PDA

View Full Version : DM Help What Skills should DMs roll for players?



ReNoid
2017-09-01, 06:08 AM
I've always worked on the assumption that Players should roll their own skills to determine the success or failure of what it is they do. Lately, I've wanted to try something different and I had heard that some DMs roll skill checks for Players instead of having them roll it.

For example: You're floating down the river on a barge with just yourself and the pole guiding your boat. A flat barge with nothing to hide behind. You spy orcs on the riverbank and attempt to hide... in the open... on a barge floating down the river. Even a natural 20 can't prevent the orcs from checking it out this "abandoned raft" (and yes, you could HIDE under the barge but then you run the risk of splashing water... drawing attention from things in the water and NOT orcs, etc)

Or another example would be Sense Motive. If an NPC is bluffing the Party, by mere mention of "Make a Sense Motive check" leads the party to already believe that the NPC is lying to them. Natural PC suspicion should be allowed, especially if "Shady Tom, the town Alchemist" comes off as really shady. But I mean, in cases where things like that occur "You believe he's telling the truth." Breaking the chance and option for a PC to metagame.

So what other skills would you think it appropriate for a DM to roll instead of their players? Would it just be situational or would it be all the time?

Darth Ultron
2017-09-01, 06:13 AM
In general a Dm only makes a roll when the character is not aware of something.

The whole ''DM tells the player to make a check, but also tells the player that the character does not know about the check '' is clumsy and awkward. And only the best gamers can play out something they know that their character does not know.

Though, in a lot of games just to save time, the DM makes a lot of skill rolls for the players.

The type of game where the players make no rolls can be loads of fun....but only for the right players.

CarpeGuitarrem
2017-09-01, 06:58 AM
I like 4th Edition's take here, just extrapolated across other skills. You use an assumed roll of 10, and if the player thinks something's up, they actively roll.

hymer
2017-09-01, 07:14 AM
Or another example would be Sense Motive. If an NPC is bluffing the Party, by mere mention of "Make a Sense Motive check" leads the party to already believe that the NPC is lying to them. Natural PC suspicion should be allowed, especially if "Shady Tom, the town Alchemist" comes off as really shady. But I mean, in cases where things like that occur "You believe he's telling the truth." Breaking the chance and option for a PC to metagame.

Well, never say 'You believe he's telling the truth.' Say 'He seems to believe in what he's saying.' Describe what PCs experience. Just because they've not caught someone in an open lie doesn't mean they can't be lying.

I call for plenty of Sense Motive checks for various minor stuff. Maybe the NPC is starstruck at meeting the PCs, or s/he resents the holy symbol they're displaying openly, or s/he's trying to pretend s/he's not drunk. Ideally, I'll only confirm what the players already suspected from my acting out the NPC. But sometimes those rolls are because the NPC is hiding something. And most of the time you don't get more than that; you don't know what the NPC is concealing, but you feel confident that something isn't right. That's the player's cue to start wondering if you can interrogate this NPC more closely, or if you can guess what's being hidden.

Glorthindel
2017-09-01, 07:21 AM
I don't like taking dice rolls out of players hands - the reality is player like rolling dice. So my solution is "blind" rolls - in situations where the character either shouldn't know the result (such as stealth rolls) or even know what they are rolling for, I keep a box behind by DM screen and make them throw their dice in to that, so I can read the result and pass the dice back. It keeps the player actually making the roll without giving away their success or faliure.

Celestia
2017-09-01, 07:48 AM
The DM should roll any time a player should not realistically know whether or not they succeeded or even whether or not there was anything to roll for. With several skills, success or failure is obvious. For others, not so much. You can tell that you failed your Climb check by the fact that you are still at the bottom of the cliff. However, when you fail to identify a creature, knowing that you were rolling a Knowledge (Dungeoneering) check greatly limits the possibilities. And if the DM calls for a Sense Motive check, you don't even need to roll to know that the NPC is lying.

Aneurin
2017-09-01, 09:42 AM
The only times I'll roll on behalf of my players is when they shouldn't know if they've succeeded or failed. Or maybe if they should never know there was a testto be made to begin with, like ambushes.

For instance, knowledge checks. They shouldn't know if the info is right or wrong, just that it's what their character 'knows' is true. If they make a knowledge test, and come up with the idea that werewolves will run away if you smack them on the nose with a rolled up newspaper... they should be willing to consider it might true, even if it sounds like complete nonsense. Or if they 'know' that werewolves can only be harmed by silver, they should actually go looking for some silver to weaponise rather than realising the whole silver thing is a complete myth because it was what they were told when they botched the test.

Tinkerer
2017-09-01, 09:49 AM
I've done a few different things to try and alleviate this problem. I mean I've got a great group of players but even then I run into a few who would still try something like "I roll a 5 to search this area... Oh hey maybe somebody should double check my work." If I think it's going to be a problem I usually either clear it with the player OR (because players like rolling dice) during the warm-up period I have them roll about 20 results and put the results on scrap pieces of paper then put those scraps in a small sack and hand it to me. Of course this works best if you're in a group of about 4-5 but I do find that they are a lot more willing to accept the results when they are the ones who rolled them. At one point I had a dice tower where you could slide the base back and forth but I can't really say that I enjoyed it due to some faults in its construction.

Additionally I think the GM should NEVER be calling for a Sense Motive check. Give the players a passive Sense Motive ability and then if they choose to they can roll (well have you roll/draw a slip) on specific statements (I usually limit this to one per conversation barring extreme circumstances).

If they don't know what they're rolling for at all I usually let them roll out in the open though.

Thrudd
2017-09-01, 10:03 AM
All perception related things
Searching and investigating
Hiding/stealth
sometimes knowledge

anything else where the character would not be able to see the result of their action right away, or would not be aware of unless the result of the roll was positive.

for example - the character forges someone's signature and hands off the letter to be delivered- the character doesn't know whether their forgery is good enough or not - they obviously tried their best. They shouldn't see the roll - you just tell them what happens after the letter is received.

Mastikator
2017-09-01, 10:52 AM
All of them.

Edit- I'll add an addendum to that, I know it feels like you're more in control when you throw the dice but surely you(everyone) must realize that that is an entirely irrational belief and if you did in fact have the capacity to actually change the outcome of the dice then you would be cheating. (in which case you especially shouldn't be allowed to throw the dice, even if you brought your own and the DM didn't)

Pex
2017-09-01, 11:16 AM
Way back in 2E it was a sometimes practice for a DM to roll a PCs saving throw behind the DM screen to reflect a magical attack the PC/Player doesn't know is happening. That fell out of favor but only partially out of caution against possible DM tyranny. When something is serious enough to affect a PC that requires a saving throw, the player wants to control his own fate. Even if he doesn't know the source it's his character. He wants to know and should know something is happening.

The same line of thinking of PC in character not knowing something leads to the DM wanting to roll, but the problem of taking away the player's agency for his own character is still there. This is where Take 10/Passive Scores can come in to solve the competing desires (accepting at face value the DM is not being tyrannical). The DM is rolling for the bad guy/trap, not the PC, or let the Take 10/Passive score determine autosuccess or not, and the player gets to play the game without unintentional metagaming.

My Pathfinder DM notes down each character's AC, saving throws, and Perception. During the game he'll ask a player to roll a d20, doesn't say for what, just roll. He uses that roll and applies it to a saving throw or Perception as appropriate. Most of the time it's Perception. This also maintains the check against unintentional metagaming and the player controlling his own character's fate.

Thrudd
2017-09-01, 04:39 PM
Way back in 2E it was a sometimes practice for a DM to roll a PCs saving throw behind the DM screen to reflect a magical attack the PC/Player doesn't know is happening. That fell out of favor but only partially out of caution against possible DM tyranny. When something is serious enough to affect a PC that requires a saving throw, the player wants to control his own fate. Even if he doesn't know the source it's his character. He wants to know and should know something is happening.

The same line of thinking of PC in character not knowing something leads to the DM wanting to roll, but the problem of taking away the player's agency for his own character is still there. This is where Take 10/Passive Scores can come in to solve the competing desires (accepting at face value the DM is not being tyrannical). The DM is rolling for the bad guy/trap, not the PC, or let the Take 10/Passive score determine autosuccess or not, and the player gets to play the game without unintentional metagaming.

My Pathfinder DM notes down each character's AC, saving throws, and Perception. During the game he'll ask a player to roll a d20, doesn't say for what, just roll. He uses that roll and applies it to a saving throw or Perception as appropriate. Most of the time it's Perception. This also maintains the check against unintentional metagaming and the player controlling his own character's fate.

But you aren't taking away a player's agency, in any way. The player was never "controlling their own fate" just because they rolled the die - since they don't get to choose when the die is rolled or what for.
Passive scores are fine, too, it's essentially the same thing since both are invisible to the player - either way does not grant them any more or less agency.
Thinking the person who rolls the die is somehow more "in control" is just superstition (unless you're cheating), and there's no reason to humor that idea. The players still get plenty of chances to roll dice for other things that actually makes sense for them to be rolling.

Nifft
2017-09-01, 04:54 PM
I'd rather have players roll everything -- so ideally, monsters would get passive Perception and passive Stealth, so players would roll against both.

The combat form of this would be player-only rolls for both attack and defense -- so players would roll an "AC Save" (maybe "Deflection check"?), and also players would roll "Will Attack" / "Weapon Attack" / etc.

D&D 4e can certainly do this.

D&D 5e can probably do this.

DungeonWorld (PbtA) does this by default, and that's part of why it's great.

Pex
2017-09-01, 05:37 PM
But you aren't taking away a player's agency, in any way. The player was never "controlling their own fate" just because they rolled the die - since they don't get to choose when the die is rolled or what for.
Passive scores are fine, too, it's essentially the same thing since both are invisible to the player - either way does not grant them any more or less agency.
Thinking the person who rolls the die is somehow more "in control" is just superstition (unless you're cheating), and there's no reason to humor that idea. The players still get plenty of chances to roll dice for other things that actually makes sense for them to be rolling.

If it doesn't make a difference except the player feels better that he gets to roll the die why not let him roll the die?

Thrudd
2017-09-01, 06:49 PM
If it doesn't make a difference except the player feels better that he gets to roll the die why not let him roll the die?

It doesn't make a difference to the outcome. It does make a difference to the perception and immersion of the players. At every point, effort should be made for the player's knowledge and perception to be as close to that of the character as possible. Every possible situation where "metagaming" could come up should be avoided if at all possible.

A player should not know there was something that their character failed to see or hear, a low roll might give that away - that a roll was asked for at all gives that away. Things invisible to the characters are ideally invisible to the players. A character doesn't know if they are sneaking quietly/carefully enough until/unless somebody sees or hears them - and they don't necessarily know that someone sees or hears them until something happens to alert them to that fact.

So there are real reasons to hide these sorts of rolls, and no real reasons to have the players make them.

Tanarii
2017-09-01, 08:01 PM
In 5e, ideally none. That's exactly what the Passive score mechanic is for. Any time a check needs to be a secret to the player, ie any time they can't even know a roll was needed, nor the result of a roll.

Passive (wisdom) perception gets a special call out, because it's by far the most common check that requires a secret roll.

Pex
2017-09-01, 09:25 PM
It doesn't make a difference to the outcome. It does make a difference to the perception and immersion of the players. At every point, effort should be made for the player's knowledge and perception to be as close to that of the character as possible. Every possible situation where "metagaming" could come up should be avoided if at all possible.

A player should not know there was something that their character failed to see or hear, a low roll might give that away - that a roll was asked for at all gives that away. Things invisible to the characters are ideally invisible to the players. A character doesn't know if they are sneaking quietly/carefully enough until/unless somebody sees or hears them - and they don't necessarily know that someone sees or hears them until something happens to alert them to that fact.

So there are real reasons to hide these sorts of rolls, and no real reasons to have the players make them.

That's not "should". That's your preference.

For some players it will take a while to get in the habit of understanding that a low roll is not an autofail. DCs aren't that high or the DM rolled low himself for the bad guy's opposed roll. I find myself reminding players of that, and they are pleasantly surprised when sometimes they do succeed. This is true for 5E and Pathfinder. They do get it eventually, except for my benign dedicated munchkin friend even after 12 years of playing with him. Can't win them all. :smallsmile:

Dragonexx
2017-09-02, 12:30 AM
I'll second (or third or fourth or whatever) rolling for things the player shouldn't know whether they succeed or fail at (some spot or listen checks, sense motive, ect.).

In forum PBP games I roll saving throws for players as well for the sake of expediency.

Mastikator
2017-09-02, 03:18 AM
If it doesn't make a difference except the player feels better that he gets to roll the die why not let him roll the die?

I sympathize with the need to feel in control of the effort you put into your actions and its real affect on immersion for the players (which is important) but you know what also ruins immersion? When the flow of the game is disrupted because someone threw the dice funny and it fell off the table and under the sofa and now everyone has to wait for you to get the dice back and throw it again. Not only does it waste several people's time (which is disrespectful and shame on everyone who does it, including me (who should know better because I constantly complain about it (also triple parenthesis for the win))) it ruins immersion.

Most are too afraid of confrontations so it's better to just deny everyone their dice privileges and let the DM do it.


The best of all worlds would be to have an app on a tablet in the middle of the table that you can push a button to get an appropriate random number, that way players can get their immersion and nobody's time is wasted by That Guy™

Pelle
2017-09-02, 04:40 AM
It doesn't make a difference to the outcome. It does make a difference to the perception and immersion of the players. At every point, effort should be made for the player's knowledge and perception to be as close to that of the character as possible. Every possible situation where "metagaming" could come up should be avoided if at all possible.

A player should not know there was something that their character failed to see or hear, a low roll might give that away - that a roll was asked for at all gives that away. Things invisible to the characters are ideally invisible to the players. A character doesn't know if they are sneaking quietly/carefully enough until/unless somebody sees or hears them - and they don't necessarily know that someone sees or hears them until something happens to alert them to that fact.

So there are real reasons to hide these sorts of rolls, and no real reasons to have the players make them.

I basically agree, but feel there are times when letting players know that the characters failed checks can be good for the session tension. It can for example make an ambush or assassination feel less like a screwjob.

Frozen_Feet
2017-09-02, 12:21 PM
Most games I play are pretty neutral as to who is rolling the dice. It could be the GM, it could be the player, it could be whoever's closest to the proper die. The biggest reason to let players roll their dice is to keep their attention in the game. It keeps the players' eyes and thoughts on what's happening.

I have a preference for all dice to be rolled in the open, so everyone can see the result. If I feel like keeping secrets, I just don't tell what a roll is for.

FreddyNoNose
2017-09-02, 02:54 PM
I basically agree, but feel there are times when letting players know that the characters failed checks can be good for the session tension. It can for example make an ambush or assassination feel less like a screwjob.

Just have them roll randomly every few minutes for no reason.

Max_Killjoy
2017-09-02, 03:49 PM
I've always worked on the assumption that Players should roll their own skills to determine the success or failure of what it is they do. Lately, I've wanted to try something different and I had heard that some DMs roll skill checks for Players instead of having them roll it.

For example: You're floating down the river on a barge with just yourself and the pole guiding your boat. A flat barge with nothing to hide behind. You spy orcs on the riverbank and attempt to hide... in the open... on a barge floating down the river. Even a natural 20 can't prevent the orcs from checking it out this "abandoned raft" (and yes, you could HIDE under the barge but then you run the risk of splashing water... drawing attention from things in the water and NOT orcs, etc)

Or another example would be Sense Motive. If an NPC is bluffing the Party, by mere mention of "Make a Sense Motive check" leads the party to already believe that the NPC is lying to them. Natural PC suspicion should be allowed, especially if "Shady Tom, the town Alchemist" comes off as really shady. But I mean, in cases where things like that occur "You believe he's telling the truth." Breaking the chance and option for a PC to metagame.

So what other skills would you think it appropriate for a DM to roll instead of their players? Would it just be situational or would it be all the time?

The GM should only roll for players when it's absolutely crucial -- when asking the player(s) to roll would blatantly and openly telegraph something that would ruin immersion, suspense, etc, for the table. The cost in player agency that comes with taking the roll out of the players' hands, demands a critical benefit.

Your barge example has nothing to do with who rolls the dice, it's purely a matter of there being nowhere to hide. The dice can't enable the impossible.

Dimers
2017-09-02, 06:18 PM
As GM I roll for players when the character wouldn't know how well they did, which basically boils down to knowledges and senses. I presume they do know how stealthy they're being ... though not whether an opponent has sensed them anyway.

Without question, this works best when the players can and do trust the GM.

Tanarii
2017-09-02, 06:24 PM
The cost in player agency that comes with taking the roll out of the players' hands, demands a critical benefit. Thats not what player agency means.

Max_Killjoy
2017-09-02, 06:42 PM
Thats not what player agency means.

So you'd be fine with your GM making all the rolls for everyone?

Because most players I know wouldn't -- they'd feel that something critical about their character's success / failure had been taken out of their hands.

Frozen_Feet
2017-09-02, 06:45 PM
Yeah, it's more about player involvement. When a player is rolling, their eyes are on the game and the feeling of participating is tangible - even if dice are doing all the decisions and the course of the game is not influenced by who rolls them.

Knaight
2017-09-02, 06:45 PM
So you'd be fine with your GM making all the rolls for everyone?

No, but given rolls made in the open that objection is based entirely on rolling being part of the experience* and not agency in any way.

*To some extent dice are just physically fun to roll.

Max_Killjoy
2017-09-02, 07:02 PM
Yeah, it's more about player involvement. When a player is rolling, their eyes are on the game and the feeling of participating is tangible - even if dice are doing all the decisions and the course of the game is not influenced by who rolls them.


No, but given rolls made in the open that objection is based entirely on rolling being part of the experience* and not agency in any way.

*To some extent dice are just physically fun to roll.


If people don't like this being included under "agency", that distinction isn't important to my point. Call it involvement, call it having hands-on, call it whatever, it's the sense of having this important thing that drastically affects their character, in their hands, and not in someone else's.

If the roll is going to fail, a lot of players want it to at least fail coming out of their fingers and not someone else's.

Beleriphon
2017-09-02, 07:32 PM
If people don't like this being included under "agency", that distinction isn't important to my point. Call it involvement, call it having hands-on, call it whatever, it's the sense of having this important thing that drastically affects their character, in their hands, and not in someone else's.

If the roll is going to fail, a lot of players want it to at least fail coming out of their fingers and not someone else's.

And that's usually a very human response to the idea that we have some control over what amounts to pure random luck.

I'm in the same boat as Max though either way. The DM/GM/Storyteller/Arbiter/whatever should only roll player skills when the players doing so would ruin some surprise or other mystery. If you tell the players to roll some kill involving finding things they're going to expect to need to find something, which is okay if the idea of the current in game activity is to find something. It gets to be different when you ask to roll for "figure out of that guy is lying" skills, because you've just telegraphed there is something about "that guy" that might, or might not be truthful.

The best way to think of it is would being told Bruce Willis is dead at the start of The Sixth Sense ruin the movie? Probably. Would being told The Village was actually a bunch of modern weirdos ruin the movie? Probably not, since it was awful.

Max_Killjoy
2017-09-02, 07:51 PM
Would we all agree that any roll initiated by the player / PC should be rolled by the player?

Pex
2017-09-02, 08:15 PM
If people don't like this being included under "agency", that distinction isn't important to my point. Call it involvement, call it having hands-on, call it whatever, it's the sense of having this important thing that drastically affects their character, in their hands, and not in someone else's.

If the roll is going to fail, a lot of players want it to at least fail coming out of their fingers and not someone else's.

Yes.

This.


And that's usually a very human response to the idea that we have some control over what amounts to pure random luck.

I'm in the same boat as Max though either way. The DM/GM/Storyteller/Arbiter/whatever should only roll player skills when the players doing so would ruin some surprise or other mystery. If you tell the players to roll some kill involving finding things they're going to expect to need to find something, which is okay if the idea of the current in game activity is to find something. It gets to be different when you ask to roll for "figure out of that guy is lying" skills, because you've just telegraphed there is something about "that guy" that might, or might not be truthful.

The best way to think of it is would being told Bruce Willis is dead at the start of The Sixth Sense ruin the movie? Probably. Would being told The Village was actually a bunch of modern weirdos ruin the movie? Probably not, since it was awful.

I knew that was what they were doing when he got shot. It would have been better if Willis' first appearance was in the chair in the living room across from the mother after spending the opening scene learning about the boy instead of Bruce Willis and his wife.

Knaight
2017-09-03, 12:32 AM
Would we all agree that any roll initiated by the player / PC should be rolled by the player?

No. As just one example, I have no issue with an enemy rolling to dodge for cover after a PC lobs a grenade at them. Similarly, when a PC manages to crash their vehicle by doing something stupid I have no particular issue with the GM rolling on a table for what happens to that vehicle.

Pex
2017-09-03, 12:51 AM
Would we all agree that any roll initiated by the player / PC should be rolled by the player?


No. As just one example, I have no issue with an enemy rolling to dodge for cover after a PC lobs a grenade at them. Similarly, when a PC manages to crash their vehicle by doing something stupid I have no particular issue with the GM rolling on a table for what happens to that vehicle.

Depends on your semantic meaning of a roll initiated by the player. The enemy dodging for cover is the NPC's response. Of course the DM rolls for that. The "roll initiated by the player" would be the roll that determines how accurate the player's character threw the grenade presuming there is a roll to hit. Likewise let the player roll the stealth check when he wants to be sneaky. The DM rolls the NPC's perception behind the screen to his heart's content, even if there's no NPC in the area to have possibly noticed the PC.

Tanarii
2017-09-03, 02:04 AM
So you'd be fine with your GM making all the rolls for everyone?

Because most players I know wouldn't -- they'd feel that something critical about their character's success / failure had been taken out of their hands.Yes. I'd also be fine with it all being computer generated. I'd also be fine with a DM making all the decisions by DM fiat, although that would pretty drastically remove various elements of tensions. And if those were the rules.


If people don't like this being included under "agency", that distinction isn't important to my point. Call it involvement, call it having hands-on, call it whatever, it's the sense of having this important thing that drastically affects their character, in their hands, and not in someone else's.

If the roll is going to fail, a lot of players want it to at least fail coming out of their fingers and not someone else's.Involvement and hands on are a totally different matter. That's active player engagement. That's a great reason for the reason for players rolling dice. It keeps them in the game. So does the DM rolling the dice for reasons. Players perk up and pay attention when they see the DM roll a die.

But superstitions about controlling your own fate / agency by personally rolling the dice are a terrible reason to require dice be rolled by players.

Simple example. 5e Firebolt is an attack roll. 5e Sacred Flame is a saving throw. I've considered moving all cantrip attacks to players because it's rolled most rounds. This reduces my overhead in dice rolling as DM, and adds player engagement by giving the Cleric player something more to do on his turn.

But there's no particular reason to do it in terms of this thing that drastically affects their character being in their hands. The vast majority of DnD spells in all editions except 4e are saves rolled by the DM. And that's fine, despite outcomes of actions initiated by their characters drastically affecting their characters.

Last session I had players on the edge of their seat because the BBEG had just cast Conjure Animals, and they were going to get slaughtered if they could not disrupt its concentration. When one plinked it with a javelin, I told them the target number and bonuses and rolled in the center of the table. They were holding breath for that dice roll more than they had been for any other roll in the session. It was an amazing moment, and they were literally cheering when it failed. Anecdotal, but I've got an anecdote from my very last session that shows me your theorycrafting is completely wrong.

Max_Killjoy
2017-09-03, 08:14 AM
Yes. I'd also be fine with it all being computer generated. I'd also be fine with a DM making all the decisions by DM fiat, although that would pretty drastically remove various elements of tensions. And if those were the rules.


GM fiat, even worse. Most players would refuse to play under those rules. Just look at all the threads and comments that touch on that subject.




Involvement and hands on are a totally different matter. That's active player engagement. That's a great reason for the reason for players rolling dice. It keeps them in the game. So does the DM rolling the dice for reasons. Players perk up and pay attention when they see the DM roll a die.

But superstitions about controlling your own fate / agency by personally rolling the dice are a terrible reason to require dice be rolled by players.


It's hardly superstitious, don't be insulting.

Max_Killjoy
2017-09-03, 08:17 AM
Depends on your semantic meaning of a roll initiated by the player. The enemy dodging for cover is the NPC's response. Of course the DM rolls for that. The "roll initiated by the player" would be the roll that determines how accurate the player's character threw the grenade presuming there is a roll to hit. Likewise let the player roll the stealth check when he wants to be sneaky. The DM rolls the NPC's perception behind the screen to his heart's content, even if there's no NPC in the area to have possibly noticed the PC.

That's along the lines of what I meant, yes.

The PC attacks, or dodges, or sneaks, or whatever, that's the player's roll.

Tanarii
2017-09-03, 08:26 AM
GM fiat, even worse. Most players would refuse to play under those rules. Just look at all the threads and comments that touch on that subject. Every single game has stuff decided by GM Fiat. And stuff that isn't. They just draw the line of what is and what isn't in different places. I do get what you mean. But that's because players trust DMs to be random less than they trust dice to be random. And a lack of randomness is what isn't considered 'fair'. That's why many players can't stand the idea of a DM fudging either.


It's hardly superstitious, don't be insulting.If you're going to claim that somehow rolling a die yourself puts the result of the roll, and thus your fate / ability to make decisions effect only, somehow more under your control ... you're being superstitious.

Edit: unless you know how to manipulate a die roll? Which is cheating, not controlling your own fate.

Max_Killjoy
2017-09-03, 08:31 AM
If you're going to claim that somehow rolling a die yourself puts the result of the roll, and thus your fate / ability to make decisions effect only, somehow more under your control ... you're being superstitious.


Actually, it's not, and my old gaming group got to the point where we recorded 1000s upon 1000s of die rolls because of an argument on the matter. For whatever reason -- there's no claim of explanation here, just observed phenomena -- some people are more streaky, and some people just roll "off average" even across statistically significant numbers of rolls.

And even if that weren't the case, if it's my character, I'm rolling -- if someone's going to invoke random chance regarding my character's fate, it's going to be me.

Tanarii
2017-09-03, 08:38 AM
Actually, it's not, and my old gaming group got to the point where we recorded 1000s upon 1000s of die rolls because of an argument on the matter. For whatever reason -- there's no claim of explanation here, just observed phenomena -- some people are more streaky, and some people just roll "off average" even across statistically significant numbers of rolls. Buy better dice.


And even if that weren't the case, if it's my character, I'm rolling -- if someone's going to invoke random chance regarding my character's fate, it's going to be me.How do you handle games that make some of your attacks an attack roll, and some of your attacks a saving throw for the enemy? And vice versa, some of the time the enemy gets to attack you, and other time you get to save vs the enemies attack? How do you distinguish between what 'should' be an attack, and what 'should' be a dodge?

Or do you just not play games like that,mane instead play ones where the players roll all the dice when their character is involved in any way in something? (AW comes to mind. I don't think the MC ever rolls does he?)

Max_Killjoy
2017-09-03, 08:55 AM
Buy better dice.


We accounted for that.




How do you handle games that make some of your attacks an attack roll, and some of your attacks a saving throw for the enemy? And vice versa, some of the time the enemy gets to attack you, and other time you get to save vs the enemies attack? How do you distinguish between what 'should' be an attack, and what 'should' be a dodge?

Or do you just not play games like that, mane instead play ones where the players roll all the dice when their character is involved in any way in something? (AW comes to mind. I don't think the MC ever rolls does he?)


If the attack auto-hits, and the enemy just gets to roll a save, then you've skipped the part where the dice and the PC's action intersect.

As already noted above.

Grod_The_Giant
2017-09-03, 09:12 AM
But superstitions about controlling your own fate / agency by personally rolling the dice are a terrible reason to require dice be rolled by players.
Something that seems to get forgotten sometimes in theoretical discussions: perception is everything in a game. We play RPGs for a purely subjective reason-- to have fun. And because our goal is subjective, objective truths become less important than how we feel about things. It doesn't matter if a method of rolling is mathematically fair if it feels unfair; it doesn't matter if a class is balanced if it feels too strong or weak.

For many of us, rolling the die feels like you're in control of the situation. Removing that feels like losing control, even if we intellectually know that the odds are the same.

Max_Killjoy
2017-09-03, 09:32 AM
Something that seems to get forgotten sometimes in theoretical discussions: perception is everything in a game. We play RPGs for a purely subjective reason-- to have fun. And because our goal is subjective, objective truths become less important than how we feel about things. It doesn't matter if a method of rolling is mathematically fair if it feels unfair; it doesn't matter if a class is balanced if it feels too strong or weak.

For many of us, rolling the die feels like you're in control of the situation. Removing that feels like losing control, even if we intellectually know that the odds are the same.

Or close to the same.

Tanarii
2017-09-03, 10:53 AM
Something that seems to get forgotten sometimes in theoretical discussions: perception is everything in a game. Im not talking theory here. I'm countering a false theory, that perception / superstition matters, with the actual table reality. Once you're at the table, it doesn't make any difference who rolls a die in terms of controlling the result. (And if it does in a statistically significant way, it's because someone is using bad dice. Or cheating.)

Especially since, as Max_Killjoy has demonstrated, he's drawing lines parsing what counts as 'player action' vs 'not-player action' somewhat arbitrarily. As do many games, especially including most editions of D&D in its attack roll vs save mechanic. There's nothing special to determine when something has the attacker as the active party who should roll, and the defender who should be the active party who rolls. Or on rare occasion, something suddenly becomes an opposed roll, reflecting both parties are active ... even though in almost all cases both parties were already actively opposed.

Max_Killjoy
2017-09-03, 11:08 AM
Im not talking theory here. I'm countering a false theory, that perception / superstition matters, with the actual table reality. Once you're at the table, it doesn't make any difference who rolls a die in terms of controlling the result. (And if it does in a statistically significant way, it's because someone is using bad dice. Or cheating.)


I'm reporting observed phenomena. It had nothing to do with cheating or bad dice. The most bizarre results came from a player who didn't even own his own dice, so he was borrowing different dice all the time -- and there's no way to cheat a d20 roll openly across a table.

If you would rather believe that I'm lying, go for it.




Especially since, as Max_Killjoy has demonstrated, he's drawing lines parsing what counts as 'player action' vs 'not-player action' somewhat arbitrarily. As do many games, especially including most editions of D&D in its attack roll vs save mechanic. There's nothing special to determine when something has the attacker as the active party who should roll, and the defender who should be the active party who rolls. Or on rare occasion, something suddenly becomes an opposed roll, reflecting both parties are active ... even though in almost all cases both parties were already actively opposed.


It's not at all arbitrary.

In the example of an attack by a PC, the attack is the PC's action, and if a roll is required, the player rolls. Dodge or save or parry or whatever is the target's action, so the target's controlling player (GM or otherwise) rolls.

Dimers
2017-09-03, 11:54 AM
If "player-initiated" is the measuring stick for what players roll, then nearly all saving throws in 3.x and most in 5e would be rolled by the DM. You aren't doing or initiating something with the great majority of saves, you're just resisting an incoming effect. 4e's design is more consistent regarding attacks/resistances, but even then, saving throws to overcome lasting problems would get really fuxxored by that "player-initiated" thing.

Tanarii
2017-09-03, 12:16 PM
If "player-initiated" is the measuring stick for what players roll, then nearly all saving throws in 3.x and most in 5e would be rolled by the DM. You aren't doing or initiating something with the great majority of saves, you're just resisting an incoming effect. 4e's design is more consistent regarding attacks/resistances, but even then, saving throws to overcome lasting problems would get really fuxxored by that "player-initiated" thing.
If "playe initiated" is the yardstick. Almost all saving throws for spells cast in 3e AND 5e should be rolled by the player. They're usually casting spells at the enemy.

Most saving throws players make should instead be made by the DM. Often they're traps or effects initiated by the monsters.

The 'initiated' vs 'resisting' line is almost always arbitrary. As I pointed out.

Max_Killjoy
2017-09-03, 12:33 PM
As with the "agency" thing before, you're fixating on the terminology instead of the actual issue -- which is maybe why it seems arbitrary.

Who's "doing the thing"? The attacker attacks, the defender defends, the resistor resists, etc. Whoever is "doing the thing" should be rolling the dice.

Thrudd
2017-09-03, 12:57 PM
I don't disagree that rolling dice is fun, and a part of what makes playing RPGs fun. I want players to roll dice. Just not in certain circumstances, mostly perception and searches. When the characters are surprised, I also want the players to be surprised. If I have something hidden (implying there is a chance it won't be found) and the characters miss it, I don't want the players suspecting that there is something they missed because of dice being rolled (suspecting there is something based on general paranoia or deduction is great, of course). I would prefer my players not ask to "roll perception" or anything similar. You describe actions from the character's POV, and the DM will tell you if you should roll a die for something.

It also really makes a difference what game is being played. Not all are equal. My emphatic stance on perception and other hidden rolls applies to D&D, primary. If we're talking about games where searching for hidden things and surprise attacks are not really so integral, then there may be nothing the GM needs to keep hidden. In Feng Shui, I can't imagine ever needing to disguise anything from the players - you move from one action set-piece to the next. The only difference your investigating-type rolls will make is perhaps the details, nature and venue of the next big action sequence. Whether the roll is high or low, something awesome is going to happen, regardless.

Knaight
2017-09-03, 12:59 PM
Who's "doing the thing"? The attacker attacks, the defender defends, the resistor resists, etc. Whoever is "doing the thing" should be rolling the dice.

In that case almost all combat rolls should be opposed, and static values never make sense.

Cluedrew
2017-09-03, 01:27 PM
For many of us, rolling the die feels like you're in control of the situation. Removing that feels like losing control, even if we intellectually know that the odds are the same.Yeah I'm a bit too intellectual even for that... no I don't mean I'm smarter, I just mean I don't feel like luck is a thing most of the time.

I mean it sort of did during the 3-4 months where I was having about half of my attacks, which required a 4+ on 2d6, miss. But I've been playing dice games for years and outside of that one period I've never had any particular streaks. We make jokes about it for years though.

On Main Topic: find the solution that works for your group, whatever that might be.

Max_Killjoy
2017-09-03, 02:14 PM
In that case almost all combat rolls should be opposed, and static values never make sense.

One could make that argument for multiple reasons.

FreddyNoNose
2017-09-03, 07:42 PM
So you'd be fine with your GM making all the rolls for everyone?

Because most players I know wouldn't -- they'd feel that something critical about their character's success / failure had been taken out of their hands.

I'd have zero problem with a DM rolling all rolls.

Pex
2017-09-03, 10:12 PM
I'd have zero problem with a DM rolling all rolls.

Define "all rolls".

That could mean every skill check. Every attack roll. Every saving throw. Every damage roll. Why bother having the players there?

The Extinguisher
2017-09-03, 10:34 PM
Define "all rolls".

That could mean every skill check. Every attack roll. Every saving throw. Every damage roll. Why bother having the players there?

To actually play the game? Rolling isnt the game, it just simulates random chance. when i play a video game, the system does all the randomization for me, but it still doesnt run on its own.

theres certainly an argument to be made about keeping your players engaged, especially in games like d&d where combat rounds can be long and youre often just waiting for your turn (weve started using active defense rolls for all our dnd games) but there are plenty of systems that dont use dice at all that keep players involved and in control. its the role playing, or tactical simulations, or whatever else that goes into the game that should keep you engaged, not the plastic shapes.

also, Max, if youre dice rolls arent producing fair results, your dice arent fair, or your sample size is too small. theres no answer to that problem that is "some people ignore probability"

Thrudd
2017-09-03, 10:50 PM
Define "all rolls".

That could mean every skill check. Every attack roll. Every saving throw. Every damage roll. Why bother having the players there?

The players describe what their characters are doing. The GM tells them what their characters perceive and what happens when they take actions. The dice are just the way to have an objective means of determining outcomes and introducing random chance into the world of the characters - the person rolling the dice does not have an impact on what it is the players are actually doing in an RPG.

Having players roll dice sometimes is a psychological ploy that encourages players to feel more engaged with the game, and also a means of establishing some transparency to discourage GM abuses. In a vacuum, with players and GM perfectly following the rules, trusting each other, and completely engaged and paying attention to the game, it would not make a difference if a player ever laid hand on a die.

goto124
2017-09-04, 12:44 AM
For instance, knowledge checks. They shouldn't know if the info is right or wrong, just that it's what their character 'knows' is true. If they make a knowledge test, and come up with the idea that werewolves will run away if you smack them on the nose with a rolled up newspaper... they should be willing to consider it might true, even if it sounds like complete nonsense.

Personally, I wouldn't say that sort of thing unless someone rolled a nat 1, and even then it's only for fun and I'll mention it's a nat 1 and completely untrue. Otherwise, Knowledge checks always turns up true information. Even with low rolls, the players just get "you don't know".

Imagine making Knowledge checks, not knowing how high or low you rolled, and being given false information. Why make a Knowledge check if you have no idea if the knowledge is real? Red herrings are things that games don't need even more of, the players are more than capable of coming up with them without the DM's help.

Also, I now want to play in a campaign where werewolves run away if you smack them on the nose with a rolled up newspaper :smallamused:

oxybe
2017-09-04, 01:14 AM
Ideally none, I would say.

As a person I can generally tell if I mucked up or did really well. I need to be seriously off my game to bugger up a situation where I think I did super awesome at a thing I'm familiar with but in reality botched it badly.

I see no problem with letting players roll their stealth or perception. Their characters are likely trained for these situations.

For situations like actively rolling to see if they notice a hidden attacker (VS a passive perception), I generally ask for the roll immediately before the attack takes place, if they fail the roll, then they get attacked and everything goes as normal, if they succeed I roll back the clock until the enemy gets into perception range a few rounds prior and basically "redo" that part of the scenario but the player in question is aware that the enemy is there.

Pex
2017-09-04, 01:19 AM
To actually play the game? Rolling isnt the game, it just simulates random chance. when i play a video game, the system does all the randomization for me, but it still doesnt run on its own.

theres certainly an argument to be made about keeping your players engaged, especially in games like d&d where combat rounds can be long and youre often just waiting for your turn (weve started using active defense rolls for all our dnd games) but there are plenty of systems that dont use dice at all that keep players involved and in control. its the role playing, or tactical simulations, or whatever else that goes into the game that should keep you engaged, not the plastic shapes.

also, Max, if youre dice rolls arent producing fair results, your dice arent fair, or your sample size is too small. theres no answer to that problem that is "some people ignore probability"


The players describe what their characters are doing. The GM tells them what their characters perceive and what happens when they take actions. The dice are just the way to have an objective means of determining outcomes and introducing random chance into the world of the characters - the person rolling the dice does not have an impact on what it is the players are actually doing in an RPG.

Having players roll dice sometimes is a psychological ploy that encourages players to feel more engaged with the game, and also a means of establishing some transparency to discourage GM abuses. In a vacuum, with players and GM perfectly following the rules, trusting each other, and completely engaged and paying attention to the game, it would not make a difference if a player ever laid hand on a die.

Again, if it makes no difference who rolls the dice but a player will feel better if he rolls the dice when his character does something, why deny him?

Pelle
2017-09-04, 04:08 AM
I'm reporting observed phenomena. It had nothing to do with cheating or bad dice. The most bizarre results came from a player who didn't even own his own dice, so he was borrowing different dice all the time -- and there's no way to cheat a d20 roll openly across a table.


You know, when 5 guys are rolling dice at a table, someone has to be the most lucky one, nothing special about that. This is determined after seing the results though, you can't extrapolate this to mean the same person will be as lucky in the future. People thinking otherwise is the reason why casinos make money.

Also, some investment funds can show good history records of beating the market, but it's no gurantee that they will in the future. That is because if different funds invest in different parts of the market, per definition some have to be better than the market average. Doesn't mean they will continue doing that when the market change. Buy index funds instead :)


The psychological effect is a good reason to let players roll, even though it is completely irrational. IMX, my players won't feel they have accomplished anything in my sessions, unless I give them a roll and a chance to fail when they want to take actions.

Knaight
2017-09-04, 05:00 AM
Again, if it makes no difference who rolls the dice but a player will feel better if he rolls the dice when his character does something, why deny him?

This is a dramatically different argument than "why even have players", which is what was being addressed there.

Aneurin
2017-09-04, 05:03 AM
Personally, I wouldn't say that sort of thing unless someone rolled a nat 1, and even then it's only for fun and I'll mention it's a nat 1 and completely untrue. Otherwise, Knowledge checks always turns up true information. Even with low rolls, the players just get "you don't know".

Imagine making Knowledge checks, not knowing how high or low you rolled, and being given false information. Why make a Knowledge check if you have no idea if the knowledge is real? Red herrings are things that games don't need even more of, the players are more than capable of coming up with them without the DM's help.

Also, I now want to play in a campaign where werewolves run away if you smack them on the nose with a rolled up newspaper :smallamused:

It's what they get for saying "I roll a Common Lore (Werewolves) check" rather than, say, going to a library. Or asking random people on the docks. Or doing anything at all that might get them better information than random hearsay and urban legends.

But you're right, red herrings aren't particularly interesting - at least, not unless they add something to the game. And if something isn't being added to the game by the roll, why are you even bothering having it rolled for in the first place?

My example was, I'll admit, rather silly and taken to ridiculous extremes just to make the point obvious. A better example is investigation-heavy scenarios, where Inquiry and Charm tests might not only fail to turn up relevant information, but actually provide incorrect answers in situations where learning who's really the secret cultist is actually important, or what sort of bodyguards someone might employ. It means it's important to correlate information, and rely on multiple sources to check for inaccuracies... or the party could just barge in there and hope for the best and suffer for it.

It probably sucks in D&D, or anything where you throw dice to make the talking stop, but more for more investigative and social games it works nicely because it's relevant and adds to the game. Knowing your group's tolerance for these things also helps.

Dimers
2017-09-04, 06:30 AM
If "playe initiated" is the yardstick. Almost all saving throws for spells cast in 3e AND 5e should be rolled by the player. They're usually casting spells at the enemy.

Most saving throws players make should instead be made by the DM. Often they're traps or effects initiated by the monsters.

The 'initiated' vs 'resisting' line is almost always arbitrary. As I pointed out.

I completely agree, and I have a mild preference for games that are more consistent about the difference. I mean, I prefer playing games where the acting creature/effect is what makes the roll, whether that's my PC or the gamemaster's ambush.

Not specifically @Tanarii: While that does address opposed sensory rolls, though, it still doesn't touch rolls to define what a character already knows. There's no action involved at all, no initiator. And when something isn't trying to hide but the PC might not notice it anyway -- an unopposed sensory roll -- there's again nothing performing an action.

If the player actively rolls for that sort of thing, not only are they potentially getting metagame information, they also have to be mind-readers. How else could they guess that a gamemaster has info an Arcana check will uncover, or a Scent roll, or whatever? Unless they waste game time rolling for each knowledge and sense in every situation? And what if the character could be aware of something but the player doesn't happen to ask about it? The GM should roll any knowledges and senses as soon as soon as they become relevant, not waiting for the PC to "try" to know or notice. That solves the metagame issue and the proactivity issue all at once.

Tanarii
2017-09-04, 11:14 AM
Define "all rolls".

That could mean every skill check. Every attack roll. Every saving throw. Every damage roll. Why bother having the players there?


Again, if it makes no difference who rolls the dice but a player will feel better if he rolls the dice when his character does something, why deny him?
That's a pretty huge shift in the goalposts. :smalltongue:

Pex
2017-09-04, 11:26 AM
This is a dramatically different argument than "why even have players", which is what was being addressed there.

It was the first time I asked if it doesn't matter who rolls the dice but the player feels better rolling so why deny him that led to this. Hence the "again". It's circular.

Passive Perception, "Take 10", etc. solves the problem of not wanting to alert players of a possible ambush by telling them to roll. If the player initiates wanting to avoid an ambush in a particular area he gets a roll. Likewise let the stealthy player roll his stealth/move silently/hide all he wants. It doesn't matter if there's an NPC to have noticed him or not. If there is the NPC gets his own perception check. If there isn't the DM gets his own nothing roll. The NPC can have its own Passive Perception, "Take 10" the player is rolling his stealth against, no DM die roll needed, if the DM feels the act of rolling the die behind the screen is itself too much information.

Thrudd
2017-09-04, 02:17 PM
I said it would make no difference if a player never rolled a die. It does make a difference in certain situations that the DM keep rolls secret rather than having players roll them. It would have no impact (other than psychological) if the DM rolls all attacks and damage and saves and skills behind the screen. It does have an impact in certain types of games when the player rolls and sees the results of perception and searching and certain other things. These are the cases in which the DM should always roll secretly. If the very knowledge that a roll is happening, let alone seeing its result, can cause a player to suspect things their character would not, then it should stay secret.

However, if the game is one where the flow works differently than my assumption for D&D - where dice are never rolled unless and until something is about to happen or be revealed to the players - then it wouldn't matter that players roll these types of perception skills. So the failed perception roll will always mean that something immediately happening to the characters has a slightly different effect than if the roll was successful - rolls only occur when there are things immediately happening to the characters. There are not random or hidden things the players might miss, there is no delay that allows player action between when the roll happens and when the prompt for the roll (ambush, trap sprung) has effect. There is never a missed secret door or a hidden treasure or a clue that goes unnoticed, or a stalking creature or a spy following at distance that isn't attacking at that very moment.
So one's perception of what rolls are and aren't appropriate for players to see depends on your game's format.

Using "Take 10" or passive perception are the same thing as the DM rolling those skills - suggesting the use of those mechanics is confirming that certain checks should be kept secret from the players, not denying it. It is just an argument of semantics, at that point, to claim that this is different than the DM secretly rolling the perception skill behind the screen. It only has a difference in the probabilities involved.

Max_Killjoy
2017-09-04, 02:47 PM
You know, when 5 guys are rolling dice at a table, someone has to be the most lucky one, nothing special about that. This is determined after seing the results though, you can't extrapolate this to mean the same person will be as lucky in the future. People thinking otherwise is the reason why casinos make money.

Also, some investment funds can show good history records of beating the market, but it's no gurantee that they will in the future. That is because if different funds invest in different parts of the market, per definition some have to be better than the market average. Doesn't mean they will continue doing that when the market change. Buy index funds instead :)


The psychological effect is a good reason to let players roll, even though it is completely irrational. IMX, my players won't feel they have accomplished anything in my sessions, unless I give them a roll and a chance to fail when they want to take actions.


I'm not talking about 5 people rolling dice at a table for one night. I'm talking about recording 1000s and 1000s of rolls over a couple years.

I have no idea why, but the numbers that came out showed some players with small but consistent and long-term deviations from the "statistical" norm that would have been expected. Different dice, different surfaces, whatever, it didn't matter.

Koo Rehtorb
2017-09-04, 04:43 PM
I'm not talking about 5 people rolling dice at a table for one night. I'm talking about recording 1000s and 1000s of rolls over a couple years.

I have no idea why, but the numbers that came out showed some players with small but consistent and long-term deviations from the "statistical" norm that would have been expected. Different dice, different surfaces, whatever, it didn't matter.

Isn't this an argument for one designated person rolling all the dice? If some people actually had some demonstrable magical ability to alter the laws of probability then this directly goes against the whole intent of a random number generator.

If this ability was intentional then it would be called cheating, if it was unintentional then it would be called a bug (in the fabric of reality?)

PhoenixPhyre
2017-09-04, 04:52 PM
Isn't this an argument for one designated person rolling all the dice? If some people actually had some demonstrable magical ability to alter the laws of probability then this directly goes against the whole intent of a random number generator.

If this ability was intentional then it would be called cheating, if it was unintentional then it would be called a bug (in the fabric of reality?)

Low probability events like this happen. It's not a bug, it's an essential part of the system. Unlike most physical laws, statistical laws only give probabilities, not certainties.

Koo Rehtorb
2017-09-04, 04:54 PM
Low probability events like this happen. It's not a bug, it's an essential part of the system. Unlike most physical laws, statistical laws only give probabilities, not certainties.

I don't think it's real. I am dealing in hypotheticals.

The Extinguisher
2017-09-04, 05:02 PM
I'm not talking about 5 people rolling dice at a table for one night. I'm talking about recording 1000s and 1000s of rolls over a couple years.

I have no idea why, but the numbers that came out showed some players with small but consistent and long-term deviations from the "statistical" norm that would have been expected. Different dice, different surfaces, whatever, it didn't matter.

I find it really unlikely you've been meticulously recording every dice thats rolled around you, and you're just finding patterns where there are none. Selection bias is a hell of a thing. Or your sample size is too small and these deviations aren't statistically significant.

The alternative is that people you are rolling dice with do actually distort probability in which case you should probably publish your results cause thats pretty important news.

FreddyNoNose
2017-09-04, 05:11 PM
The players describe what their characters are doing. The GM tells them what their characters perceive and what happens when they take actions. The dice are just the way to have an objective means of determining outcomes and introducing random chance into the world of the characters - the person rolling the dice does not have an impact on what it is the players are actually doing in an RPG.

Having players roll dice sometimes is a psychological ploy that encourages players to feel more engaged with the game, and also a means of establishing some transparency to discourage GM abuses. In a vacuum, with players and GM perfectly following the rules, trusting each other, and completely engaged and paying attention to the game, it would not make a difference if a player ever laid hand on a die.

Not to forget that we had situations with pre-generated characters back in the day. So if you can play with a pre-generated character, you can play with the DM rolling all the rolls.

This following the rules is back to metagaming. You want to know what to do based on the game rules and not what is my character trying to do. The DM can change anything at anytime just to keep it unpredictable. Perhaps the marys around here can't handle that.

Max_Killjoy
2017-09-04, 05:37 PM
I find it really unlikely you've been meticulously recording every dice thats rolled around you, and you're just finding patterns where there are none. Selection bias is a hell of a thing. Or your sample size is too small and these deviations aren't statistically significant.

The alternative is that people you are rolling dice with do actually distort probability in which case you should probably publish your results cause thats pretty important news.


It wasn't me writing it down, it was the obsessive meticulous math nut who was trying to prove that we were just imagining things. And it was in the mid-late 90s, not current.

As one example, we played A LOT of HERO-system, with its big pools of d6. The average on d6 over time should be 3.5... my average was 3.46 or 3.48, something like that.




Low probability events like this happen. It's not a bug, it's an essential part of the system. Unlike most physical laws, statistical laws only give probabilities, not certainties.


Which kinda relates to the facepalm-worthy notion I hear repeated as gospel fact all the time, where "average" is supposedly defined as "half above, half below". :smalleek:

flond
2017-09-04, 05:40 PM
Not to forget that we had situations with pre-generated characters back in the day. So if you can play with a pre-generated character, you can play with the DM rolling all the rolls.

This following the rules is back to metagaming. You want to know what to do based on the game rules and not what is my character trying to do. The DM can change anything at anytime just to keep it unpredictable. Perhaps the marys around here can't handle that.

Yes having a consistent baseline is so much worse than playing a game of "Adjudicated freeform with lies."

Pex
2017-09-04, 05:51 PM
Not to forget that we had situations with pre-generated characters back in the day. So if you can play with a pre-generated character, you can play with the DM rolling all the rolls.

This following the rules is back to metagaming. You want to know what to do based on the game rules and not what is my character trying to do. The DM can change anything at anytime just to keep it unpredictable. Perhaps the marys around here can't handle that.

Pre-gens happen at gaming conventions (before RPGA and after for non-RPGA games) because they're one-shot adventures just for the fun of playing. There isn't time to create characters, so the DM did it beforehand. Occasionally a DM would let a player play his own character from whatever campaign after looking it over to approve or disapprove particulars, usually magic items.

For an on-going campaign I want to play the character I want to play not what the DM wants me to play. I will allow for using a pre-gen if for a particular campaign there's no session 0 we're playing as soon as we meet just to get started, but I want to be able to tweak the character for my own taste starting session 2. That's how I ended up playing a paladin in one of my 5E games. Normally I play humans, but the character was a mountain dwarf for the pre-gen. I decided to keep the race, but I tweaked many things including ability score allocation.

Max_Killjoy
2017-09-04, 06:03 PM
Pre-gens for conventions aren't really a good example, they're kinda necessary when you only have 2 or 3 hours, and some players might never have played the system before.

FreddyNoNose
2017-09-04, 09:10 PM
Pre-gens for conventions aren't really a good example, they're kinda necessary when you only have 2 or 3 hours, and some players might never have played the system before.

Not the only pre-generated characters out there. But you knew that before you wrote that didn't you.

Max_Killjoy
2017-09-04, 09:21 PM
Not the only pre-generated characters out there. But you knew that before you wrote that didn't you.

I've never been in a non-convention game that used pre-gens... or heard of a non-convention game that required pre-gens... so, um... no, I didn't?

Psikerlord
2017-09-04, 11:27 PM
Generally speaking, if the roll is a contest, the player always rolls, but I might make my roll in secret to keep them guessing.

If it is a player only roll, like perception vs a trap or ambush, mostly I'll still get the player to roll. Because if the result is about to be revealed anyway (trap about to be sprung, ambush about to be launched, etc) - it makes no difference if the player rolls - they're about to find out the result either way regardless.

The only tricky one is something like a secret door or being lied to. The result isnt immediately obvious if missed. In that case, I prefer to fall back on the old "have the players make 10 x d20 rolls before each session. if you need a secret roll from them, roll a d10 and consult the list. keep a list of the PC's perception and insight scores on that same list". So the players do all their rolling, but some stuff might be resolved secretly. I very much disike take 10/passive perception so I dont use that at all.

Having said all that, sometimes I just roll for the PCs anyway, hhaha. I just prefer to try not to.

Sharur
2017-09-04, 11:49 PM
The only tricky one is something like a secret door or being lied to. The result isnt immediately obvious if missed. In that case, I prefer to fall back on the old "have the players make 10 x d20 rolls before each session. if you need a secret roll from them, roll a d10 and consult the list. keep a list of the PC's perception and insight scores on that same list". So the players do all their rolling, but some stuff might be resolved secretly. I very much disike take 10/passive perception so I dont use that at all. .

My solution is that instead of a PC making a skill check, to have a NPC skill check. For example, rather than having a passive perception check to detect a secret door (or a trap or ambush for that matter), I would make a Stealth check for the ambusher, the trap, or the secret door. The latter two I'd give a Stealth bonus (or penalty on one occasion, due to the politics of the adventure) based on how well hidden they were. Advantage/Disadvantage would be reversed (that is, a PC who had advantage to detect traps would cause the trap to have disadvantage on its Stealth roll), and the DCs would be 8+Skill bonus of opposed skill. So Stealth would be opposed by Perception, and Insight (to see if they were being lied to) would be opposed by Deception.

Pelle
2017-09-05, 04:48 AM
I'm not talking about 5 people rolling dice at a table for one night. I'm talking about recording 1000s and 1000s of rolls over a couple years.

I have no idea why, but the numbers that came out showed some players with small but consistent and long-term deviations from the "statistical" norm that would have been expected. Different dice, different surfaces, whatever, it didn't matter.

I believe you, I just think you are putting more weight into the results than you should. Are you saying that some people have an inherent property that makes them more "lucky" than others, meaning that you can also expect them to be lucky also in the future? That sounds like superstition to me. Sure someone can be said to be more lucky than others based on their roll history, but that is based on what was rolled, not what will be rolled. Someone having been "lucky" is normal, it is statistically unlikely that there are no outliers.

If your point is that because it is possible for someone to be "lucky", they should roll the dice themselves, I would just say that the same person can be just as lucky if the DM rolls on their behalf. It's not an argument for or against players rolling, since the purpose of the dice is to be a fair number generator.


It wasn't me writing it down, it was the obsessive meticulous math nut who was trying to prove that we were just imagining things. And it was in the mid-late 90s, not current.

As one example, we played A LOT of HERO-system, with its big pools of d6. The average on d6 over time should be 3.5... my average was 3.46 or 3.48, something like that.

Which kinda relates to the facepalm-worthy notion I hear repeated as gospel fact all the time, where "average" is supposedly defined as "half above, half below". :smalleek:

I think you are confusing average and expected value, or projecting other peoples confusion. The average is the total sum of the results, divided by the number of outcomes. This is after rolling. The expected value is the sum of the different possible outcomes, divided by the number of possible outcomes, before rolling. There is no guarantee that these numbers are the same, because it is random. The die dosn't remember what it has rolled before, the expected outcome for the next roll is not dependent on the roll history. Check out the Gamblers Fallacy, interesting psychology.

For illustration I just did a quick Monte Carlo simulation in Excel, rolling 10000 d6, and taking the average. I did this 10 times, and these are the averages:
3.4859 3.4827 3.5218 3.5229 3.482 3.4931 3.5107 3.5296 3.518 3.5013

I first did it for only 1000 d6, and the results varied more. So you see, your own numbers aren't so special. Play around youself to see how many rolls you need for it to converge on 3.5. Even if you are going to roll many times, there is always some chance of the average not being 3.5.

Max_Killjoy
2017-09-05, 06:17 AM
I believe you, I just think you are putting more weight into the results than you should. Are you saying that some people have an inherent property that makes them more "lucky" than others, meaning that you can also expect them to be lucky also in the future? That sounds like superstition to me. Sure someone can be said to be more lucky than others based on their roll history, but that is based on what was rolled, not what will be rolled. Someone having been "lucky" is normal, it is statistically unlikely that there are no outliers.

If your point is that because it is possible for someone to be "lucky", they should roll the dice themselves, I would just say that the same person can be just as lucky if the DM rolls on their behalf. It's not an argument for or against players rolling, since the purpose of the dice is to be a fair number generator.



I think you are confusing average and expected value, or projecting other peoples confusion. The average is the total sum of the results, divided by the number of outcomes. This is after rolling. The expected value is the sum of the different possible outcomes, divided by the number of possible outcomes, before rolling. There is no guarantee that these numbers are the same, because it is random. The die dosn't remember what it has rolled before, the expected outcome for the next roll is not dependent on the roll history. Check out the Gamblers Fallacy, interesting psychology.

For illustration I just did a quick Monte Carlo simulation in Excel, rolling 10000 d6, and taking the average. I did this 10 times, and these are the averages:
3.4859 3.4827 3.5218 3.5229 3.482 3.4931 3.5107 3.5296 3.518 3.5013

I first did it for only 1000 d6, and the results varied more. So you see, your own numbers aren't so special. Play around youself to see how many rolls you need for it to converge on 3.5. Even if you are going to roll many times, there is always some chance of the average not being 3.5.

I get all that, and I'm making no claims of cause, just the observations. To me, the bigger thing was how consistent it was by player over time. I rolled under average a lot. Another guy in a different game the math nut was in rolled 1 and 20 on d20 more than he "should" have, consistently. I wish I had access to all the recorded numbers and notes, but as far as I know he lost them in the intervening years. Again, I have no idea why this was happening, and the main point is that people have reasons to want to roll the dice themselves, the observation is real regardless of the underlying cause, and calling it "superstition" is both insulting, and a misunderstanding of what the people are thinking.

(And in the second quote box, to be fair, I was replying to one thing in the first two lines, and another thing in the third line -- which disappears in your quotation of me, making it look like a single statement.)





I find it really unlikely you've been meticulously recording every dice thats rolled around you, and you're just finding patterns where there are none. Selection bias is a hell of a thing. Or your sample size is too small and these deviations aren't statistically significant.

The alternative is that people you are rolling dice with do actually distort probability in which case you should probably publish your results cause thats pretty important news.


It wasn't me writing it down, it was the obsessive meticulous math nut who was trying to prove that we were just imagining things. And it was in the mid-late 90s, not current.

As one example, we played A LOT of HERO-system, with its big pools of d6. The average on d6 over time should be 3.5... my average was 3.46 or 3.48, something like that.




Low probability events like this happen. It's not a bug, it's an essential part of the system. Unlike most physical laws, statistical laws only give probabilities, not certainties.


Which kinda relates to the facepalm-worthy notion I hear repeated as gospel fact all the time, where "average" is supposedly defined as "half above, half below". :smalleek:


I should have written should as "should" to make it more clear what I meant in the first part. Again, I'm not making any claims as to why, but when after all those roles a pattern emerges, I'm not going to ignore the pattern because pop-statistics nerds keep screaming that it's "impossible".

In the second part, I was making a side-comment on the way people misunderstand what "average" means, how they make jokes about more than half of people believing they're "above average" as if that's impossible, or the old snark "picture the average person and then think about how half of everyone is dumber than that".

Tanarii
2017-09-05, 08:00 AM
Again, I have no idea why this was happening, and the main point is that people have reasons to want to roll the dice themselves, the observation is real regardless of the underlying cause, and calling it "superstition" is both insulting, and a misunderstanding of what the people are thinking.Why is because random, not statistics.

And no, they don't have a reason to roll themselves based on random variatiom from a statistical mean or average. It doesn't matter who rolls, there will be random variation. If you had picked 1000 random numbers from the mix of players you recorded values from, it would also have been different from the expected average.

It is exactly superstition if you believe your measured and observed outcome provides any justification for people rolling themselves. It's the perfect example of that superstition, the human propensity to find any justification to believe they have some personal special affect on the outcomes.

Max_Killjoy
2017-09-05, 08:27 AM
Why is because random, not statistics.

And no, they don't have a reason to roll themselves based on random variatiom from a statistical mean or average. It doesn't matter who rolls, there will be random variation. If you had picked 1000 random numbers from the mix of players you recorded values from, it would also have been different from the expected average.

It is exactly superstition if you believe your measured and observed outcome provides any justification for people rolling themselves. It's the perfect example of that superstition, the human propensity to find any justification to believe they have some personal special affect on the outcomes.

If you're first going to call me a liar, and then call me superstitious, for reporting what was observed, and giving people some slack for putting a little stock in their observations...

/plonk (https://en.wikipedia.org/wiki/Plonk_(Usenet))

If anyone is superstitious here, it's those who expect large-scale descriptive statistics to be small-scale predictive.


E: Probability is interesting, but it's a little worrisome that people put so much emotional stock in it that I've been called a liar, and incompetent, and superstitious, for posting an empirical observation that they even suspect might hint at taking the supremacy of probability and chance into question.

It's as if I'd said "I saw some strange lights in the sky last night", and someone immediately said "you're an idiot for believing in UFOs".

Pelle
2017-09-05, 08:35 AM
Why is because random, not statistics.

And no, they don't have a reason to roll themselves based on random variatiom from a statistical mean or average. It doesn't matter who rolls, there will be random variation. If you had picked 1000 random numbers from the mix of players you recorded values from, it would also have been different from the expected average.

It is exactly superstition if you believe your measured and observed outcome provides any justification for people rolling themselves. It's the perfect example of that superstition, the human propensity to find any justification to believe they have some personal special affect on the outcomes.

I agree with this, but maybe "superstition" has some negative connotations in English that I'm not aware of. Call it irrational psychological effects :)


A friend of mine is the running joke in our group for "always" rolling low on new HD for his character. We have fun with it, but that doesn't mean I expect him rolling low next time, it's just extra fun if it occurs again. Which is still likely, since it's 50% chance he rolls below the expected value.

Sure, I can understand that players want to roll dice themselves, and have no problems with letting them most of the time. The chances of success is independent of who rolls the dice however, so that itself is not an argument. Keeping players happy, because people in general are irrational, that is an argument for letting them roll though.



(And in the second quote box, to be fair, I was replying to one thing in the first two lines, and another thing in the third line -- which disappears in your quotation of me, making it look like a single statement.)


Indeed. I guess I just didn't understand where you were going with the third line, since it didn't seem like anyone here was claiming something like that.

Tanarii
2017-09-05, 08:49 AM
I agree with this, but maybe "superstition" has some negative connotations in English that I'm not aware of. Call it irrational psychological effects :)Yes, it often has negative connotations. So would irrational psychological effects. Either one means believimg something is real that is actually not real. And unsurprisingly, they get bent out of shape when you call them on it, and attempt to justify their beliefs.

I mean, not like I haven't done that plenty of times, even if I'm only counting posts on these forums. I can empathize with the feeling of being insulted. But nonsense justifications are nonsense. Better to have them called out, even if we get bent because of it.

Max_Killjoy
2017-09-05, 09:03 AM
I agree with this, but maybe "superstition" has some negative connotations in English that I'm not aware of. Call it irrational psychological effects :)


It's deeply insulting.

And ironic, given that it's being directed at someone who takes empirical data more seriously than oracular predictions about the future.




A friend of mine is the running joke in our group for "always" rolling low on new HD for his character. We have fun with it, but that doesn't mean I expect him rolling low next time, it's just extra fun if it occurs again. Which is still likely, since it's 50% chance he rolls below the expected value.


So what percentage of his new character HD rolls would have to come in lower than average, before you started to wonder, just a little? Or would you maintain your confidence in probability and chance as a "rule of reality" in the face of any amount of evidence?

Would you let him roll the HD for your new character?




Sure, I can understand that players want to roll dice themselves, and have no problems with letting them most of the time. The chances of success is independent of who rolls the dice however, so that itself is not an argument. Keeping players happy, because people in general are irrational, that is an argument for letting them roll though.


I don't think it's irrational, regardless of where one sits on the entire side-discussion. Even assuming that the results are 100% random, there's nothing irrational about people wanting to take the roll themselves. It's their character's success or failure, "life or death", on the line.

Pelle
2017-09-05, 09:53 AM
It's deeply insulting.
And ironic, given that it's being directed at someone who takes empirical data more seriously than oracular predictions about the future.


Ok. Following this will imply that your next die roll is not independent of the previous rolls, though. How can you explain that? These "oracular predicitons of the future" are just posing that if you roll a d6, there will be equal chances (16.7%) of rolling either a 1, 2, 3, 4, 5, or 6. Even if your current average is 3.48. How would you assign your odds differently for rolling these outcomes?



So what percentage of his new character HD rolls would have to come in lower than average, before you started to wonder, just a little? Or would you maintain your confidence in probability and chance as a "rule of reality" in the face of any amount of evidence?


He would have to roll quite many more HD rolls than he will have the chance to in this campaign, let's say that. If anything, I would start to wonder if his dice aren't properly balanced. That makes more sense to me than that he is inherently "unlucky". Equal chance doesn't mean that you will have equal amount of outcomes when it's random.

Sorry, I don't mean anything ill, but that you are asking this makes me believe that you are indeed commiting the gamblers fallacy. If not, can you explain why not? Please go ahead and read about it if you are not familiar, it's quite interesting.



I don't think it's irrational, regardless of where one sits on the entire side-discussion. Even assuming that the results are 100% random, there's nothing irrational about people wanting to take the roll themselves. It's their character's success or failure, "life or death", on the line.

It's not irrational to want to roll, but it's irrational to believe it will change the chance of success.

Max_Killjoy
2017-09-05, 10:10 AM
Ok. Following this will imply that your next die roll is not independent of the previous rolls, though. How can you explain that? These "oracular predicitons of the future" are just posing that if you roll a d6, there will be equal chances (16.7%) of rolling either a 1, 2, 3, 4, 5, or 6. Even if your current average is 3.48. How would you assign your odds differently for rolling these outcomes?


He would have to roll quite many more HD rolls than he will have the chance to in this campaign, let's say that. If anything, I would start to wonder if his dice aren't properly balanced. That makes more sense to me than that he is inherently "unlucky". Equal chance doesn't mean that you will have equal amount of outcomes when it's random.

Sorry, I don't mean anything ill, but that you are asking this makes me believe that you are indeed commiting the gamblers fallacy. If not, can you explain why not? Please go ahead and read about it if you are not familiar, it's quite interesting.


It's not irrational to want to roll, but it's irrational to believe it will change the chance of success.


Wouldn't the Gambler's Fallacy be the opposite?

Instead of expecting the rolls to follow the ongoing pattern, the Gambler's Fallacy would appear to view probability as a hard law such that future rolls "correct" for past rolls to bring the overall result average in line with the mathematical average.

I recall many times over the years, sitting there at the gaming table, looking at the dice and thinking "each new roll is utterly disconnected from all the previous rolls... and yet the pattern is a real thing, it's not imaginary... my damage rolls suck every session, even with different dice."

Pelle
2017-09-05, 11:20 AM
Wouldn't the Gambler's Fallacy be the opposite?

Instead of expecting the rolls to follow the ongoing pattern, the Gambler's Fallacy would appear to view probability as a hard law such that future rolls "correct" for past rolls to bring the overall result average in line with the mathematical average.

I recall many times over the years, sitting there at the gaming table, looking at the dice and thinking "each new roll is utterly disconnected from all the previous rolls... and yet the pattern is a real thing, it's not imaginary... my damage rolls suck every session, even with different dice."

To be fair, that's correct, but there are variations on the fallacy. This situation is more properly the reverse gambler's fallacy. It has the caveats that the dice are fair, though, which is hard to prove without doing lots of rolls. Hence suggesting getting new dice, check balance etc. Both versions of the fallacy implies that the dice have a memory of previous results, and that rolls are not independent events.


Getting streaks is normal, it is statistically unlikely (low probability) to not get them.

Max_Killjoy
2017-09-05, 11:34 AM
To be fair, that's correct, but there are variations on the fallacy. This situation is more properly the reverse gambler's fallacy. It has the caveats that the dice are fair, though, which is hard to prove without doing lots of rolls. Hence suggesting getting new dice, check balance etc. Both versions of the fallacy implies that the dice have a memory of previous results, and that rolls are not independent events.


Getting streaks is normal, it is statistically unlikely (low probability) to not get them.


Both the suggestion of "bad dice" and the mistaken notion that the dice have "memory" would seem to be nullified as concerns by the clarification that different dice didn't seem to alter the patterns.


E: Also, what about streaks lasting years? Would that be normal?

Tinkerer
2017-09-05, 02:12 PM
Just to check everyone here knows how to check a dies balance right? Epsom salt and water? It's a good idea.

Max_Killjoy
2017-09-05, 02:23 PM
Just to check everyone here knows how to check a dies balance right? Epsom salt and water? It's a good idea.

Just make sure the faces / pips aren't stickers first...

Psikerlord
2017-09-06, 12:55 AM
My solution is that instead of a PC making a skill check, to have a NPC skill check. For example, rather than having a passive perception check to detect a secret door (or a trap or ambush for that matter), I would make a Stealth check for the ambusher, the trap, or the secret door. The latter two I'd give a Stealth bonus (or penalty on one occasion, due to the politics of the adventure) based on how well hidden they were. Advantage/Disadvantage would be reversed (that is, a PC who had advantage to detect traps would cause the trap to have disadvantage on its Stealth roll), and the DCs would be 8+Skill bonus of opposed skill. So Stealth would be opposed by Perception, and Insight (to see if they were being lied to) would be opposed by Deception.

Oh yeah excellent I really like this actually, great idea!

Velaryon
2017-09-07, 11:09 AM
I almost never make rolls for the players. I'm fortunate to have players that will not metagame. If they fail a Spot check, the players may joke about it, but will have their character act accordingly. The most I'll do to keep them on their toes is occasionally call for meaningless checks even when there's nothing to be seen/heard/found/etc.

When it comes to Sense Motive, the decision to roll is usually player-initiated at our table. Whoever is DMing will usually just make a statement, and if the player wants to try and tell if they're lying, they'll make the roll. This sometimes results in PC's mistrusting a character even if they fail the roll (or if the NPC is being truthful), but I'm okay with that.

FabulousFizban
2017-09-09, 08:55 PM
any roll that reveals information on a success, but not on a failure, should be rolled in secret by the DM.

Earthwalker
2017-09-12, 02:48 AM
My general feeling on this has changed over the years.

I only really play now with friends and don’t join pick up games so I know who my players are and trust them. I mostly play games with meta currencies that allow players to control / change the dice rolls. I think the GM should never roll dice for the players.

With that in mind it does change the nature of the game and some points need to be considered.

1) The impact of failed rolls should be instantly apparent.
Fail a perception and the trap goes off or the ambush is sprung.

2) Rolls should only be made for things that are important.
If the GM calls for a roll then a success should get something, normally when possible failure should also get something.

3) Players need to know what the dice roll is for, what is the success and what is the fail result.

Dr_Dinosaur
2017-09-12, 07:29 PM
The skills players roll should ideally be active attempts with immediately-recognizable results. If the character doesn't know immediately if it worked or is passively knowing/sensing something, the DM should roll.

I like the Passive Perception idea from 5e personally, and generally apply it to Perception, Sense Motive, Knowledge, and Appraise. If something calls for a check with an immediate benefit I let them roll because that's an actual ability.

AMFV
2017-09-13, 03:32 AM
I personally am becoming very fond of the idea that PCs get passive skills rolled for them (or automatically take 10), and active skills they roll. I like the realism of that. So taking perception as an example. If a person is intentionally scouting, they'd roll their own perception, or if they're on guard duty they might roll perception checks every so often. Whereas in a situation where somebody is passively perceiving, or not actively working on it, they'd get whatever their passive threshold is as a minimum to notice things like stealthed characters.

This can apply to other things as well. Take Sense Motive (or it's equivalents). If you are deliberately paying attention to determine if someone is lying, like actively looking for signs of it, you get the sense motive roll, because it's active. If it's just somebody may be lying and you're not really paying more attention than you would in a regular conversation you'd get the passive and the DM would roll, or you'd default to the equivalent of taking 10.

If I were actually implementing this system I'd probably impose a (slight) penalty on passive rolls, like make it taking 8 rather than 10, so there's a concrete to actively doing things rather than passively doing them.

Tinkerer
2017-09-13, 02:38 PM
If I were actually implementing this system I'd probably impose a (slight) penalty on passive rolls, like make it taking 8 rather than 10, so there's a concrete to actively doing things rather than passively doing them.

I quite like that "take 8" thing. It also helps stealth PCs by offering a reward for having the enemy on low alert. I've often found it odd that RAW aside from the frequency of patrols there isn't much of a difference between a building on high alert and one on standby.

AMFV
2017-09-13, 02:45 PM
I quite like that "take 8" thing. It also helps stealth PCs by offering a reward for having the enemy on low alert. I've often found it odd that RAW aside from the frequency of patrols there isn't much of a difference between a building on high alert and one on standby.

I think that would work quite well, although I might modify the numbers more or what-not once I've seen it in practice.