Ability Scores

[Li Shenron]

I think that linearity is a better math solution than the old "curves" for ability bonuses. It is more streamlined, easier to remember, allows for easier design of challenges and circumstances modifiers (including equipment).

To me your concerns about the distribution of ability scores/bonuses over the population can still be solved by adopting an alternative way of rolling for ability scores, without the need to also go back to a non-linear definition of ability bonuses.

All that matters after all is how exceptional is a +3 bonus (rather than how exceptional is a 16) if the bonus will be used in many more mechanics than the score itself. So if you want to "compress" the distribution more around the average, you can do it by changing the rolling method to something with a lower variance.

 

[Rogue Agent]

In 2nd Edition, the same difference in ability scores only creates a difference of 5% chance of success.

10% on hit probability, damage, missile attacks

16% on bend bars/lift gates.

20% on physical saving throws

30% on mental saving throws and chance of spell failure

40% on resurrection survival

44% on system shock

Sorry. I lost track of what we were talking about.

 

[Chris_Nightwing]

I want the translation from ability score to its effects on play to be straightforward. Before 3E, whilst it was fun to look through the tables, it wasn't straightforward. I don't think there was a flaw with the 3E ability score/bonus system itself, but there was with the surrounding ways of increasing scores and precisely what depended on them.

A system that had a large middle range of +0, with +-1/+-2 at the extremes seems friendly on paper, and it is simple. It doesn't prevent the player desire to want more though. Mongoose Traveller (which is great fun) for instance, has a similar system, and during character creation players will continually try to push their abilities up to get that extra +1, because it applies across a broad range of skills and checks (and therefore is better than a +1 to a single skill). The difference is that once the character is made, that's it; there are few ways to increase stats during gameplay (unless the GM is generous with cybernetics). Whatever system you use, I guarantee players will hunt out small ways to increase their chances of success, and it's not the ability score system that's at fault (unless you strip it down to a no-bonus system), but the subsequent rules.

So how can we work with this in practice? Don't put any items in that allow characters to increase their ability scores. I'd be tempted to remove natural increases too (though given the flat math, I don't know whether that leaves no room for improvement). We can also select how the ability score bonus applies to game mechanics carefully. It's foolish to think that every ability is equally useful, but we can certainly make them all useful at some point. For instance, there is worry that Dex is a super-stat again (because you can choose weapons that use it, it applies to AC and initiative and obviously a bunch of skills). This can be ameliorated by making armour selection less Dex-biased (I think the original 3rd edition armour system worked well here, sans Mithral of course), or perhaps making Initiative based on Int (or no ability at all).

 

[justinhalliday]

The bell curve is already in place with the base score; it doesn't need to be applied again to the bonus. It just looks linear when put in tabular form. The +4 bonus is low probability already simply by being associated with an 18 score. Unless you defeat the distribution with point buy, of course, but if you do that you'll defeat the bonus distribution you propose as well.

Damned right. The bell curve is in the ability score generation method (to represent the various proportions of physical ability in the population). The scale of the possible scores is linear, so that it can encompass all of the total range of capabilities, from a 40lb weakling all the way up to a titan (in the case of Strength).

I think what I'd really like to see is hit rate divorced from stats - no matter how much of a bonus it is, as long as your chance to hit is involved people will game for the higher bonus. Take that off the table and other options open up.

If you're not using the ability scores (whether they represent relative Strength, Intelligence, or Dexterity) to determine who wins a contest, then what are you using?

 

[Minigiant]

If you're not using the ability scores (whether they represent relative Strength, Intelligence, or Dexterity) to determine who wins a contest, then what are you using?

I think what I'd really like to see is hit rate divorced from stats - no matter how much of a bonus it is, as long as your chance to hit is involved people will game for the higher bonus. Take that off the table and other options open up.

This is what I was getting to..

The playtest fighter is +6 attack (+3 from proficiency and +3 from Str)

If you drop 16 to a +1 bonus, you have to raise proficiency to a +5 bonus.


Look at 4E.
A similar fighter has +7 (+3 from strength, +2 prof, +1 feat, +1 fighter)

Remove the STR bonus and you lose +3. You can't hit jack squat with a +4 melee attack in 4E. That mean something else have to increase by +3 to make of the difference.

 

[justinhalliday]

This is what I was getting to..

The playtest fighter is +6 attack (+3 from proficiency and +3 from Str)

If you drop 16 to a +1 bonus, you have to raise proficiency to a +5 bonus.

Look at 4E.
A similar fighter has +7 (+3 from strength, +2 prof, +1 feat, +1 fighter)

Remove the STR bonus and you lose +3. You can't hit jack squat with a +4 melee attack in 4E. That mean something else have to increase by +3 to make of the difference.

You could always drop the AC against which you're testing, or simply live with a reduced hit chance...

There seems to be a gradual but clear trend to higher frequency of successful attacks. D&D Next is the most recent in this line (especially with advantage).

 

[howandwhy99]

The thing with pre-d20 Ability Scores is they weren't simply results from dice rolls. They were core descriptors of a character's abilities.

The hidden modifiers that followed the bell curve expression were each uniquely their own. Some were +/- modifiers, some were 1d6 rolls, some %ile, and some flat out resource additions like bonus languages or encumbrance coin.

Instead of having these systems in place to model the game world Ability Scores have become a referentially meaningless skill system - what in Forge terms means "Mother May I". This is popular though and a unified modifier table is much easier to track, especially when everyone at the table is understood to be doing so.

That the modifiers of this new Check [Skill] system aren't expressed as a parabolic relationship isn't that big a deal. The parabolic relationship isn't there anymore anyways. The die roll generation method for "assigning" Ability Scores became largely a legacy mechanic in 2000 and the modifiers became the only element useful element for play.

The Ability Score totals as Saving Throw DCs may change this, but I'm dubious it will turn out well. Unless they do a lot of Ability Score advancement than these DCs won't progress anywhere near what the Saving Throw matrix of the past did. Also, how will the 3d6 + keep method really hold up for these DCs? Maybe the 8 = zero point buy won't be bad, but it may still be undercutting Saving Throw advancement. We don't know yet.

The good thing is there's a lot more still to come and many changes will still be made. Otherwise, we can just house rule it.

 

[Minigiant]

You could always drop the AC against which you're testing, or simply live with a reduced hit chance...

Then if you shift AC down to match the lower attack roll then then unarmored is AC 7 and Clump Plate is AC15.

Feels weird.

There seems to be a gradual but clear trend to higher frequency of successful attacks. D&D Next is the most recent in this line (especially with advatage).

Lower accuracy is bad.
Missing both moralizes the player by plainly ending his turn and grinds the game down by forcing more turns to dofinish battle.

The problem with 4E's combat length was not high HP.
It was high HP plus a moralizingly poor 50% accuracy.

It took 4 hits to kill a standard. With 50% accuracy, that is 8 actions on average.

If 4E was like Pokemon where most attacks have 95% or better accuracy and the worst usable attacks are 75%, the game would be MUCH faster.

 

[Luce]

While high stats are always desirable it is possible to play PC with some weaknesses. In 2ed DMG there is a section about playing “hopeless” characters that is PC with scores below average (3rd tried to do something similar in an article “Wise As An Ox, Strong As An Owl” drg 284). Since ability scores have to be higher before starting to make a difference in play char with 8 and another with 12 in a non class requisite score were mechanically similar and it fell on the player to rp them differently. In 2ed I had seen fighters with str of 14 and/or wisdom of 8, not in 3rd or at least not for long (they died off). In 4th that would be even more suicidal by RAW. I am not saying one is a better system, just that they make different assumptions. I just like being able to play the stereotypes (too) without being over penalized.(big dim-witted fighter, the old mage with a bad cough)

In short I do not that much mind having (some) exceptional starting abilities (16+), but do mind having a large difference between starting and high level characters. Take 3e, in 18 levels a melee PC can get +5 (enchantment, wishes or tomes) +4(from levels) +5 (flame steward PrC) to STR. SO what started as what I see a reasonable STR 16 character turns into a titan wrestling STR 30 monster, before buffs. Yes, yes PrC are optional and the DM is in charge of what he hands out. Still in my (limited) experience the system itself requires it by assuming what the PC are capable of at a given level.

 

[keterys]

The problem with 4E's combat length was not high HP.
It was high HP plus a moralizingly poor 50% accuracy.

It took 4 hits to kill a standard. With 50% accuracy, that is 8 actions on average.

Whenever I see statements like this, I am reminded of how different some of our experiences in 4e were... I'm used to folks hitting on 3s and two rounds resolving most combats to the point where they're decided.

Missing on a 10? Maybe a fort attack on the Tarrasque.
4 hits to drop a standard? About two hits at low level. At higher levels, it does take about 4, but you're making two+ attacks per standard action, plus getting immediates.

I'd actually like to see something a little more tame than 4e.