known for anybody aside from wizards. Even for wizards, the max spell level for a 9 Int is higher than the max spell level seen in 85% of actual games, so in only determines number of spells in spellbook and what numbers you have to fudge the percentile roll too for groups that actually used it.

In 2e, there is zero difference between an 8 and a 14 Wisdom for anybody aside from Clerics.

In 2e, a 9 Cha and a 12 Cha gave everyone a +0 reaction adjustment (although 12 got one more henchmen, I don't think there was ever a game where henchmen were anything other than offscreen help or fireball fodder.

Like most all things 2e, that's ass for no fewer than 7 different reasons

1. Since the stat method is a bell curve, most characters will end up in the middle range, where all the modifiers are +0 to anything. Since most characters will end up here, this means that stats are not a reliable way of differentiating one PC from another.
2. Since most characters will end up with +0 to everything, this means that it's uncommon for player to roll up a character with exciting stats. Personally, I think player excitement is a good thing, but it's obvious that 2e defenders don't.
3. While the range of meaningless stats is always in the middle of the bell curve, it sits at different points along the bell curve for no discernable reason. This means that players have to memorize that a 15 dex matters but a 15 strength doesn't and they can safely sink their Con to 7 but don't want to sink their Cha below 9 and similar inconsistant breakpoints 3e's "innovation" of putting all stats on the same scale was a huge improvement in Clarity, Comprehensibility & Coherence. The further "innovation" of tying all numeric bonuses to a linear formula means that a player knows their modifier is always ( stat-10 ) / 2 rounding halves down, their max spell level is always stat-10 and they just have to look at the chart for bonus spells - thereby further increasing Clarity, Comprehensibility and Coherence and also likely improving Speed of Play.
4. In 2e, some stats outright did not matter at all for some characters.
5. The original version optional 2e NWP system that tried to make the raw stats matter was all-but-incapable of generating a character who was even competent at their specialty - so either players accepted their characters failing at mundane tasks within their competencies or it was magic tea party that didn't use those rules. A fucking 10th level fire wizard with Max intelliegence who burnt all his slots on Blacksmithing got to roll a d20 for less than Str+6, with the example (page 55) saying that making a wrought iron-cage is an arbitrary -3 to that roll. So if he has a 12 or less Str, then he fails 25% of the time - even though he can fucking cast Wall of Fucking Iron twice a fucking day with zero chance of failure. While that is a skill keyed to an attribute other than primary, blacksmithing isn't even cross-class, multi-slot or bundled with an inherent penalty like most of the NWPs characters are supposed to care about. Blacksmithing also doesn't let characters make swords or weapons - so it is even less likely that a 10th level PC would sink all of their slots into it than it is that you'll ever see a 10th level game in 2e.
6. Okay, that's not actually 7, but if you're defending 2e, you think THAC0 is easier than BAB, so you can't count number of steps anyways and are incapable of noticing.

Now if you want to rephrase you argument to say that "3e's decision to set the linear scaling of stat bonuses by dividing by 2 instead of a larger constant was a poor design choice" - you'd still be wrong, but at least you wouldn't be crazy and you do make a reasonable point about the chance of stat bonuses pushing things off the RNG.

There is exactly one optimum design choice for that constant, and that is to divide by 1 - which is to say don't divide at all and use the original stats.


Because if you only use derived stats, then the original stats don't matter and you have added needless complexity to chargen. If you only use original stats, then definitionally you are not using derived stats. And if you use a mix of both original and derived stats then there are two possibilities:
Either the tasks which use original attributes and the tasks which use derived attributes have greatly varying difficulties when a character rolls against the the same target number - which is potentially confusing and potentially unbalancing.
Or the difficulties between such tasks will be largely similar and a task with target number N will be about the same difficulty whether it's resolved with direct attribute rolls or with some derived number - and in that case you have a system that workks when difficulties are similar, so there is no obvious reason why it can't work when they are exactly the same - which is to say you can get the exact same results without calculating derived values at all and having simpler chargen.

[Psychic Robot]

If it took 3 stat points to get +1 mod then stats would be handed out in 3s instead of 2s and nothing would change.

I meant that if the current system was kept the same and instead 13 was +1, 16 +2, and so on, the system would be better

Now if you want to rephrase you argument to say that "3e's decision to set the linear scaling of stat bonuses by dividing by 2 instead of a larger constant was a poor design choice" - you'd still be wrong, but at least you wouldn't be crazy and you do make a reasonable point about the chance of stat bonuses pushing things off the RNG.

that's what I meant

[Emerald]

PR, this isn't your usual trolling, instead you are talking like a certifiable loony here - has shadzar or condorDM or darwinism hijacked your account?

Lemme refresh your memory:

In 2e an 8 Strength and a 15 Strength only differ by encumberance numbers and changing BBLG from 1% to 7% (both of which are too low to matter)

In 2e a 7 Dex and a 14 Dex are +0 to everything

In 2e a 7 Con and a 14 Con only differ by System Shock and Rez survival rolls

In 2e, a 1 int and an 18 int only differ in languages known for anybody aside from wizards. Even for wizards, the max spell level for a 9 Int is higher than the max spell level seen in 85% of actual games, so in only determines number of spells in spellbook and what numbers you have to fudge the percentile roll too for groups that actually used it.

In 2e, there is zero difference between an 8 and a 14 Wisdom for anybody aside from Clerics.

In 2e, a 9 Cha and a 12 Cha gave everyone a +0 reaction adjustment (although 12 got one more henchmen, I don't think there was ever a game where henchmen were anything other than offscreen help or fireball fodder.

[...]

1. Since the stat method is a bell curve, most characters will end up in the middle range, where all the modifiers are +0 to anything. Since most characters will end up here, this means that stats are not a reliable way of differentiating one PC from another.
2. Since most characters will end up with +0 to everything, this means that it's uncommon for player to roll up a character with exciting stats. Personally, I think player excitement is a good thing, but it's obvious that 2e defenders don't.
3. While the range of meaningless stats is always in the middle of the bell curve, it sits at different points along the bell curve for no discernable reason. This means that players have to memorize that a 15 dex matters but a 15 strength doesn't and they can safely sink their Con to 7 but don't want to sink their Cha below 9 and similar inconsistant breakpoints 3e's "innovation" of putting all stats on the same scale was a huge improvement in Clarity, Comprehensibility & Coherence. The further "innovation" of tying all numeric bonuses to a linear formula means that a player knows their modifier is always ( stat-10 ) / 2 rounding halves down, their max spell level is always stat-10 and they just have to look at the chart for bonus spells - thereby further increasing Clarity, Comprehensibility and Coherence and also likely improving Speed of Play.

Assuming you get rid of the variations in the stat modifiers so every ability has 15 is +1, 16 is +2, 6 is -1, 5 is -2, and so forth, is there any merit at all to the 2e-esque ability system? In 3e you have many more ways to customize a character than ability scores, so you don't really need a, say, 14 Str to be vastly different from a 10 Str to make two fighter-types different. The range would be standardized, so you wouldn't have that confusion.

[CatharzGodfoot]

Assuming you get rid of the variations in the stat modifiers so every ability has 15 is +1, 16 is +2, 6 is -1, 5 is -2, and so forth, is there any merit at all to the 2e-esque ability system?

No.

[Josh_Kablack]

Assuming you get rid of the variations in the stat modifiers so every ability has 15 is +1, 16 is +2, 6 is -1, 5 is -2, and so forth, is there any merit at all to the 2e-esque ability system?

You'd likely be better off with an ArM style attribute system of -5 to +5, or whatever your mods actually were. Maybe they shoulda brought one of the early ArM devs in on D&D.

But if you are wedded to the D&D concepts of 3-18 stats modifying a d20 roll then having the modifiers only kick in at the far ends and each point at the far ends mean more than 2-3 points in the middle is a way to reinforce the bell curve - most characters will be mediocre and stat any bonus at all is noteworthy and memorable. Having a medium or large bonus is rare and special. I'm plenty mediocre in real life, so I personally regard that sort of thing as a failure in my escapist fantasy, but as near as I can figure, the attitude of a lot of pre-3e grognards is "that's a feature not a bug".

But even for that sort of thing, I kinda prefer the old edition Rolemaster's percentile stats to 3d6. At least in a percentile system where you have a linear chance to roll any stat, but only a 10% chance to get a stat in the bonus range, and individual points only matter if you roll in the top 5%, the odds are more obvious to players than having to compute permutations out of 216 possibilities. Maybe they shoulda brought one of the early Rolemaster devs in on D&D

Even if you're not rolling for stats, you can achieve similar bell-curve like results by just using a point-buy system where additional points in higher stats cost more than the same number of additional points in lower stats. Such as the default point-buy in 3rd ed or the triangular attribute costs used in Storyteller.