D&D General Blending individual checks into group checks

[Tigris]

Uhm... I am not sure if it wasn't clear, but I meant 5e group checks, meaning that the group succeed if at least half characters succeed . These group checks do not increase the odds "a lot", in fact they generally decrease the odds compared to when a single PC attempts a task, because in most cases the PC attempting first is the one with the best chance. If the "group" in "group checks" is made by 2 characters, certainly the group checks increases the odds (two rolls, only one need to succeed). But as soon as there are at least 3 characters, things are not that obvious, even assuming all characters have the same chance... I can't do a detailed analysis, but my feeling is that if the individual task is easy (>50%) then the more characters try the better, while if it's hard (<50%) then the more character try the worse.

Well if I first can try alone, and then if I fail we can do a group check, yes this does increase the chance of success a lot compared to if 1 person failed no one else can try, which kind of is the situation without adding group checks.


Of course if you else would just allow everyone to try then that chance would be bigger.

 

[Koloth]

One thing to keep in mind with group checks is how many characters can reasonably attempt the task at the same time. For opening a stuck/locked/etc door, the limit may be how many can fit next to the door. For many doors(think the standard modern house/apt front door), 1 or 2 may be the max. A group knowledge check could well let the entire party contribute. Same for helping build a rope bridge across a chasm, assuming at least one character knows how to build the thing. Or the group succeeds at a group knowledge check on rope bridge building...

 

[billd91]

I think the nature of the individual task is getting kind of muddled here. There are some tasks where I think a group check would be appropriate, where treating things as a help action would be more appropriate, and some where neither really applies.

Take the door opening issue. For one thing, retries aren't a problem here - after all, even the strongest people in the group don't necessarily apply all their strength perfectly on one attempt. If a second person can fit, use the help action and give a subsequent attempt advantage. Group checks don't make much sense here since there's no reason to assume that half the group needs to succeed to open the difficult door.

Similarly, with Intelligence (lore of some sort) checks, I don't see much value in requiring a group check. People may know different things and from different sources completely independently. Those should be independent and not subject to majority success requirements to be useful, UNLESS the PCs specifically say they're consulting together for a moment - then I allow help/advantage.

As I see them, group checks really are for situations where multiple PCs have to do something where one's successes can help someone else succeed even if they're not doing that well on their own - climbing, sneaking, convincing, bluffing, etc.

 

[Thommy H-H]

The theoretical group check for breaking down a door doesn't mean everyone has to be able to shoulder the door down simultaneously. It means the Fighter who fluked a failure on their roll softened it up for the others who tried afterwards. Everyone contributed, and that's abstracted by a group check.

 

[Lakesidefantasy]

I'm interested in reining-in the Chimera. I see this sometimes with skills like Deception and Persuasion. For instance when one character is quite clearly having a conversation with a non player character and we determine that Deception is being employed so we call for a check. They fail and the next Player declares they would like to have go at it. Fine, but it's a Group Check now and we have one failure.

Sometimes a Player will even jump in and claim they have the better Persuasion skill so they should be rolling the check. That when I ask them to please let the other player have their time in the limelight. Otherwise our Chimera wanders through the adventure with all the best skills.

I like a blended group check for these types of situations, but I agree, a stuck door should not be an impediment if time is not a pressure.

Maybe some group checks can be blended whereas others are obligate. For instance getting the party around to lift a heavy stone would be an obligate group check that they all roll simultaneously.

 

[saviirkad]

I generally take a stance of no more than two checks for the purpose of game flow and moving forward, with the exception of when the first two people to jump at a thing leaves out the character most suited for the task.

I want the game to flow and not get bogged down, but I also don't want the player whose character was made for this task to be frustrated over it.

If the Goliath Barbarian and Dragonborn Fighter both try to shift this book case and it doesn't work, the rogue or monk are unlikely to think they can make a difference.

Now if the rogue and monk were the first to try, maybe because they were in the room alone or the Goliath was busy, it'd make sense after their failure that they'd go get their big and burly friend.

But turning it into a group check makes sense too, I can use that. I'm wondering if we've been house ruling group checks though, because we don't do "over half succeed," we average everyone's results. In my experience this tends to be in the group's favor, as the extreme most likely to affect the average is when the expertise players also get a high roll.

 

[Charlaquin]

Well I can see how different people might want to try, but in the end this just makes every check a lot easier, and rewards the "not accepting negative outcomes" player behaviour.

This isn’t actually true. The way the math on group checks work is that the more characters are involved, the less likely they are to succeed. An even number of characters is always better than an odd number, but it’s always worse than a smaller even number.

 

[Charlaquin]

Uhm... I am not sure if it wasn't clear, but I meant 5e group checks, meaning that the group succeed if at least half characters succeed . These group checks do not increase the odds "a lot", in fact they generally decrease the odds compared to when a single PC attempts a task, because in most cases the PC attempting first is the one with the best chance. If the "group" in "group checks" is made by 2 characters, certainly the group checks increases the odds (two rolls, only one need to succeed). But as soon as there are at least 3 characters, things are not that obvious, even assuming all characters have the same chance... I can't do a detailed analysis, but my feeling is that if the individual task is easy (>50%) then the more characters try the better, while if it's hard (<50%) then the more character try the worse.

There was another thread about this idea once before where the OP did the analysis, and the conclusion is what I’ve been saying, that the best number of characters to have on a group check is the smallest even number possible - the more characters involved, the lower the total odds, but even numbers are always better than the next-lowest odd number and worse than the next-lowest even number. I’ll see if I can find the thread and link it.

 

[Charlaquin]

Well if I first can try alone, and then if I fail we can do a group check, yes this does increase the chance of success a lot compared to if 1 person failed no one else can try, which kind of is the situation without adding group checks.

It’s not a new group check, it’s making your failure retroactively count as your contribution to the group check. You don’t get to retry, but if they succeed, then at least half of you have succeeded, so it becomes a successful group check. This does increase the odds of success, by slightly less than granting advantage would increase them by.

Of course if you else would just allow everyone to try then that chance would be bigger.

Thats incorrect. The more characters involved in a group check, the less likely it is for at least half of them to succeed. Adding another character only improves the group’s chances if doing so brings the total to an even number; if makes them worse if it brings it to an odd number. And a larger even number is always worse than a smaller even number.

 

[saviirkad]

There was another thread about this idea once before where the OP did the analysis, and the conclusion is what I’ve been saying, that the best number of characters to have on a group check is the smallest even number possible - the more characters involved, the lower the total odds, but even numbers are always better than the next-lowest odd number and worse than the next-lowest even number. I’ll see if I can find the thread and link it.

Thanks for searching, Im very curious about this math. Like is it a bell curve, what's the optimal number of participants?