Dealing with MAD; Stacking vs. Substitution

[Mendicant]

I've been toying with a design concept that I'd like some input on. In most design I've seen that actively tries to manage MAD, there is a tendency to have one stat do the job of another one. This can be accomplished with a hard substitution, such as Weapon Finesse or Agile Maneuvers, or to a lesser extent through a stacking effect that is so cost-effective it tends to work as a direct substitution anyway--Divine Grace comes to mind here. The result (and, circularly, the justification) is that an optimized character typically has an array that's as spiky and specialized as you can get away with.

As a way to offer an alternative, I've considered expanding the number of stats that contribute to things, and making sure any contributions stack. For instance, if Dex and Strength both contribute equally to your attack bonus, there's an increased incentive to spread resources more evenly there, as two 14's are cheaper than a single 18. This wouldn't always work the way I want, obviously: if, say, Wisdom and Charisma both contribute to Will saves, then a Sorcerer is going to have even less incentive to put points into Wis, not more. Overall though, I think if done right this would punish people less for spreading their stats around, and make it a bit easier to build effective non-casters without a laser focus.

Is this a good idea? How would you execute it? How invasive or disruptive would it be to current game balance? If this was done, what things should stay solely the province of a single stat?

[OldTrees1]

Here is my experience with MAD via playing Fighters:
I am for versatile fighters and thus the class acts like a Str, Dex, Int, Cha > Con >> Wis MAD class. Yet it can be played reasonably well with 14/14/12/14/08/14(or 28 points). The real difference between a 14 and an 18 is a mere +2 bonus on a d20 die roll. That is really not much of a difference except at really low levels. In addition the first +3 has a point buy cost equal to the last +1. So point buy already has a lot of factors favoring the 4x14(24) vs the 2x18(32). But why does this play reasonably well if you are still that +2 behind on every d20 roll? The answer is stacking, albeit not directly. The 14 Str powers the attacks that 14 Dex grants, while the 14 Cha buffs the skills that the 14 Int grants. Now 1 example class style does not a conclusion make, but this is evidence that synergy between medium scores can match a high score.

[Mendicant]

The problem is that without direct stacking, the opportunity costs of spreading your abilities thin outweigh the synergy benefits. Frankly, the abilities frequently don't really "synergize" at all. There's no amplification of your social skills happening when you give up some Cha for a higher Int, for instance. At best you're using Pathfinder and the rank buys a trained bonus you wouldn't have otherwise--in that instance you gained a +4 bonus in one skill where a higher charisma would have given you a +1 bonus in four skills. At no point is the combination of Int and Cha giving a synergistic effect where 1+1+2.5. The same is true of Dex v. Strength--the damage bonus from your mighty composite longbow isn't making up for the damage penalty for lower accuracy, or is at best breaking even.

[OldTrees1]

The problem is that without direct stacking, the opportunity costs of spreading your abilities thin outweigh the synergy benefits. Frankly, the abilities frequently don't really "synergize" at all. There's no amplification of your social skills happening when you give up some Cha for a higher Int, for instance. At best you're using Pathfinder and the rank buys a trained bonus you wouldn't have otherwise--in that instance you gained a +4 bonus in one skill where a higher charisma would have given you a +1 bonus in four skills. At no point is the combination of Int and Cha giving a synergistic effect where 1+1+2.5. The same is true of Dex v. Strength--the damage bonus from your mighty composite longbow isn't making up for the damage penalty for lower accuracy, or is at best breaking even.

Huh? I just cited a case where the synergy drastically outweighed the opportunity cost.

By having 14 Int & 14 Cha instead of 10 Int & 16 Cha I get -1 to Cha skills but gain 2 more skills at max ranks. This allows you to go from just Intimidate to Intimidate, Bluff, & Diplomacy. At first level I trade a +1 ability modifier to 3 skills for +4 ranks to 2 skills so a net swing of 8-3=5points. This only increases with level.

By having 14 Str & 14 Dex instead of 16 Str & 10 Dex, I get -1 attack, -1 damage, +2 AoOs per round. That is quite the trade IMO.

So, yes there are cases like the composite longbow where there is insufficient synergy, but there are also good examples with sufficient synergy that can be used as models to fix the other cases.