Not only do different parts of the game need to interact easily, but the terms on which they do so need to be known and consistent. When you have qualitatively different sets of resolution mechanics and a module developer can't know for instance how combat is going to work in the game his module will be used in then how does that work?
The two games that I know that have a simple contest/complex contest mechanic that are meant to be used in tandem (depending on the table's preferences as to how much detail is warranted) are BW and HW/Q. I imagine there are other games like this too.
In both these games, the character build rules yield featurs/attributes/abilities/whatever that can be used in either simple or complex checks. This is a bit like 4e's skills - these can be used for a skill check, or a skill challenge; and a bit like 4e's attack bonuses - these can used both to attack a minion, or to attack a normal monster. If we think more about the 4e example, we also note that many powers do variable damage and/or inflict effects on a hit - if all the combat was against minions, this would be wasted rules text. So the existence of this rules text creates obvious pressure to have some combats against non-minions (of course, pressure for that comes from other places as well!). HW/Q and BW don't tend to have this sort of stuff - PC abilities/resources that can
only be used in complex challenges.
The promise of those two systems is that, as far as prospects for success go, it shouldn't matter whether you use a simple or a complex resolution mechanic, and as far as choosing a mechanic is concerned, it shouldn't matter which PC is brought to the table. The choice should be driven only by considerations of pacing/player buy-in/etc. In practice, I assume that the odds of various outcomes
are affected by choice of resolution mechanic - apart from anything else, I'd expect going for multiple die rolls over one die roll to reduce the swinginess somewhat. But the mathematical comparison between simple and complex is very involved and I haven't done it. I don't know, but I suspect, that the designers have relied more on playtesting than on maths to achieve a sense of "balance" or "equality" between simple and complex mechanics. (Notoriously, 4e is still grappling with this issue in its ongoing revision of DC numbers, skill challenge 'advantages', guidelines on what DC to use in what sort of challenge involving what sort of PC, etc.)
Systems that are qualitatively different will at the very least create different values for different things.
I'm not sure what you have in mind here, so am not sure if I agree. I've tried to explain how in HW/Q and BW, the qualitively different complex resolution mechanicsm aren't
meant to make a given score in a given attribute more or less significant in one version of resolution than the other.
You see this already a bit with 4e and the different scaling of skills vs attacks.
<snip>
Even if they all use the same d20 mechanic (and suggestions that they might not are frankly anathema to me) they also need to agree that a '15' means the same thing under the same circumstances or you'll have situations like "bull rush becomes worthless at high levels".
This is just a flaw in design. For modules to work in the way Mearls is talking about, I think this just has to be corrected. (Before the brawler fighter was published, or the improved grappling feat, I had implemented a house rule that grabs, bull rushes etc get a +2/+4 bonus at paragon/epic tier.)
As for my equating of combat with gamism and exploration with narrative
<snip>
Combat can be more centered around using mechanics in creative ways, while exploration could be more focused on the interactions between in-game resources. In combat you might think in terms of using the rules provisions for gaining an advantage over your opponent. In an exploration type activity you might consider using story resources for a more strategic advantage, which isn't likely to be modeled by a specific rule.
<snip>
Some people seem determined to have each activity focus be approached in the same way, and I don't think that's needed. In fact 4e seems somewhat uninterested in that kind of rigidity. It seems quite happy to have the focus in combat be very mechanical and somewhat gamist, but leaving other activities to much looser generalized approaches. That forms a type of expectation of design influence on play that can't hold when different modules don't have a knowable sort of approach. Detailed rules for feeding your PC don't make sense in the context of an overall system where travel is modeled by high level SCs.
I don't know if I follow all of this, but I think I follow at least some of it.
The bit about "approach" and "PC food rules" seems right to me. It's going to be hard to have a good, coherent game in which it is meant to be viable to treat rations and other resources both in a very abstract, "test your resource attribute" sort of way, but also in a "did you remember to buy enough food while you were in town" sort of way. The second approach, for example, requires fewer attributes on the PC sheet (no "resources" attribute) but brings a particular element of D&D play that's always been very central - gp - into the equation. How is all this to be balanced and reconciled? Dunno.
The earlier paragraph, about combat vs exploration, is a bit more hazy to me. But why, in exploration, can't I use rules provisions to get an advantage (eg I use a charge of my Wand of Metal Detection or take a swig of my Treasure Finding Potion). And in combat why can't I draw on my story resources - like using my dedication to Bahamut to help me fight this cleric of Bane, or - to be even more metagamey - using my earlier pleasant conversation with the captain of the guard to bring it about that he turns up to help me out when I'm being overwhelmed by a couple of assassins in a dark alley (The Riddle of Steel allows the first sort of "story resource", andh HeroWars/Quest both sorts of "story resources", to figure into combat).
I don't really know how you deal with this kind of thing in a BW game.
I've tried to describe above. I'm not drawing on a lot of play experience with BW, though - I mostly rely on it to give me guidance on how to run my 4e game! - so it would be interesting for someone else with more experience to chime in.
I think being able to use whichever subsystem you want on a case-by-case basis at least means you get to think ahead of time about what types of factors are likely to play a part in a given combat for instance and pick a system that deals with that on the level of abstraction it deserves. At least if there are going to be factors which are important that will say only come out in a tactical resolution you might pick that system instead of the 'fast and dirty' system that won't really model those factors. If you are then talking about a game where the resolution systems are fixed for different things to start with that's fine WRT that game, but it becomes pretty hard to create material for the whole system. Is a Roper a tough monster? Well, it relies on a specific tactic to do its thing. It may be tough for one party and easy for another depending on what builds they have and what tactics they use. None of that detail is likely to be captured using a combat module that doesn't bother with modeling tactics in a detailed way. Things have different values under different mechanics. I don't know how you have a really coherent system that way.
For these sorts of cases - like "will the muddy ground factor into the fight" or "will they have a special advantage against a Roper because they've got a lot of bottles of spirits and really good Escape checks" - I think you're right. A game like HW/Q relies on these sorts of tactical issues not factoring into resolution in the same way as in D&D - rather - to the extent that they are relevant at all - they generally provide augments, and augments can be applied in either simple or complex contests.
In BW, I think that that sort of detail
is going to get lost if you go for simple rather than complex resolution - just as damage rolls and effects, in 4e, get lost if all you're fighting is minions. I think the BW attitude to this is the same as the 4e attitude to fights with minions - you win some, you lose some, it will all come out in the wash! The less plausible, and the more like desperate handwaving, that that seems, the less viable the "modules" idea will look, I think.
Another game I know of that has modules, and uses them in what seems to be Mearls' envisaged way - ie choosing by campaign rather than from situation to situation - is HARP. It has three or four (maybe more?) combat mechanics - the core rules, the Martial Power rules (which are like a more nuanced version of the core rules), Hack and Slash (which makes the rules a bit closer to classic Rolemaster), and another mechanic based on the Rolemaster Express combat system. And some people just plug in the classic Rolemaster mechanics.
This is mostly viable as far as character build is concerned, because all the mechanics look for the same numbers as relevant inputs - how much attack bonus, how much defence bonus, etc. (Although even here, there are a few issues - different modules make weapon choice, or armour choice, matter a bit more or a bit less). And the outputs are also described in the same way across all the systems - concussion hits delivered plus individuated injuries and other conditions inflicted. (Here, again, though, there are some minor implications for healing rules.) But the actual effect in play of the different modules is fairly different - Hack and Slash, and core Rolemaster, are swingier than core HARP, for example, because they make it easier to kill on a typical attack. (So Hack and Slash is often suggested as an antidote to combat grind in HARP.)
Personally, I would expect Mearls' modules - if they move from hypothetical to actual - to look more like HARP than like BW. Which gives rise to the "fracturing" concern that I voiced upthread (as did others - including you?).
