Solving DCs - Know Your Group

[NewfieDave]

I keep seeing complaints about skill challenges (and DCs/defenses in general) being too easy/too hard. I've read through some detailed math that shows the system is flawed, and it makes a compelling argument. This seems strange to me. How could such a flawed system make it through playtesting? Why don't I hear as many complaints about the system from playtest reports as I do from statisticians?

The answer is that D&D is not 100% math (to most people). There's a lot of art to it as well. In a game where your imagination is the only limiting factor, designing a system to account for every possible variable and complexity is impossible. The tables in the DMG for setting skill challenge and other DCs are merely reasonable guidelines to follow when crafting your own encounters. The best encounter is one that engages the PCs and makes the session fun for everyone.

So ultimately, the best "fix" for the skill challenge system is very simple: know your group.

Don't put a skill challenge involving jumping over 50 pits of lava against a group a PCs who are all untrained in Athletics/Acrobatics.

Don't force a reasonable PC choice to go against an out-of-whack DC just because the tables tell you "that's the way it is, so buck up!"

Do put the PCs up against challenges that are related to their abilities and jive with their personalities/goals.

A good rule to follow is the 70% rule. A roll against moderate difficulty should succeed 70% of the time. In d20 terms that makes the DC 7 + Skill Mod. On the flip side, harder checks should have around a 30% success rate (DC = 14 + Skill Mod). These numbers are averages, and it's recommended to deviate from these averages (both up and down) to make things more interesting.

The other complaint about skill challenges (and perhaps it's the more valid complaint) is that the complexity system is out of whack. I'm going to be ignoring the complexity tables altogether when I DM because it seems like an unintuitive approach. I'm comfortable with letting common sense be the final arbiter on where the party succeeded or failed because the complexity system math doesn't model the art very well. It shouldn't be how many failures the party has earned that ends the skill challenge; it should be how they failed that ends the skill challenge.

 

[Terwox]

I respectfully disagree.

The best fix is one that gives you a table, statistically derived, that will do this work for you. 4E isn't about careful allocation of skill points.

This is especially true for new DMs who will be using skill challenges -- it is hard to know your group and their capabilities if you just started running the game.

 

[Tony Vargas]

A simple 'fix' to the skill challenge system is to ignore the number of failures. Just go until you get the propper number of successes, and only call it a failure if there's a very bad failure that would logically blow everything.

Personally, I wonder if the skill challenge system isn't designed to result in failure more often than the combat system. A failure in a combat challenge can equal a TPK or a capture scenario. A failure in a skill challenge might just dish a little damage from a trap, force the party to face an extra combat challenge, use up some resources, or whatever. The 'story' can survive a lot more skill challenge defeats than it can combat defeats.

 

[TheGlow]

Nod, I saw all these posts about Skill challenges and read into and saw that one could design a quick, simple method for giving out a little XP without fighting.
So with so much riding on a few toss of the die instead, I guess I can see why they lowered success/raised failure rates.
And as noted earlier, failing a conversation or so can probably still carry the adventure whereas catching a beating by some trolls takes a little more to work with.

 

[Saeviomagy]

The problem is that the skill challenge rules provide you with a system that is incredibly sensitive to change.

Suppose you're using the skill challenge system as written, with the sole exception that you're carefully targeting DCs to match your party's capabilities.

You want a decently long experience, so you go with a complexity 5 challenge (12 before 6).

You've decided you want a win rate of 50/50.

Imagine your surprise to find out that it's impossible to give that win rate.
Your only alternatives are 40% win rate or 60% win rate, and that difference occurs over a single point of DC.

Oh well, lets give em 60%.

What happens if you misjudge?

Well, one point of DC too high and your 60% win rate turns into a 40%.

One point too low and your 60% turns into a 75%.

 

[NewfieDave]

The best fix is one that gives you a table, statistically derived, that will do this work for you. 4E isn't about careful allocation of skill points.

While I agree that such a table would be awesome, I don't think it's possible given the breadth of options available at the gaming table. There are simply too many "What if?"s to make a simple level by level table work for all possible party combinations and applications of creativity. It could come close though, and I think the tables in the DMG are already close enough for my needs.

 

[ravenight]

Basically, what you describe in the OP is simply throwing out the skill challenge system and designing more free-form social encounters with the DM's intuition. So that isn't really a fix to the skill challenge system. The point is, the DMG provides a system with a good base concept (designing social encounters with clear rules for success and failure) but completely flawed math. Changing the math to something that works well (essentially, discovering the "reasonable guidelines" that should have been in the DMG) means doing statistical analysis of some sort.

 

[VannATLC]

As far as I can see, the biggest problem with Skill Challenges is the general idea that the party is intended to succeed more often than fail, although I acknowledge there has been solid mathematical analysis that shows a vast amount of swingy-ness, based on certain assumptions.

 

[LostSoul]

The key is to know, generally, what the chance of success is and to negotiate success with your group.

If failure is as interesting as success - even if it's not what the PCs or players want, it's still acceptable to the players - then a high failure rate is okay if the rewards are big enough.

 

[NewfieDave]

Basically, what you describe in the OP is simply throwing out the skill challenge system and designing more free-form social encounters with the DM's intuition.

Sort of. I think the current system lays a good framework for what skill challenges should do, but it has two flaws...

The first flaw is that a table of DCs can't model every possibility (although it can certain serve as a handy guideline). Yes, you can map the roll probabilities and use all the advanced statistics you want to come up with a perfectly balanced table. That's the "science" half of things, but how do you account for the "art" side of playing D&D? Player creativity and smart thinking the DM didn't account for (it's happened to all of us at one time or another) requires free-form and intuitive thinking on the DM's part.

The second flaw is that the only thing complexities add to skill challenges is... complexity. And limitations. I would rather leave the number of successes and failures as open variables with a page explaining what adjusting those variables does to success rates. What if I want a skill challenge with 7 successes and 5 failures? What if I want to allow failures to be overcome by secondary rolls?

There are so many potential applications of the skill challenge system, and even more ways that it could be modified. Unfortunately, the current complexity system does not support that very well. I think the reason we have the complexity table instead of an open variable approach is to make the game more streamlined (2X Successes = X Failures is pretty easy to remember).

I'm hoping we'll see lots more expansion on the skill challenge system in future supplements like the DMG2.