Collaboration is a basic facet of human connection. Sometimes you roll for it, sometimes it costs you a resource, sometimes it’s an automatic boon. But despite all its variation, I’ve never met a Help action that I particularly liked.
Almost every TTRPG I’ve played distinguishes between the helper and the helped. Occasionally, these roles follow the fiction. One PC does the talking while another looms menacingly; one performs the surgery while another holds the tools; one’s getting their ass kicked by the monster until another steps in with a well-timed brick.
But they’re often nonsensical. When two PCs kick down a door together, who is “doing” the kicking and who is “helping” with it? When the whole party talks over each other in an attempt to convince the monarch that their advisor can’t be trusted, who is “leading” the persuasion check? The obvious answer is to figure out who has the highest bonus and let that person make the primary roll. This is the worst type of metagaming: completely above the table, breaking immersion for the sake of an abstract rule that means nothing to the fiction.
My ideal Help action therefore removes the helper/helped distinction. When multiple people collaborate on the same task, their rolls are treated the same. Characters can make different types of rolls if they’re performing different actions—one holds the reins while another telepathically soothes the horse—but neither takes precedence.
Here are some more important features of the ideal Help:
Generally, the more people who work together, the better the chances of a successful outcome.
Attempted help can sometimes fail. Assistance from someone who doesn’t know what they’re doing is bad.
With these guidelines in mind, I’ve started using the following system whenever two or more characters pool their strengths and succeed or fail as one.
First, everyone who’s contributing to a task makes individual rolls. The roll each person makes is based on the nature of their contribution, which might (but doesn’t need to) vary from person to person. Bonuses and penalties to the rolls might also vary. Either way, these individual rolls are binary: either a character succeeds on their roll or doesn’t.
Then count the number of characters with successful rolls to determine whether the group effort succeeds.
Everyone rolls well = Critical success
At least two people roll well, but not everyone = Standard success
Exactly one person rolls well = Mixed success
No one rolls well = Failure
(Sometimes more people working together should always increase the risk of failure, such as during a group stealth check. In these cases, the resolution rules are changed to: At least two people roll poorly = failure; exactly one person rolls poorly = mixed success; everyone rolls well = success; everyone rolls exceptionally well1 = critical success.)
I haven’t gotten to play with it too much, but it’s been fun so far. Most noticeably, players don’t feel like they’re leaving money on the table by failing to declare Help on every single roll. (Looking at you, 5e.)
One quirk is that a two-person group can never roll a standard success—only a failure, mixed success, or crit. I’m not sure how to feel about this. I do like the implication that, when possible, two competent people pooling their strengths is the optimal way to get shit done.
Another quirk is that a thousand people with even the tiniest die-sized chance of success will, like monkeys at typewriters, inevitably do the trick. But my games tend to have less than a thousand players, and I’m not enough of a simulationist to care. It would bother me if a mere six PCs could consistently succeed together where they’re all supposed to suck; this is something to keep at eye on, though I think it’s more likely to manifest as a numbers-need-tweaking problem than a back-to-the-drawing-board one.
And of course this would need to be adapted for games without four tiers of outcomes. For example, in a binary success/fail system, you could say that at least 50% of individual rolls need to succeed for the overall task to go well—though come to think of it, this creates an even weirder inflection point at group size 2. I guess that’s why 5e’s group checks are distinct from its Help.
What this means varies from system to system—something like beating the DC by a certain number, or rolling well on more dice than necessary.