D&D 4E Heavy Concrete Data on 4e's Skill Challenge System (long, lots of tables)

[gonesailing]

Even if the system worked it's still just a couple die rolls, what the players do matter little. There is no fun in a skill challenge.

Really...Even the modified version Stalker0 created? I admit I am still trying to understand how to use them effectively, but I think they can be fun if used right. Maybe this belongs in another thread, about how to use Skill Challenges (any variety) effectively in an Adventure/Campaign. I will let the Mathematical Gurus come up with their systems and use the one(s) that fit my needs.

 

[Stalker0]

Stalker, this is a great thread. You always have something interesting to say, and this is no exception. In fact, this probably takes the cake. Thanks for doing all this.

I have a question, and please forgive me if it's been addressed already. What happens if, at each difficulty level, we increase the number of failed rolls required to fail by one or two? It seems to me that a good way to affect overall success rate is not by decreasing individual DCs, but decreasing the chance of overall failure.

I appreciate the compliments. While I understand not everyone is going to agree with the numbers I've presented (or more precisely the interpretation of them) its nice to see so many taking the time to look them over. This data is for you all!!

To your question, let me make sure I understand what you are asking. Are you saying change a complexity 5 challenge to 12 failures and 7 successes?

If so, if we look at our standard 50% mark, which normally has a 7% chance of success. Going to 12/7, that number becomes 11.89%.

At 65%, normally its 41.97%. With 12/7, its 59.4%. The variance is still very high, but it does help some.

 

[Dayspire]

Speaking of Stalker0's Revised system - any update on when the new version is coming out, Stalker?

 

[Stalker0]

Speaking of Stalker0's Revised system - any update on when the new version is coming out, Stalker?

I've got all the core mechanics I want in place, now I'm tweaking numbers to get everything good and ready. That's actually just as hard, getting the variance down in this kind of system is very very tricky, but its coming along well. I hope to have it done in a day or two, heck if I'm awesome maybe even today.

 

[Soyokaze]

Ah, but pg42 isn't the be-all end-all.

I think I may have found a large part of our problem. As pointed out on this thread http://forums.gleemax.com/showthread.php?t=1044606 the DCs on table 42 aren't always the easy, medium, or hard DC of your level.

Look in the PHB. Look at the DCs for Balance under "Narrow or Unstable." DC 15. So, if I have a lv18 character who has to retain her balance while running across a roof that's a DC15 in a situation without further complications. It doesn't magically get higher because they are stronger.

So, not all of the DCs are going to be in that 20, 24, 28 range, even though she is a lv 18 character.

That being said, I don't think a DM should design any skill challenge in which there are auto-successes. There might quite simple, with only a 10% or so chance of failure, but auto-successes are the skill-challenge equivalent of the monster that's so weak it doesn't give you experience any more. It just means getting those successes may be easier.

 

[Obryn]

The idea of analyzing all skill challenges mathematically is flawed. Every skill challenge is meant to be unique, and tailored to the individual party, per the DMG. There is no such thing as an 'average' skill challenge.

Saying that parties generally have a 8.73% chance to win an average skill challenge is like saying they have a 8.73% chance to win an average combat. What's an average combat? No such thing. The term is way too vague to have meaning, let alone be statistically analyzed. In this case, it simply means your concept of an 'average skill challenge' is flawed. By the DMG, an average skill challenge is one in which the the party has roughly the chance to win it that the DM desires for them to have (since he has prepared it custom, for that party).

I definitely agree that each skill challenge is different, and the tables in the DMG aren't a straitjacket - just a starting point.

However, as a DM, I like to have some idea of the odds when I tailor-make or adjust a challenge for my players. For combat, there's too many variables to measure, but for a simple skill check, it's an easy percentage.

That's what - in theory - game designers do. They actually check the math - that's their job. It's the reason, for 3.5, why Power Attack was changed for 2-handed weapons. It's the basis behind the CR systems. In theory, the math is done well and it becomes more transparent to us, the players. We, as DMs, are meant to work from there, adjusting the scales upwards or downwards.

The problem is that a lot of the math that goes into skill challenges is counter-intuitive. For example, I would not have guessed that a 70% success rate per-check swings the odds in the players' favor the longer the challenge goes on. That's surprising to me, but it's clear now that Stalker0 has pointed it out. By the same token, I need to know that - by default - a lot of the DCs in the DMG require me to give a lot of ad-hoc bonuses if they're going to work out in the players' favor.

Tables and discussions like this help DMs make exactly the kinds of rulings you're encouraging us to make.

-O

 

[Cadfan]

Hmm. I think its pretty clear there's a problem. Specifically, the problem has to do with the fact that a "medium difficulty" on a single d20 roll generates a "very high" difficulty when you have to make many d20 rolls. Even a "low difficulty" on a d20 roll produces a higher than expected difficulty if you have to do it multiple times.

That being said, I'll have to look more closely to see if I expect any real problems of my own when using the skill challenges. Take the Negotiation skill challenge from the excerpt. It has 3 skills which are recommended as "moderate," and one which is recommended as "easy." I will almost definitely be granting +2 bonuses for roleplaying, and the use of "aid another" is explicitly noted as a possibility (though not for Diplomacy, which I don't get, I'm assuming that's an oversight since using aid another for bluff makes equal sense as using it for diplomacy). There's also a note that the PCs might carry out tasks to obtain the NPC's favor, which in a way is a free success in the skill challenge.

Other skill challenges have similar matters which reduce their difficulty. For example, a skill challenge relating to physical activity is going to end up being a mixture of "easy/moderate/difficult" DCs, and objective DCs from the skill section of the PHB. The objective DCs could either increase or decrease the overall difficulty, but at least they give the players a chance to estimate the difficulty before deciding whether to go that particular route.

So... I'll have to read more carefully. It may play acceptably well. But I do agree with the OP that giving the players a bunch of tasks with a 60% chance of success on each individual one, then expecting them to attain 8 successes before 4 failures, is not a good idea at all.

 

[Wolfwood2]

I think the underlying assumption of level-appropriate challenges is going to prove a lot less true for skill challenges than for combat. Sure there are going to be some skill challenges that the DM has set up in advance, but I expect that the vast majority of skill challenges will be a result of "exploring the world" when the PCs decide on the spur of the moment that they want to do something.

Example: The PCs are investigating some bad guys, and they suddenly decide they want to sneak into the (access restricted) Royal Library and see if they can find any information there. This is something that could reasonably happen with 1st level PCs, 4th level PCs, or 8th level PCs. The difficulties won't change based on the level of the PCs.

The natural effect is that if level-appropriate skill challenges don't succeed very often, the players will naturally start seeking out only level inappropriate challenges beneath their level. (And they should be able to judge reasonably well after playing with a DM for a while.) On the other hand, if level appropriate skill challenges succeed a lot, players will more often seek out level inappropriate skill challenges because the gamble is worth the risk.

Sure they'll get more or less xp, but most skill challenges are going to be driven by things the PCs want to accomplish in-game rather than particularly caring how much XP they get out of it.

 

[GoodKingJayIII]

To your question, let me make sure I understand what you are asking. Are you saying change a complexity 5 challenge to 12 failures and 7 successes?

If so, if we look at our standard 50% mark, which normally has a 7% chance of success. Going to 12/7, that number becomes 11.89%.

At 65%, normally its 41.97%. With 12/7, its 59.4%. The variance is still very high, but it does help some.

Yep, that's what I was asking. Basically, I wanted to make overall failure more difficult (but still possible) while keeping the total number of successes the same, to ensure that everyone has an opportunity to succeed and that the skill challenge wouldn't hinge on one or two characters.

I think I understand your overall problem with the system. It's not just the failure rate (which is too high in standard cases and far too low at extreme cases), but the high level of variance in the system.

I don't have any answer for the second. As for me, I'll probably end up increasing the amount of failures by two, and reducing the amount of successes by one.

Ultimately, I'm just going to have to see how this works out in game. I'm not entirely sure how I will introduce skill challenges, or how my players will approach them. Also, there's a good chance that I will be running a game for one or two people, in which case the skill challenge system would probably not work at all, as written.

 

[fanboy2000]

I called WotC customer service today and they said that yes, you can use Aid Another on skill challenges.