Skill Challenge DCs

[Skywalker]

Am I doing something wrong or are the errata dcs a little too low?

IMO the DCs aren't too low, though they can appear that way. There has been a recent WotC article explaining the math and I think the design is sound

It is worth noting that in order to get an even chance of success for the entire challenge, you need individual skill rolls to have the following chance of success:

Complexity 1: 67%
Complexity 2: 75%
Complexity 3: 80%
Complexity 4: 83%
Complexity 5: 86%

As you can see these are pretty high.

The trick though is presenting a range of DCs that allow PCs to have a chance of success, without just relying on their highest skill (which is dull). Even if a PC has pretty inappropriate skills, you want to leave some avenue open for them to roll provided they act in a clever or creative manner.

The current DCs reflect this. Easy DCs are really only able to be failed by the untrained. Mod DCs are generally able to be succeeded by the trained. Hard DCs remain a challenge.

The difficulty comes when you present your Skill Challenge. If you do so transparently and reveal all the options and mechanics to the players, then they simply align their skills and the DCs are too low unless you have made them all Hard DCs.

If you do as the DMG suggests and simply note Primary Skills to the players, but don't reveal the exact mechanics of the challenge, then the DCs work just fine.

I think it is important to make sure that you follow the action and not the skills. So the skill rolled is determined by the narration made and not vice versa.

Beyond this there are some advanced tips and tricks you can add to a Challenge to make work even better. Allow PCs to increases DCs for increased successes, put in time limits, have rolls interact or provide autofailures, can all help to focus the players on getting clues from the narration and not on the math.

For example, in Demon Queen's Enclave, I made a Skill Challenge where the DC to intimidate the Drow Matron was low for the first two successes (she was in a poor position and respected strength). However, further uses of intimidation produced auto-failures as there was a tipping point from showing strength to making her look foolish. The idea of a tipping point was described through the narration of the scene.

FWIW I would avoid the suggestion to add +5 DC across the board. This tends to promote Skill Challenges where PCs only use their best skills, reducing the creativity and variability that is in the existing system if used correctly.

 

[Jhaelen]

IMO the DCs aren't too low, though they can appear that way. There has been a recent WotC article explaining the math and I think the design is sound.

Unfortunately, not being a DDI Insider I couldn't read that article.

Based on my admittedly very limited experience with skill challenges the new values really are too low. During our test week my players breezed through a Complexity 4 skill challenge without a single failure. And that was with adjusted DCs (the original DCs were in the range of 10 to 15 which meant some pcs would have had a higher skill bonus than the DC!).

They figured out almost immediately that they should try to let those who were trained in a certain skill make that check and used aid another to bump the results even higher.

Imho, the new DCs only make sense if
a) aid another is not allowed
b) everyone is forced to participate every round
c) everyone is required to make a particular skill check at some point (i.e. forcing pcs to use untrained skills)

The new DCs also don't make sense when compared to the static DCs given in the PHB and opposed skill checks, because those are always way higher.

One thing that particularly irked me was that the module we used (the Ashen Crown adventure for Eberron) suggested perception DCs that didn't make any sense, e.g. in one encounter it suggested a DC of 10 to detect a hidden gnoll hunter with a a Stealth modifier of +12 (I'm not 100% sure about the correct numbers, but you get my drift). Umm, yeah.

What I did was this:
I looked at the ability modifiers of monsters at every level and used those values as a basis for setting DCs at that level. What I ended up with was a DC progression that lies somewhere between the original DCs and the errataed DCs.

Imho, having skill challenges with a 100% success chance is a waste of time. Where's the challenge in this? Since failed skill challenges should never be show stoppers, it doesn't matter if the pcs don't always succeed.

 

[Mustrum_Ridcully]

Unfortunately, not being a DDI Insider I couldn't read that article.

I think the most interesting excerpts are in my post, directly adressing the issues you have with the system.

-

Can characters win an encounter just by always using your at-will powers with no tactics and strategy? Anyone ever tried that? Somehow we still don't do it.

The problem with the framework is not the DC, but that the challenge doesn't get more interesting or you get better benefits (like not losing healing surges) if you use "tactics".

So I think part of the responsibility of a good skill challenge is to add some tactics.
1) Certain skill checks open up new skills to use.

2) Skills can't be used infinitely often. Among that - I don't think aid another should be generally applicable. It is a "tactic", but one that's not interesting.

3) Finite time. If you don't succeed within a certain time frame (represented by checks you make), you can't succeed or at least don't get the best results. This can even counter the "everyone uses aid another" trick. If you have only X "rounds" in each which every character can attempt one skill check, and you need 2X successes, some players have to take risks.
Maybe the time limit is not strict in the sense you fail big time. But if you don't manage in time, you lose an advantage. (For example, if you don't chase after the criminal fast enough, he will find some additional allies to aid him. If you fail entirely, he also sets up an ambush for you and/or destroys evidence.)

4) Different skills change the nature of your success. If you use Diplomacy, you might get an ally. If you use Intimidate, you lose an ally but gain a reputation. If you use Arcana, you gather some Residuum, if you use Religion, you recover a healing surge or divine power.
The minimum result should always be a different narrative, but it should also provide some "tangible" effects - be it represented by mechanics or by story awards.
Even if there is still a straight route to success - just use Diplomacy over and over, knowing that a different route might change the outcome can be enough to make things interesting.

The biggest trick might be coming up with restrictions and special benefits (especially ones that feel appropriate). It's not something the framework can give you, just like the encounter design rules can't tell you whether you should use Goblins or Kobolds in your adventure.

 

[Kestrel]

Imho, having skill challenges with a 100% success chance is a waste of time. Where's the challenge in this? Since failed skill challenges should never be show stoppers, it doesn't matter if the pcs don't always succeed.

Exactly. In fact, I want more failures because to me, these are just complications. If the party fails 3 times, I place an obstacle in the PCs' way (like a combat challenge or something deterimental to the overall goal) and then if they are successful removing the obstacle, they continue on the same challenge with the previous successes still tallied. If they fail again, they deal with another complication, maybe tougher this time. Continue on until they reach the goal.

Complexity is really about how long its going to take to get to the ultimate goal, and how many obstacles they will have to deal with on the way. DCs control how hard it is to reach the goal without complications.

 

[UngeheuerLich]

you could easily combine 2 challenges into one:

8 successes before 5 failures and make something happen when the 3rd failure is gathered.

Also you should see building skill challenges more as a guideline. And generally don´t allow aid another against dc 10, but instead aid another against an easy, medium or hard dc and maybe allow a different skill to aid someone in another (maybe a hard knowledge nature check to aid someone on a hard heal check)

 

[Skywalker]

Based on my admittedly very limited experience with skill challenges the new values really are too low. During our test week my players breezed through a Complexity 4 skill challenge without a single failure. And that was with adjusted DCs (the original DCs were in the range of 10 to 15 which meant some pcs would have had a higher skill bonus than the DC!).

I note that with the standard rules, the avergae result is normally just 1 failure for 4 to 12 success. So, getting no failures happens. This doesn't mean the mechanics don't work as intended.

They figured out almost immediately that they should try to let those who were trained in a certain skill make that check and used aid another to bump the results even higher.

OK this is your first issue. As with play outside of Skill Challenges, if the players are allowed to analysis and organise themselves in an optimal manner then you should expect them to do better than normal. However, a good Skill Challenge will include all kinds of pressures that make this kind of optimisation difficult. Time may be limited, narration may influence what skills are brought into play, perhaps the party's focus gets split. My recommendation is to try and keep the pressure on in Skill Challenges like a DM does in combat.

Imho, the new DCs only make sense if
a) aid another is not allowed
b) everyone is forced to participate every round
c) everyone is required to make a particular skill check at some point (i.e. forcing pcs to use untrained skills)

The DCs certainly work better with that kind of pressure in play. Ask yourself this: how often does aid another get used outside of Skill Challenges? If your answer is much less than in Skill Challenges then you aren't applying enough pressure.

One thing that particularly irked me was that the module we used (the Ashen Crown adventure for Eberron) suggested perception DCs that didn't make any sense, e.g. in one encounter it suggested a DC of 10 to detect a hidden gnoll hunter with a a Stealth modifier of +12 (I'm not 100% sure about the correct numbers, but you get my drift). Umm, yeah.

I wouldn't use a single Skill Challenge to criticise the mechanic as a whole.

Imho, having skill challenges with a 100% success chance is a waste of time. Where's the challenge in this? Since failed skill challenges should never be show stoppers, it doesn't matter if the pcs don't always succeed.

FWIW did you arrive at 100% mathematically or are you exaggerating the probability of success to support your conclusion?

As mentioned above, in a Complexity 4 Skill Challenge, an even chance of success overall requires an average of 83% chance of success on individual skill checks. Putting that into perspective, that's needing no more than a 4 or more on the D20. It feels like success is high, but you only need 3 failures to fail.

 

[Jhaelen]

First, thanks to Mustrum for the article overview!

I note that with the standard rules, the avergae result is normally just 1 failure for 4 to 12 success. So, getting no failures happens. This doesn't mean the mechanics don't work as intended.

Maybe, but the question is: What _is_ the intention?

Ask yourself this: how often does aid another get used outside of Skill Challenges? If your answer is much less than in Skill Challenges then you aren't applying enough pressure.

Nope, they've been using aid another all the time, in and outside of combats, when scouting and (re)searching stuff. I guess, they're just experienced and clever players

I wouldn't use a single Skill Challenge to criticise the mechanic as a whole.

The example I gave wasn't part of a skill challenge. It was a regular perception check at the start of one of the encounters. And it's just a single example of dozens of DCs that are completely off. The party's elven ranger had a spot modifier of +12. So, the passive perception was high enough to spot absolutely everything in the whole adventure with but a glance.

Part of the point I was trying to make here is this: The suggested DCs may work or even be required for skill challenges to work but it's a mistake to use the same DCs for regular skill checks. Regular skill checks (where a single success is sufficient) should have higher DCs to be meaningful.

FWIW did you arrive at 100% mathematically or are you exaggerating the probability of success to support your conclusion?

I'm not exaggerating. All but one pc had skill modifiers of +9 or higher. A skill DC of 10 (i.e. moderate) means auto success. Add the bonuses from (multiple) aid another actions and a 'hard' DC is also an auto-success.

As mentioned above, in a Complexity 4 Skill Challenge, an even chance of success overall requires an average of 83% chance of success on individual skill checks. Putting that into perspective, that's needing no more than a 4 or more on the D20. It feels like success is high, but you only need 3 failures to fail.

I know, and I've done the math. When the DMG was released there was a huge discussion which led to the birth of the Obsidian system as an alternative to the 'official' skill challenges.
That was when I created an excel sheet and a sample party to investigate the inner workings of skill challenges. I've adjusted the sheet to use the errataed version and it shows that low complexity skill challenges will auto-succeed at every level. Higher complexity skill challenges have a success rate of about 80% +/- 5%. This is assuming the pcs will always use their highest skills but doesn't include the effects of aid another (with aid another skill challenges of every complexity at every level will succeed 100% of the time).

Back then it was only an assumption, but now playtesting has shown it to be correct.

What I would like to see is a 50% success rate _at most_. And this requires the applications of restrictions and/or increasing the DCs.

Unfortunately, I'm also still waiting for my copy of the DMG2 which is supposed to contain an update on skill challenges.

 

[CapnZapp]

To me it seems that removing counting failures is key. Then you can have DCs that doesn't feel like you're being baby-sat by the game.

 

[Mustrum_Ridcully]

What I would like to see is a 50% success rate _at most_. And this requires the applications of restrictions and/or increasing the DCs.

I would like to note: Combats typically have a success rate of close to 100 %. If we count permanent character death and Total Party Kills as "failure" and nothing less. But we can live with that just fine because we know we worked for that. We spend the healing power at the right moment, the Wizard managed to push the enemies away before they could strike down the fighter.

But the standard skill challenges lack this type of details. I don't think it is the success rate that's the problem, but the "boringness" in which it is achieved.

Maybe it would be fine to raise all DCs by 5 points _if_ you also add an ability to undo failures (that doesn't cause new failures). So if the party face fails his Diplomacy Check, the Cleric can make an Insight check, negating the failure and allowing the face a new try.

The difficulty is, as always - we don't seem to have a good way to "codify" this without referring to specifics of a skill challenge. The above example needs a context to make sense? How can the Cleric use Insight to help now?
And also difficult - how can we avoid simple "button-pushing" - how does the player feel smart about using Insight in this situation?

I guess the DMG III or 5E needs a skill challenge maneuver/power system...

 

[Kestrel]

Adding a way to counteract failures is pretty easy. Just add a mechanic for it if the skill challenge calls for it.

I recently did one where each pc had a choice to help succeed or defend the party. If they chose to defend, they could make an attack roll (using a power if they chose) to remove a failure.

The big dissapointment for me in the challenge was the low dcs. They didn't really have to use the defend mechanic because they never failed a check (using the errata dcs). Thats what prompted the post.