Fancy title, huh? Let's see if I can live up to it.
Introduction, In Which I Rant and Rave
(You should probably skip this section if you're easily offended by pro-4E, anti-3.5 sentiments. Or just don't care about my personal opinions in the matter.)
D&D 4E is my D&D. I've tried some older versions, primarily 3.5, but never liked them. There were too many things that rubbed me the wrong way: the Linear Fighter-Quadratic Wizard problem, the restrictive classes, the many traps in character design, the the often bizarrely specific spells... Y'know, let's not go there. Let's just say that while I could play 3.5 I had to constantly fight my urge to overhaul the system from the ground up.
Enter 4E. Here was a game that was brave, bold and unabashed, a game that seemed to say "Yeah, sure, I can let you do that deep roleplaying if you really want to... but if you want to kill monsters and steal their stuff, nobody does it better". And I think that's true. If you want a game that has tactical heroic fantasy combat* as its core functionality, there is no better alternative.
Which is not to say that 4E is without faults. I'm not talking about balance problems as any sufficiently complex game will have those, and 4E has done an amazing job of keeping those minimal in any case, or the inability to do things that 3.5 could do, which I think is highly overstated. (Multiclassing is what's usually brought up, but to that I say that in 3.5, multiclassing was good for exactly three things: finding overpowered builds, overcoming the inherent flaws of the class designs and giving the illusion of choice. Sure, you could add any class at any level, but at least 90% of the time it was a pure mistake. Many of the complaints are also based on identifying the name of the class with the character's place in the world, like it's impossible to be a druid without being a Druid.) No, the main problem with 4E is the skill challenge system.
Don't get me wrong - I like the 4E skill system. The skill list is just right for me, with enough diversity to make it interesting but not so many skills that you forget what they're for or that any seem so specific as to be unnecessary. (Use Rope, anyone?)
The Skill Challenge Challenge
But for me, skill challenges are a big disappointment. The idea is great, that you have a mechanical framework for making something kind of like a combat but with skills, but the execution is sorely lacking. The first attempt, in the original DMG, fell completely flat. It was fun as a change of pace once or twice, but soon it was obvious that all you would want to do was use your best skills over and over and over again. There was no room for tactics or variety, no need to think at all. The ideas that have been added later have been interesting, but mostly they boil down to "here are some ideas, make it up for yourself". When compared to the solid and exhaustive support the combat system gets, with a neverending flood of monsters and combat powers (not that I'm complaining about that!), the skill challenge system is sadly neglected, like a red-headed stepchild that smells weirdly of what might just be leprosy.
In my fairly long-running 4E game (something like 30-40 sessions so far) I have made many attempts at spicing up the skill challenges I have thrown at my players, treating them as mini-games to the main game of combat encounters. Unfortunately, since there is so little support in the rules it's a lot of work. I can whip up an interesting combat encounter in less than half an hour if I have to (and that includes transcribing the monster stats to a format I like better), but an interesting skill challenge takes days of pondering and hours of actual design work because I basically have to design a game from the ground up. Sure, it's a tiny, throw-away game with some pre-defined rules, but still. If I have to create a skill challenge in half an hour it will almost certainly be a dull optimized-skill abuse contest because anything else requires adding new mechanics.
So what I want to do is create a framework for designing skill challenges, set up some general rules that form a design space that is general enough to handle any kind of situation but still complex enough that the choices it creates are interesting. Preferably it should allow for the creation of new powers that further increase the tactical and strategic choices for the players.
Does it sound impossible? Perhaps it is. But I'm still going to try, and here are the first 2.5 ideas.
Idea the First: Effect Rolls
This is something I came up with when I was looking at the combat system for ideas on how to improve the skill challenges. In a vanilla skill challenge all that matters is success or failure but in combat it also matters how well you succeed, primarily measured in damage. So why not import this feature?
In practice this means that instead of requiring a certain number of successes you require a certain amount of effect, with each successful skill check followed by an effect roll. A base effect roll might be 1d6 plus a relevant ability modifier, probably the one the skill is based on. For example, a successful use of Diplomacy might add 1d6 + Charisma-bonus to the effect. Being trained in the skill could add another +2 to the roll, with a further +2 for having a feat that gives a bonus to the skill or a skill utility power based on it. This already adds something to look at beyond "which is my best skill, I'll use that... again", but there are more possibilities. For example, each use of a specific skill (by a specific character or the party as a whole) could reduce the effect roll for that skill. Suddenly hammering away at the Stealth button doesn't help as much! But is it still better than using a skill with a lower chance of success? A choice - an interesting choice.
Now add another wrinkle: multiple effect tracks. I'll quote an example from my comment on this post, which followed this one - the discussions in the comments are very interesting, and while I think The Jester is a bit too harsh in his criticism of the 4E skill system he makes many good points:
For example, if the goal is to convince the baron to send out his men to fight the goblins in the woods, using Diplomacy might add 1d6+Cha-modifier to the "Convince the Baron" track with no consequences for failure, while using Bluff might add 1d10+Cha-mod for a success but add 2d4 to the "Baron's impatience" track, and successfully using Intimidate might add 1d6+Cha-mod to both. Once either track reaches a certain threshold the outcome is decided or the challenge changes, perhaped becoming "Escape the baron's guards" or "Lead the baron's soldiers against the goblins".
You can also do more interesting things with support skills than give someone else a bonus on the "real" roll to succeed, like making minor changes to the tracks or modifying the effect rolls for success or failure for another roll. In the example above someone could make a Diplomacy check before the Intimidate check to reduce or prevent the effect on the "Baron's impatience" track for example.
Idea the Second: Posture
The second idea is to give multiple ways to use the existing skills. This could be likened to maneuvering on the combat map: do you go straight for the enemy, support your allies or maneuver around the fringes looking for an opening? The trick here is to find modes of behavior that are abstract and general enough that you can apply them to social situations, traps and puzzles, action sequences, investigation scenes and so on without being bland. There should also be enough of them that there isn't a single best option in every situation. Three is probably the bare minimum, and ten too many; something like 4-8 should be ideal.
An example of how you could do this would be to have the following postures:
* Direct: You attack the problem head-on. You will get a big positive effect if you succeed but probably a big negative effect if you fail.
* Cautious: You proceed with care, observing the effects of each action. A success won't give you much direct effect but could give you information about how to best proceed, and a failure won't cost you much.
* Indirect: You skirt around the edges of the problem, looking for other ways to attack it. A success can have a big effect on a secondary goal or give bonuses to characters taking a more direct approach, with failure having similar but negative effects.
* Defensive: You concentrate on preventing failure. A success strongly mitigates the effects of failures for other characters and if nobody fails will have a moderate positive effect. A failure will have very little effect.
* Supporting: You back up another character in their actions. If you succeed the one you're helping will either get a bonus to their rolls or have reduced negative effects on a failure, but if you fail they will get penalties to their checks and if they also fail the effects will be stronger.
Ideally these choices will give the players more interesting choices, and an optimal approach should involve a mix of postures. If the challenge limits which skills can be used with which posture things become even more interesting: do I use my best skills in a risky or otherwise suboptimal manner, or do I make a safer move with a different skill?
Idea the Second-and-a-Half: Status Effects
If we have the equivalents of damage and maneuvering, status effects seem like a natural continuation. However, this is trickier since it seems even more tricky to come up with generic status effects that work as well with a ball at the king's castle as a chase through the secret passages inside it. However, by going to what status effects do in the combat game, which is primarily to reduce what actions you can take, I can see a couple of ideas:
* Checked: The opposition is blocking you from taking a direct approach. You cannot use the Direct posture or use the Supporting posture to support an ally's Direct posture.
* Isolated: You are unable to help or receive help from your allies. You cannot use Supporting posture and an ally cannot help you with the Supporting or Defensive postures.
* Misled: You have misunderstood the situation. You cannot use the Cautious posture and any negative effects of failure are increased.
I would like to add one more dimension to the effect tracks-posture field to give additional options, but I'm not sure what that would be. Status effects only affect how you can act on the other two dimensions so it's not an independent axis but more like terrain on the other two. Or perhaps choosing skill and posture under the limitations of status effects and effect rolls is enough? I'll have to cobble something actually usable together to see.
One thing I would definitely like to have are standardized opponents to add to these challenges. Probably not as detailed as specific characters or situations, but more like types of situations. Is it a race against time (getting across a bridge before it gives out, getting a trade-deal signed before the end of the month), a chase (trying to catch a spy fleeing over the roof-tops), a delicate situation with unknown consequences of failure (a tense dinner with foreign dignitaries, a strange mechanical trap) or something else? Perhaps that's not possible and each situation will have to be designed by hand.
As I mentioned above, I think that these ideas could give skill challenges the complexity needed for it to be possible to design interesting powers focused solely on these situations. When there is more to manipulate than the success or failure of individual skill checks there is much more room for designing interesting effects.
Currently, skill challenges have great potential but the lack of a formal structure makes it far too difficult to realize that potential. Combat encounters are easy to design because there are so many ready-made monsters that you can just drop in, but there is no equivalent for skill challenges - if combat encounters worked like skill challenges currently do, not only would you have to design the opponents from scratch, you wouldn't even have anything to base your work on but a level one kobold and a zombie minion. Of course skill challenges are unsatisfactory - imagine if all combats consisted of beating up various combinations of kobolds and zombies, with the most creative being a zombie kobold.
What skill challenges need is a stronger formal structure that allows for interesting choices while being general enough that you don't need special cases for different types of skill challenges - a social challenge should work the same as a chase or a dangerous journey through a wasteland, because the skills themselves work the same, just like every combat uses the same rules. That's what I mean by formalizing the design space: creating that general structure that allows the GM to focus on creating a specific interesting skill challenge instead of trying to find a way to make skill challenges interesting in the first place.
I guess my players will soon be guinea pigs again.
* I'm aware that this is pretty specific, but it's also broadly appealing. There are plenty of games that do other (also pretty specific) things better, but if you want a generic, combat-heavy fantasy romp RPG then 4E is the game.