Most rolls in D&D are made by a single character. Only the cleric makes a saving throw when she looks at a medusa. Only the rogue tries to disarm a trap. Only the fighter swings his sword at the goblin. The party isn’t Voltron (note to self: make rules for a Voltron session). But sometimes there are opportunities for a check from the entire group because everybody is engaged in the same activity, like making Constitution checks as the party travels across the desert or dodging the same cone of cold. It’s a moment of camaraderie as everybody faces one challenge together, and it’s a chance for the most skilled party members to shine as they shrug off effects that would destroy a normal person and/or bard.
The group skill check, however, is a special sort of problem. It either goes very well or very poorly, almost every time, and that’s because the rules have been doing it wrong.
A group skill check occurs when the entire party makes the same skill check at the same time for the same reason. In D&D skill checks either result in a success or a failure; partial successes are a great idea but are technically a house rule. That’s all well and good. The problem lies with the definition of “success” and “failure”. In my experience, there are only two types of group skill checks:
Type A: A failure for one is a failure for all. If any person does not succeed at the check, the entire party is penalized for it. The archetypal example is a Stealth check. The rogue may be good at it, and the ranger may be good at it, and maybe the wizard rolled high today. But the dwarven paladin in full plate with a Dexterity of 8 drags everybody down, and asking her to sneak is like trying to smuggle a collection of cuckoo clocks into a library. It would take a miracle for her to sneak around, and if any enemies hear her, they know intruders are nearby no matter how sneaky the other characters are. They’re going to raise an alarm, and the only advantage the players have is that the enemies don’t know how many people they’re ambushing.
Avoiding this problem is usually a matter of mitigating the check entirely. The barbarian ties a rope to the sorcerer, swims across a river, and pulls the sorcerer along because asking her to roll an Athletics check is a recipe for disaster. The monk navigates the chasms in combat, because the cleric is never making the Acrobatics check he needs to reach the other side of the battlefield. Unless the entire party is built for a specific task, forcing everybody to make a type A check is almost a guaranteed failure.
Type B: A success for one is a success for all. If any person succeeds at the check, the entire party is rewarded for it. The archetypal example is a Knowledge check. The bard has a high Intelligence, good proficiencies, and class features that buff his rolls further. If the party needs to make a History check, he has all the tools to provide an answer no matter how ill-trained the rest of the party is. Only the worst roll would foil him, and even a mediocre check on his part towers over anyone else’s contribution. The party is going to get the information they need from him, and everybody else can only offer tidbits and trivia.
Parties do this by design. It’s the logical extension of role specialization; there’s no need to have two characters who are good at History or Arcana or Nature when one will suffice. Putting the onus for a given skill on one character frees everybody else to specialize elsewhere. But it means there’s largely no purpose in using a group skill check in the first place, and it encourages players to avoid certain skills because they already have somebody who’s very good at them, even if training in those skills is logical, wise, or in-character. Examples here include Perception (only one person needs to hear the kobolds approaching to warn everyone else), Insight (only one person needs to know the duke is lying), and Intimidation (only one person needs to interrogate said duke). This is one character’s show, and pretending otherwise is disingenuous. Unless nobody in the party is good at a specific task, a type B check is almost a guaranteed success.
Both of these are the success/failure dichotomy writ large. There’s no partial success or failure for a group skill check. Either it’s a type A check, and the only way to succeed is for everybody to succeed, or it’s a type B check, and the only way to fail is for everybody to fail. Often type A checks are physical and type B checks are mental but that’s not always true; one failure on an Animal Handling check can spook the party’s horses when kobolds attack, and one success on a Sleight of Hand check can let a rogue smuggle enough weapons for everybody into the royal ball. If a DM asks the entire party to make a check and only cares about the highest or the lowest number, every other roll is kind of a waste of time and effort.
I’ve been playing around with an alternate way to run checks like this, and so far I’ve been very happy with it. It’s buried in the 5E Player’s Handbook, in a section that’s easy to miss if you think you already know how the game works:
To make a group ability check, everyone in the group makes the ability check. If at least half the group succeeds, the whole group succeeds. Otherwise, the group fails.
This suddenly makes things very interesting. It means that if the majority of the party succeeds on a Stealth check, they provide enough advice or cover for the paladin to sneak by regardless of her actual result. If the majority of the party fails a Knowledge check, they provide enough bad information (and they’re so certain about it) that the bard has to second-guess what he knows. It’s definitely better for tasks with lower DCs, but it means there’s some purpose to rolling checks that used to be foregone conclusions. Essentially, the party is making a single skill check rather than each character making it individually.
It’s such a quick, easy, elegant solution, it’s no surprise the rules manage to miss the boat in the next paragraph:
Group checks don’t come up very often, and they’re most useful when all the characters succeed or fail as a group.
We’re back to the success/failure dichotomy. Yes, the party can either succeed or fail. But those aren’t the only options. What about just barely failing, or blowing it out of the water? Three characters succeeding should not give the same result as all five characters succeeding. There should be some synergy, and total success should give some bonus above and beyond a minor success, like combining History checks to get a complete view of a past event from several sources and angles. Similarly, three characters failing should put the party in a better situation than all five characters failing. On a minor Stealth failure, the party might knock over a bucket and give away their position, but a total failure might have them bumbling into a meeting or tripping over a candle that catches a tapestry on fire. There’s a brilliant opportunity to talk about partial successes and failures, and the rules see it coming and Matrix dodge out of the way.
We’ve been using the partial success/failure version in Under the Stars in a way the rules clearly did not intend, as a sort of mini-skill challenge. The party decides they want to accomplish some task, and each character decides how he or she wants to contribute. Unlike a normal group skill check, these checks do not have to use the same skill. In fact, I encourage them to be different. For example, the party wants to build some defensive spikes around the Outpost. The barbarian digs holes for them with a Strength check, the ranger assists her, the druid makes the spikes themselves with carpentry, and the bard lays out where everything should go using Intelligence (History) to represent his knowledge of military tactics. A failure on one check doesn’t ruin the entire endeavor. Maybe the holes aren’t deep enough or the spikes are shoddy or they’re placed in a bad position, but they’re still effective, and as more characters succeed on their checks, the more effective the spikes become.
This is what a party is supposed to do. Everybody has his or her speciality and they all work together on a common goal. Normally group-wide checks mishandle it by forcing everybody to be good at the same things or feigning interest in anybody but the best person for the job. The group check rules in 5E make everybody a meaningful participant in simple, party-wide tasks. The mixed group check we’ve been doing is the logical extension, letting everybody contribute even for difficult tasks, and it encourages players to think creatively about how they can help instead of waiting for the DM to tell them what needs to happen. Few long-term problems can be solved with a single skill. D&D just needs a framework to adjudicate using several at once.