There is a section in the playerd handbook that mentions about the possibility of doing group checks when the group is trying to accomplish somethinhg togheter. It says that as long as half of the group succeds, then the complete group is succesful.
Last time we player, the plsyers used this to try to see if they could hife and surprise a party in an encounter. I really do not think this was a good application of the rule, but I let it go to continue with the gamr dynamics and flow. I thunk that each player individually needs to roll stealth and compare against wisdom check or at least compare against passive perception.
So the questions here are:
- when do group checks apply?
- does it apply in the example I just mentioned (trying to hide from a group they are attacking)?
Thanks
The example you give is a textbook example of when you should use group checks.
My rule of thumb is that, whenever rolling five times would give ridiculous odds on the task, because only one success or failure is needed - such as five people each needing to stealth past somewhere, or five people making perception checks to notice the same thing - you probably want to use group checks.
Atari2600 wrote:
I use group checks usually for stealth, charisma checks to blend in, strength checks for lifting (as long it's too heavy or unhandy for a single character to lift), acrobatics/athletics checks when they're tied together by a rope, etc.
Even if some of the examples doesn't some logical at first, you have to remember, that even stealthiness can be more efficient if you're working together. Like the ranger, who gives the fighter a sign, that he's about to step on a branch. And the stealth exampel prevents party-splitting while planning an ambush a viable option, which enhances the fun imo.
http://community.wizards.com/forum/rules-questions/threads/4162201
Or don't use group checks. It's a disfunctional mechanic.
I think the group check is meant more for a group of the players are all attempting to hold a door close against a foe or greoup trying to push the door in. That would be used for a group check. At least that is what I have used it for.
I don't think it's a dysfunctional mechanic - it's just not likely to see a ton of use. As I mentioned in the thread I linked, it will probably be very situational or apply only to long-term abstract challenges where the success of one PC isn't enough to carry the others. Context will tell if it's an appropriate mechanic to use.
iserith wrote:
Gnarl wrote:
iserith wrote:
FrogReaver wrote:
iserith wrote:
Gnarl wrote:
FrogReaver wrote:
FrogReaver wrote:
Gnarl wrote:
iserith wrote:
Gnarl wrote:
Gnarl wrote:
iserith wrote:
Gnarl wrote:
*http://stattrek.com/online-calculator/binomial.aspx* binomial calculator found here
At 30% success rate per player. At 4 Players
Chance for at least 1 success without group checks. (I used a binomial calculator with success chance set to .3 and number of trials as 4 and wanted 1 or more successes)
75.99% chance of at least 1 success
Chance for success using in book group check rules. (binomial calculator can be used here too. However I want chances of 2 or more successes.)
34.83% chance for success using in book group rules
------------------------------------------------------------------------------
Let's look at 30% success rate with 5 players.
Chance for at least 1 success without group checks. (I used a binomial calculator with success chance set to .3 and number of trials as 5 and wanted 1 or more successes)
83.193% chance of at least 1 success
Chance for success using in book group check rules. (binomial calculator can be used here too. However I want chances of 3 or more successes.)
16.308%
------------------------------------------------------------------------------
Now lets loot at 50% and 4 players
Chance for at least 1 success without group checks. (I used a binomial calculator with success chance set to .5 and number of trials as 4 and wanted 1 or more successes)
93.75%
Chance for success using in book group check rules. (binomial calculator can be used here too. However I want chances of 3 or more successes.)
68.75%
--------------------------------------------------------------------
Now lets look at 50% and 5 players
Chance for at least 1 success without group checks. (I used a binomial calculator with success chance set to .5 and number of trials as 5 and wanted 1 or more successes)
96.875%
Chance for success using in book group check rules. (binomial calculator can be used here too. However I want chances of 3 or more successes.)
50%
----------------------------------------------------------------------
Conclusion: the group check rules as written are better if you don't want absurdley high success rates. They will stay closer to a single characters chance to succeed than not using the group check rules.
And with Gnarl's way of handling group checks by increasing the DC and shooting for 1 success, assuming DC 15 is a 30% chance of success for all the players, that makes DC 18-20 a 5-15% chance of success individually. With four PCs, that's 48%, 34%, and 19% for DCs 18-20, respectively. With five PCs, that's 56%, 41%, and 23%.
If my math is correct (please check), then what conclusions may be drawn about Gnarl's approach compared to the standard group checks?
I did this Excel spread sheet for those that are interested in the numbers. I wouldn't mind if someone could check the numbers.
Groups: these columns contain the probability of succeess when using the group check rules.
Lore: I called it Lore because you typically make these kinds of checks when making an Intelligence check to determine if you know something or not. In this situation, if a single member of the group succeeds at the check, then the whole group succeeds.
Stealth: When a whole group is trying to sneak around, if a single person fails at his stealth check, the sneaky operation fails miserable for the whole group.
The 30%, 50%, and 70% are the probability of one person succeeding.
iserith wrote:
Gnarl wrote:
Gnarl wrote:
Could a solution be to set party check DC's for tasks and then just have each member of the party roll and see if the combined total meets the party check DC? If so pass yay!!! if not fail?
Then to decide who failed you could compare each character to the average that would need to be rolled by each character to beat the DC and have that comparison tell you who cause the failure for the narrative side?
Group Checks are a narrative tool IMO. Great for travel, great for whenyou are you have extended time to set stuff up. IN the case of group stealth. IF the group has made a plan and is sneaking in groups, then i would allow a group check.
Gnarl wrote:
Median is easy. In odd groups just take the middle roll and compare to dc. In even groups take the two middle rolls and average them and then compare that to the dc.
Thiugh it has the effect of bringing the even number of player rolls down to where the odd number ones are instead if vice versa
iserith wrote:
FrogReaver wrote:
Yea that's basically it. You can turn an odd group into an even by removing the median guy or you can turn an even group into an odd by taking the median of the two middle numbers.
Gnarl wrote:
I'm actually fascinated by the statistics conversation, but in the actual example of the OP:
I would suggest each PC who hides making their own Stealth check and that would be the DC for each character, and then have the opponent group use their highest score and roll one Perception check for each. (forgive me if these are the wrong terms) Just because one person is noticed doesn't mean they're ALL noticed -- unless they're huddling together.
Yes, the odds become greater that the group will be discovered, but that's why the dwarves sent Bilbo into Smaug's lair, instead of ALL of them. More people of varying skill equates to higher risk of getting caught -- and in a case of hiding or some similar talent, a skilled person can help another only so much.
Forgive me if that's an overly simplistic way of looking at the initial example. I'm not even sure if it addressed the main theme, but I wanted to toss my two cents in.
My general rule for when to make group checks is when the following conditions are met:
- The entire group is trying to do the same thing (climb a cliff face, etc.)
- If only one PC fails to do what the group is attempting, the attempt will not automatically fail for the entire party
If the failure of one PC will result in the failure of the entire party, as in the case of sneaking around, I simply make the PC with the worst modifier roll the check.
MechaPilot wrote:
DungeonKrusher wrote:
Advantages of median over the current group check rules include:
1. If the sneaky guy brings a non-sneaky guy with him, then he will be less stealthy.
2. If most of the group are sneaky, you can still bring non-sneakers along - they'll affect the results,but not completely destroy your chances (which IMO is a positive for the game)
3. If you have 5 people who are all equally sneaky, then the odds are the same as if you have one.
4. 4 ~= 5
I see group checks being used for any time that you don't want the players to split the party just to succeed at a roll. Stealth, climbing and social interaction all fit into that mould. If a character's best mechanical option is to stand in the back and do nothing, a group check sounds like a good idea.
Atari2600 wrote:
Gnarl wrote:
FrogReaver wrote:
MechaPilot wrote:
iserith wrote:
Gnarl wrote:
MechaPilot wrote:
iserith wrote:
Saeviomagy wrote:
I think the math analysis is ignoring the circumstances by which difficulty class is set.
bawylie wrote:
I think we're getting too hung up on Stealth check examples.
iserith wrote:
I think bawylie has a point. We're making calculations based on the idea that the DC is the same for a group check as it is for a check by an individual. That may or may not be the case given the context. With that in mind, how does that change up the numbers?
As for examples, let's go with navigating the swamp as mentioned in the Basic Rules:
"For example, when adventurers are navigating a swamp, the DM might call for a group Wisdom (Survival) check to see if the characters can avoid the quicksand, sinkholes, and other natural hazards of the environment. If at least half the group succeeds, the successful characters are able to guide their companions out of danger. Otherwise, the group stumbles into one of these hazards."
So the stakes are clear: Succeed and make it through the swamp, no problem. Fail and run into a hazard (quicksand, sinkhole, other natural hazard). What's the DC for one guy guiding the rest safely through the swamp? What's the DC if all four guys are working together at avoiding danger?
Gnarl wrote:
Thats is a valid point. However Part of the problem is that the DC for a group check cannot be easily set on the fly. If we raise or lower the DC by 2 or by 5 on a group check can the dm actually tell just how much or how little that affects the actual chances for success. The answer is he can't tell what that change in dc actually does without doing some complicated math to figure it out. Heck the dm can't even figure out the chance for success for a given DC without using the same complicated math.
a DM need to know all this to set proper DCs for group checks. It's just not going to be possible with the group check rules as written.
my newest suggestion makes it easy for a dm to set an appropriate DC for a group check. He can do this on the fly. He can choose how to set the DC almost exactly the way he would set it for a single individual check.
iserith wrote:
It's a judgment call. Is this possible or automatic? If there's a chance at success, decide if that's easy, medium, hard, or very hard. If a group is doing it, drop it down a step (maybe 2).
I anticipate group DCs 60% will be DC10 and 40% will be DC 15.
No complicated math here. In fact do this: All group DCs are 10 until 11th level and then they're 15.
Or, just use your DM judgment. There's no wrong way to do it. Except for focusing on math alone. That's simply incomplete.
bawylie wrote:
Gnarl wrote:
iserith wrote:
iserith wrote:
Gnarl wrote:
I cracked open my 4e Rules Compendium as I remembered that the rule of thumb in that game was to set DCs at Easy for group checks. There was some similar language in this section (pages 128-129) as the current rules:
"The Dungeon Master sometimes asks the adventurers to make a check as a group. Doing this is called making a group check, which is useful when a number of individuals are trying to accomplish something as a group. In such a situation, the characters who are skilled at a particular task help cover those who aren't.
To make a group check, everyone in the group makes a skill check or an ability check specified by the DM. A group check is almost always against an easy DC. If at least half the group succeeds, the whole group succeeds. Otherwise, the group fails.
Group checks might come up in a variety of situations: when the adventurers try to sneak past some sentries (using Stealth), try to scale a sheer cliff together (using Athletics), use disguises to pass as a group of orc soliders (using Bluff), and so on. Such checks are particularly common in skill challenges, discussed later in this chapter."
Of course, in that game, the DCs scale by level unlike in D&D 5e. But I wonder if setting the DC to easy is something of a clue about the math.
Gnarl wrote:
iserith wrote:
Necro'ing this thread to see if Gnarl and FrogReaver are still following and to ask them (or anyone else) to look at how I did the group check in this scenario: Quiet Please. I'd appreciate it if you can double check the math and tell me how it sits with you design-wise in light of the discussion that took place in this thread. Thanks.