PDA

View Full Version : Dissecting Skill Challenges: Revisions, errata, and supposed mechanical failures?



RedWarlock
2019-01-25, 12:30 AM
To isolate a topic I'm curious about from the Fantasy Heart Breakers (http://www.giantitp.com/forums/showthread.php?577120-Fantasy-Heart-Breakers-mechanics-that-break-the-heart) thread, there was a lot of (rather one-sided) discussion of how 4E's Skill Challenges were considered flawed or somehow a mechanical failure.

There's an article that was linked that was mentioned a few times in the discussion, either as 'obvious proof that they were broken', or a 'tautological declaration without actual factual basis' (these are my own indirect summaries of the arguments about it, not quotes). I haven't been able to find the link in a quick search, but given the discourse, I'm not sure if the article is actually useful to reference or not.

Now, obviously, they DID go through a few mechanical revisions during 4E's lifetime, so something was definitely up, but what? Can anybody break down the different versions of the mechanic, what changed from iteration to iteration, and how they succeeded or failed at providing the encounter experience they were intended to be?

And on a more homebrew-y note, does anyone have any suggestions for fixes which can help the skill challenge re-target back to that goal?

ThePurple
2019-01-25, 01:34 AM
Now, obviously, they DID go through a few mechanical revisions during 4E's lifetime, so something was definitely up, but what? Can anybody break down the different versions of the mechanic, what changed from iteration to iteration, and how they succeeded or failed at providing the encounter experience they were intended to be?

The initial skill challenge system, as presented in the DMG, was *woefully* bad. Because you needed to get a high number of successes before a tiny number (3? I think? It's been a while since I read them) of failures with no other considerations, it encouraged the party to figure out a single PC with the highest chance of success and then have them make the roll each round with the rest of the party either sitting back or rolling to assist. Failure occurring because of accrual of failures discouraged players from participating.

Another problem was how Skill Challenges were a way for players to avoid tactical combat, for which there were a lot of rules and fun, while being forced to use the rather blase Skill system in 4e, for which there wasn't really much rules or fun. The "fun" element encouraged players to straight up skip many skill challenges and just have a fight since that's what many players were there for, especially since many skill challenges were written with only a tiny subset of skills as valid uses and, due to 3.X's tendency to encourage heist-movie style hyperspecialization (e.g. the party face is the only one with high CHA and also trains in the social skills while the high INT char trains all of the knowledge skills), many players found themselves incapable of actually contributing to many skill challenges (e.g. the fighter couldn't really do anything to further the research skill challenge in the Great Library of Thuum).

There was also the issue that failing a skill challenge just 1 success shy was identical to failing a skill challenge with no successes at all: the results were binary with no real cost along the way.

There were many iterations along this line, none of which were well collated. The Obsidian Skill Challenge system, which was designed to act as a way for Skill Challenges to augment combat, fixed the first problem by requiring a given number of successes within a set period of time (3 turn limitation) and also added a bunch of tactical options, but it did away with the facet of Skill Challenges that allowed them to act as noncombat-centric risk/reward equivalents of combat encounters (consuming resources in exchange for progress and xp).

Later published Skill Challenges tried to circumvent the problem of many players feeling useless in specific skill challenges due to the limitations of the skill system tried to create extremely creative uses for the most ubiquitous skills, like Religion and Stealth, which itself created the problem that any divine character's go-to Skill Challenge action was "pray for success" and stealthy characters just started "doing it stealthily" even if it made no real sense (e.g. in an SC to climb a cliff, the cleric without athletics would just constantly pray for success while the rogue without athletics would "climb stealthily") or that Athletics suddenly became a social skill because the fighter was allowed to just start flexing for whatever reason. In allowing hyperspecialized characters to become more flexible in Skill Challenges, it slightly shifted their hyperspecialization (to the most generally useful skills) and broke credulity because, suddenly, parties could negotiate with a Duke by having their fighter flex at him via Athletics.


And on a more homebrew-y note, does anyone have any suggestions for fixes which can help the skill challenge re-target back to that goal?

I've loved Skill Challenges ever since their first inception, recognizing their limitations and trying to work around them.

The biggest problem, of course, is the Skill System in 4e itself. Training confers *way* too big of a bonus to success and the hyperspecialized nature of ability scores (you only ever get 2-3 good ones and they only get better as you get higher in level) means that the gulf between low (no stat, no training), moderate (1 of stat or training), and high (both of stat and training) continually increases, reducing the ability for characters to operate outside of their hyperspecialized choices (and making it so that, at high levels, it's pointless to even *try* unless you're both trained and have the stat for it). The math scales way too quickly.

On top of that, the whole notion of Skill Challenges operating on a bunch more binary scale perturbs me from a balance sense. All skill checks amount to either the same success or loss; it's as if, in combat, every attack did the same amount of damage. Everything happens in big chunks, such that you lose the subtlety that makes things interesting.

My solution for this was to completely revamp the Skill System so that it more closely mimics the combat system of 4e while encouraging players to diversify their skill training in order to ensure that they're functional in a large number of different situations to go with 4e's "everyone acts together" design.

First off, all skills were divorced from the ability scores with the GM determining what ability score would be used with the relevant skill (Athletics, Acrobatics, and Endurance are cut down to just be "Athletics"; what used to be Acrobatics would now just be Athletics + DEX) which allows for a lot more versatility in roleplay and action (talking about sports facts would be Athletics + INT).

Secondly, a "damage die" was added (referred to as "success die") to determine how successful the given act would be, if the skill check succeeds, so that there's some variability in success.

Skill Training, rather than adding a +5 to success, now adds a +2 to the skill check (mimicking the proficiency bonus to attack rolls) and increases the untrained success die of d4 to a d8. As such, a trained practitioner is both more likely to succeed and will contribute more to success while still giving an untrained PC a chance to contribute *something*. I also have basically made it so that players are basically forced to get at least 2 skills from each of the major skill suites (physical, social, knowledge) rather than focusing all of their training on a single skill suite (which makes sense since, even in a social encounter, you only get 1 action per round).

A slew of other numerical changes were also made (primarily making it so that gear contributes to Skill roles to match the progress made for attack rolls and damage rolls) in order to get the numbers to follow the same progression that combat numbers do.

Skill Challenges themselves now follow the same design conceit as combat encounters. You set a given soak value (required success points to end the skill challenge) and assign threats that deal damage. Each turn, the players describe an action they're doing that either advances progress (generating success against the necessary value to end the skill challenge) or protects from or diminishes the threats (reducing the damage that the threats pose). When the players' turn is over, the threats act, either forcing players to roll to avoid damage or attacking the players themselves.

I'm still messing about with the exact numbers for threat and success (there's more variability in the practically applied numbers because of the separation of the ability scores from specific skills as well as the increase in non-trained actions, which are comparable to untrained unarmed attacks, compared to combat) but all of the playtesting I've done gotten rave reviews. My players find it to be just as fun as combat, especially since the variety of potential skill/ability score combinations that can be used effectively gives them a similar array of options that combat gives them.

I'd like to look into creating some powers that can be used in SCs, but I'm afraid that would bog things down too much.

Kurald Galain
2019-01-25, 02:43 AM
The difficulty with discussing skill challenges is this: what, exactly, IS a skill challenge?

Pretty much everybody agrees that the rules as printed in the DMG1 are pretty horrible. But from there, some people made their own changes and houserules and called it an SC (because it was clearly inspired by the DMG1 rules) whereas other people made similar changes and houserules and called it NOT an SC (because it was clearly different from the DMG1 rules).

So it is easily possible that Player A will point to the DMG1 rules and say "see, this is horrible!" whereas Player B will point to whatever his own group does and say "see, this is great!" (and unless B is the DM, he might not even realize the difference). Both are true statements but this is not a productive basis for discussion.

Kurald Galain
2019-01-25, 06:52 AM
Now, obviously, they DID go through a few mechanical revisions during 4E's lifetime, so something was definitely up, but what?

Some things that are (or have been) up,

The DMG1 states that if the players try something the DM didn't think of, it automatically gets a harder DC. There was player outcry over how this discourages creativity, and it was errata'ed out. But the Rules Compendium brings it back, stating that creative solutions should not contribute successes to the SC, but give a lesser effect.
The difference between an easy and a hard DC is commonly less than the difference between a character's best skill and worst skill. Mathematically, this makes it a good tactic to use your highest skill over and over again, regardless of circumstances.
RulCom states that failing a SC means the PCs still succeed at their task, and get full XP, but they lose a healing surge or get a penalty in the next combat.
If a player wants to use an item or power in an SC (instead of a skill) then neither DMG1 nor RulCom give any guidance whatsoever. As the result, numerous DMs rule that either this is impossible (because you're in an SC now) or that after the item/power is used up, the player must still roll a skill check.
Many SCs in official WOTC adventures (including LFR) are bizarre, contradictory, or downright awful; and bear in mind that this is where many novice DMs get their ideas.


There are other examples, but that's the gist of it. In general, the more closely you follow the printed rulebooks, the worse your SCs become.

darkbard
2019-01-25, 07:20 AM
Here is an explanation of how one might run SCs, written by a friend. Note that the way we play, there are no set primary skills for use in an SC; a PC can declare whatever action they declare, and the GM decides which Skill is appropriate, and this includes deciding that in any given circumstance a different ability modifier applies. So note that while this is fairly close to the Rules Compendium iteration, there is some house-ruley stuff involved:


Skill Challenges: What are Primary Checks?
These are zoomed in moments of the greater conflict where the PCs are facing significant adversity and must make a decision under stress, for better or for worse. Something decisive is going to happen to change the situation as a result of their efforts. Players might leverage their Skills (Arcana, Athletics, Nature, et al), Healing Surges or applicable Powers for bonuses, applicable Dailies to step the DC down one (or succeed at the Medium DC automatically for at least one success towards the challenge), coin (often 10 % of an at-level item for an auto-success), or Rituals (counting for at least one auto-success towards the challenge).

A second usage of the same Primary Skill in a challenge should step up the DC (eg from Medium to Hard) or increase it by + 5 if the DC is Hard.

What are Secondary Checks?
These are tighter zoom still than Primary Checks. These are effectively augments to Primary Check efforts (+2 or - 2 success or failure). Players can aid each other or they can augment their own subsequent checks. This is most often accomplished through Skills (though this can also be achieved by the use of an Encounter Power that is not directly related to the challenge).

The number of Secondary Checks available customarily equals that of the complexity of the Skill Challenge (eg 2 for a complexity 2 challenge).

What are Advantages?

These provide a modicum of authorship to players. Thematically, these would be things like luck, digging down deep, when powerful emotions are in play (eg relationships), divine sponsorship, etc. Players gain 2 of these for complexity 3 challenges, 4 for C4 challenges, and 6 for c5 challenges. Mechanically, they can do things such as (1) stepping a DC down (eg from Hard to Medium), (2) allowing a second use of a Skill for a Primary Check at the same DC, (3) cancelling out an accrued Failure, (4) allowing a reroll (this can also be achieved by spending an Action Point).

What are the consequences of Check Failures?

• The character who failed the check loses a healing surge or (in a combat context) takes damage.
• The characters must spend time or money making up for the failure.
• For the rest of the challenge, no character can achieve a success using the same skill that was
used for the failed check.
• lf the challenge takes place in a combat situation, the character who failed the check is dazed or
even stunned until the end of his or her next turn. Or, an opponent is angered and gains a +2 bonus to its next attack roll.


Note that we both play "ficiton first" games; i.e., the PC describes what is happening in the fiction first, and then a mechanical framework for the resolution is considered. This reduces the oft-decried spamming of best Skills but still retains PC trope focus (a character who is steeped in magical lore will turn to Arcana and History skills most frequently, and so on).

Yakk
2019-01-25, 09:51 AM
I find Skill Challenges are best thought of as a response to problems in 3e D&D.

In 3e D&D, how did you steal the crown jewels?

Well, you make a plan: sneek into the castle, disable any guards you could not avoid, unlock the vault, disable any traps, grab the crown, then get out.

So on the way in you are sneaking. How many stealth checks do you make? One every round? What is the DC? After that, there are a bunch of tasks. What happens when you fail a task?

DM's would wing it, and the result was that the difficulty of doing this would vary to an infinite degree based on the DM's choice while winging it. And lacking a serious statistical background, the DM's choices *would have no direct relation to how hard the DM thought it should be*.

One DM might think it is really hard, and make certain decisions about DCs and frequency of rolls and result of failure; another might think it is really easy, and make other decisions. And the two DMs might get the exact opposite result from what they want.

Skill Challenges attempted to address this. They basically stated that the DM should decide how hard the task should be (give it a complexity) for characters of the party's level. This then gave the DM a framework; the number of successes, number of failures, and the DCs of each of the sub-checks.

This lets the DM *intentionally* decide if something is nearly impossible or relatively easy, instead of the difficulty emerging from the mechanics. The mechancies where designed to support the intention.

They then tried to write this up, and failed miserably. They had people write up hard-coded skill challenges in adventures based off those systems, and failed miserably.

Revisions where done that cleaned up the math a bit, but still, the system resulting was pretty lacking.

Now, at the same time, it *could* work with a party that was invested in participation, unbothered by failure, and did not respond to any perverse incentives the system has, together with a DM who is both willing and able to improvise whenever the players do something creative.

But that is damning a system with faint praise.

ThePurple
2019-01-25, 03:40 PM
They then tried to write this up, and failed miserably.

Honestly, I have to wonder how much the Skill Challenges (and Skill system in general) got tested while in development. The rules as written seem like a poorly constructed first draft that doesn't mesh well with the robustness that the combat portion of the system had, even from release (that was later improved).

Much of the problem with the Skill Challenge system, imo, comes from a flawed desire to fix Skill Challenges via the math rather than overhauling it and creating a completely new system, which is why pretty much everyone that still uses SCs has created their own system. 4e was *really good* about killing sacred cows and turning them into delicious divinity burgers, but the SC system, as written, was one that it created and weighed it down heavily.

Yakk
2019-01-25, 05:20 PM
No, I'd argue that Skill Challenges *as written* work fine in a non-adversarial environment.

I mean, sometimes lower complexity is harder than higher complexity, and players can min-max character actions to create bad results, and the DCs are poorly tuned, and using it doesn't generate guaranteed good results.

But if you just use it as a narrative pacing mechanic -- complexity is just about how much table-time you need, and the idea that no single success should solve the problem (but rather reveal more complications), you have a working system. The "system" is really just smoke-and-mirrors tho, and it depends on the DM already telling an engaging story.

They almost certainly used it in play, and in the hands of a skilled DM (who probably didn't need it to start with), it works.

Problem is, a good skill challenge system would instead help medocre and poor DMs generate *better results*. And it doesn't do that.

---

An untested revamping that really doesn't kill any sacred cows:

Complexity: Unchanged. 2C+1 successes to pass, 3 "failures" to stop. (Note that not all check failures are "failures", nor are all check-successes a "success")

DCs: The most recent DC table isn't horrid. The level used is the level of the SC, not the party, however; so you can give players a Complexity 5 Level 35 challenge.

Now the changes.

Time Scale: Challenges have a time scale. The time scale is how long a "round" is. A "round" is the unit of time that each character gets to try one thing and roll a skill (or use an ability).

Threat: For something to be a skill challenge, there must be a threat. The threat is the reason why "ignore it" or "let the expert handle it" isn't an optimal solution. Sometimes the threat doesn't start until the characters engage with the skill challenge.

Threats should have narrative backing; for example, the environment being harsh justifies Endurance. If targetted and group's choice, maybe someone has to ground magical energy.

Complexity 3+ challenges should have an extra threat. Paragon challenges should have an extra threat. Epic challenges should have 2 extra threats.

Tick-tock Threat: After each round, a failure is accumulated. Should only be picked once.

Endurance: After each round, each character loses a healing surge. Characters cannot regain this healing surge while the challenge is ongoing, and possibly for some period of time afterwards. If they have no healing surges, their max HP is reduced to 3/4, 1/2, 1/4 and then they enter a coma.

Alternatively, one character somehow of the party's choice suffers (complexity) healing surge loss, or one character not controlled by the party suffers (complexity/2, round up) healing surges lost.

Hitting back: Something attacks the party at the end of each round. It rolls Level+3 vs non-AC or Level+5 vs AC, and deals and average of 2*Level+5 damage (single target), or Level+3 (attacking entire party). If picked more than once, increase damage by (Level+3) and accuracy by 1.

Plot: Describe the result of the challenge in 6 different levels. It starts at 6. Each round, roll 2d6 and discard the lowest; if lower than the current challenge result, that is the new challenge result.

Taxes: In order to engage in the challenge, resources between 1x and 5x the value of an even-level magic item must be risked (maybe you need to buy a boat? Or it requires risking your magic sword. Or, you are doing a skill challenge to buy something; in which case, the "damage" is done to increase the price of the thing you are looking for). Each round, damage equal to 1/5 of an even-level magic item is done to that resource.

Hazard: May only be picked as a second threat. Every failure triggers a different Threat (except tick-tock). So, for example, every failure costs the party an "endurance" event, or a "hitting back" event.

DMs are encouraged to think up other Threats. Skill Challenges without Threats are things that are best handled by a single expert, or a series of experts, with non-experts sitting out, and should not use the Skill Challenge system.

Player Actions: Players should describe how they intend to solve a problem. Some things they do may simply happen without a check. If so, ask what they are doing next. At some point ask for a skill check for some task they are doing, or have done, to see how it impacts the challenge.

The result of such a check should have immediate narrative impact, as well as overall impact on the skill challenge success track. Something interesting should happen on both success and failure.

So if a player is asking to remember a historical fact, failure is boring; they won't remember anything, and sit there looking dumb. Instead, you could say "yes", and give them a fact; the result of the check determines if using what they think they know is useful, useless or harmful, wrong or right.

Variant Rolls and Sacrifice: The result of a check need not be "1 success" or "1 failure". And sometimes players will want to do things that involve sacrifice.

A check might have a very high DC, and generate 2 successes instead of 1; or it might have a still higher than moderate DC, and generate 2 successes on success, but also generate a Hazard on failure. This should make narrative sense.

You can sacrifice a valuable resource. Burning an appropriate daily attack power could grant a success after successfully "hitting" an appropriate defence, or boost the result of following check to 2 successes, or even nullify a failure. If your challenge is has fast rounds, even encounter resources can generate benefits. Similarly, challenges can give opportunities to "eat hazards" in exchange for benefits (lose healing surges, take an attack, etc).

When putting out a fire, carrying the bucket into the burning building in order to deliver the water to a more important spot might risk a Level+3 attack vs Fortitude or take 5+2*Level damage, but failure doesn't add to your failure track.

Group Checks: Group checks should use a moderate DC, and admit 3+ different possible skills, or 2 possible skills where one of them can use the Easy DC. Any number of those skills may be "mandatory", in that at least 1 person making the group check must use it.

Group checks succeed if at least (Complexity/2, round up) successes are obtained.

---

Basically, adding Threat and the skill challenge system is viable as-is. "Aid Another", "Let one person do it" -- they all come out in the wash. Waste your turn, and the threat hurts more.

Also note I didn't mention "make a fight bigger". More challenging fights is often fun, and we don't want to encourage people to throw skill challenges. (And yes, taking damage also makes fights harder, but in a more obviously negative way, so less likely to be embraced).

ScrivenerofDoom
2019-01-25, 11:03 PM
I recommend two things:
1. Check out anything written by Pemerton on ENWorld about skill challenges. He managed to capture the drama the WotC designers never could.
2. Ignore Mike Mearls. The guy wasn't even running 4E when he was the lead on Essentials and the mechanically unsound Heroes of Shadow and it shows.