D&D General Normal Distribution Ability Scores

[Horwath]

That's 5e, though, where 14 actually gives you a bonus.

In 1e bonuses don't start until 15, meaning the only difference between a 7 Con and a 14 Con is your resurrection survival scores and, later, your roll-under odds if-when relevant.

honestly, we could have got rid of the scores completely and just use modifiers as scores.

0 - average
+5 - best of the best
-5 - barely functional

 

[Lanefan]

honestly, we could have got rid of the scores completely and just use modifiers as scores.

0 - average
+5 - best of the best
-5 - barely functional

A root problem I see with 3e-4e-5e design is that for some reason they took the 3-18 stats that were (ideally) generated on a bell curve and slapped a linear -4 to +4 bonus structure on top of them.

Bonuses and penalties should become greater as one nears the bell-curve extremes. The difference in bonus between 17 and 18, for example, should be greater than the diference in bonus between 11 and 12.

BX and 1e had this, if not completely right, a lot closer to it.

 

[Mannahnin]

A root problem I see with 3e-4e-5e design is that for some reason they took the 3-18 stats that were (ideally) generated on a bell curve and slapped a linear -4 to +4 bonus structure on top of them.

Bonuses and penalties should become greater as one nears the bell-curve extremes. The difference in bonus between 17 and 18, for example, should be greater than the diference in bonus between 11 and 12.

BX and 1e had this, if not completely right, a lot closer to it.

"Should" in what sense? You're assuming multiple premises there.

It's entirely relative to the result you're trying to achieve.

Those editions were deliberately designed for a smoother distribution of bonuses, allowing also for a consistent and predictable effect of bonuses granted by mechanics such as ability-increasing spells and magic items, and level-linked ASIs.

Those editions also don't use or assume a bell curve distribution of ability scores, which has been much more the exception than the rule across D&D's lifespan. OE, 2E*, and the Basic/Expert/RC/Etc. side D&D line were the ones to assume it. 1E, 3E and onward assume other ability score generation methods which don't give such a simple bell curve for PCs.

If you want the Strength spell, say, to give totally different benefits to one recipient than it does for another with the same roll, sure, the old AD&D bonus structure is a good way to achieve that. OTOH if you want +4 Strength to give relatively consistent benefits to anyone, then the post-2000 bonus structure makes a lot more sense.

*(and IME no one used 3d6 down the line in 2E)

 

[tetrasodium]

"Should" in what sense? You're assuming multiple premises there.

It's entirely relative to the result you're trying to achieve.

Those editions were deliberately designed for a smoother distribution of bonuses, allowing also for a consistent and predictable effect of bonuses granted by mechanics such as ability-increasing spells and magic items, and level-linked ASIs.

Those editions also don't use or assume a bell curve distribution of ability scores, which has been much more the exception than the rule across D&D's lifespan. OE, 2E*, and the Basic/Expert/RC/Etc. side D&D line were the ones to assume it. 1E, 3E and onward assume other ability score generation methods which don't give such a simple bell curve for PCs.

If you want the Strength spell, say, to give totally different benefits to one recipient than it does for another with the same roll, sure, the old AD&D bonus structure is a good way to achieve that. OTOH if you want +4 Strength to give relatively consistent benefits to anyone, then the post-2000 bonus structure makes a lot more sense.

*(and IME no one used 3d6 down the line in 2E)

Those editions(3.x/4e) were also designed so PCs were expected to acquire a churn of magic items boosting their attributes & other more specific things at regular intervals (ie +N skill/+N weapons/etc). It was a tradeoff, each point or two of attribute mattered more but the GM could put their thumb on the scale to (dis)favor one particular PC over a different PC just by how and in what order they awarded those magic items. Now in 5e that tradeoff was destroyed with a one-two-three-punch of each point or two mattering in the extreme but the math doesn't expect magic items★ and the magic items are no longer the sort of granular +2/+4/+N stuff present in those editions because they tend to be attrib=19 or just advantage without the granular options.

★at all! it's exceeded by the first before even getting to churn

 

[TwoSix]

A root problem I see with 3e-4e-5e design is that for some reason they took the 3-18 stats that were (ideally) generated on a bell curve and slapped a linear -4 to +4 bonus structure on top of them.

Bonuses and penalties should become greater as one nears the bell-curve extremes. The difference in bonus between 17 and 18, for example, should be greater than the diference in bonus between 11 and 12.

BX and 1e had this, if not completely right, a lot closer to it.

My favorite bonus progression is Worlds Without Number.

8-13 = 0
14-17 = +1
18 = +2
4-7 = -1
3 = -2

The character creation method is "Roll 3d6 in order; then set one number of your choice to 14."

 

[ezo]

BX and 1e had this, if not completely right, a lot closer to it.

B/X, IMO, had it best. Low easy bonus for 13-15, etc. but in reality it isn't much different 5E's

12? +0 vs. +1
13? +1 vs. +1
14? +1 vs. +2
15? +1 vs. +2
16? +2 vs. +3
17? +2 vs. +3
18? +3 vs. +4
Tot: +10 vs. +14
Avg. +1.42 vs. +2

So, on average barely a half bonus more.

My favorite bonus progression is Worlds Without Number.

8-13 = 0
14-17 = +1
18 = +2
4-7 = -1
3 = -2

The character creation method is "Roll 3d6 in order; then set one number of your choice to 14."

That is a nice one in some ways.

In other ways, my favorite would be:
9-12 +0
13-15 +1
16-17 +3
18 +6.



I hate that any sort of a contested roll on d20 for a STR 9 vs STR 18 has the STR 9 winning more than 25% of the time in 5E. That is just nonsensical and ridiculous IMO.

 

[leozg]

Allow any CON save to use STR
Allow any DEX save to use INT
Allow any WIS save to use CHA

That is the easies way I can think of.

Allow any CON/STR save to use either.
Allow any INT/DEX save to use either.
Allow any WIS/CHA save to use either.

 

[ezo]

Allow any CON/STR save to use either.
Allow any INT/DEX save to use either.
Allow any WIS/CHA save to use either.

This would make it worse.

 

[Lanefan]

"Should" in what sense? You're assuming multiple premises there.

It's entirely relative to the result you're trying to achieve.

Those editions were deliberately designed for a smoother distribution of bonuses, allowing also for a consistent and predictable effect of bonuses granted by mechanics such as ability-increasing spells and magic items, and level-linked ASIs.

Those editions also don't use or assume a bell curve distribution of ability scores, which has been much more the exception than the rule across D&D's lifespan. OE, 2E*, and the Basic/Expert/RC/Etc. side D&D line were the ones to assume it. 1E, 3E and onward assume other ability score generation methods which don't give such a simple bell curve for PCs.

4d6k3 or 5d6k3 or any other rolling system still puts the stats on a bell curve. That the bell curve is skewed a bit toward the high end doesn't matter, it's still a mathematically-predictable curve with a bulge in the middle and tails at each end.

If you want the Strength spell, say, to give totally different benefits to one recipient than it does for another with the same roll, sure, the old AD&D bonus structure is a good way to achieve that. OTOH if you want +4 Strength to give relatively consistent benefits to anyone, then the post-2000 bonus structure makes a lot more sense.

You're assuming the Strength spell gives a flat +4, which sounds rather dull. In 1e it gave +d4 to +d8 depending on the recipient's class, meaning you often didn't know just how useful it was going to be.

 

[Lanefan]

In other ways, my favorite would be:
9-12 +0
13-15 +1
16-17 +3
18 +6.

I'd quickly smooth that out to

9-12 +0
13-15 +1
16 +2
17 +4
18 +6

And would then ask myself how (or if!) it would work with stats beyond 18, which while uncommon are certainly achievable in BX or 1e; never mind percentile Strength which is, as usual, its own can of worms.