If an NPC is telling the truth, what's the Insight DC to know they're telling the truth?

[Oofta]

Well that’s interesting. And I suppose I could posit you a half dozen “what-ifs” and that you might answer something like “that’s why ‘I roll perception’ works in all cases.” Fair enough - it does work.

But let’s say I’m DM-ing a game and you’re a player. As a table, we’ve agreed (for whatever reason; mystery, whatever) that I will make all dice rolls behind the screen for all players and NPCs alike. All the players will need to do is say what they’re characters are doing in the game world. Does/would that change how you play?

I ask this because when I first started I was 8 years old and playing with the neighbor boys, he oldest brother was the DM (and this was the old AD&D) and he kept literally everything behind the screen, rolling all checks. Also we only had one set of dice, and maybe that was a big reason he rolled everything. Later in life when I played 2nd edition onward, the ThAC0 roll and the attack matrix were player info and we did get into the habit of calling rolls instead of stating actions. Only since like 4E have I (and my tables) gone back to narrating actions. So I just wondered whether “who rolls” would change the way you might approach the game world.

Personally I wouldn't do anything all that differently. If the player says they're checking the door for traps, they're checking it for all sorts of traps. I'm also going to assume they know what they're doing and I'll give them the benefit of the doubt on the approach. If they're trying to notice something out of the ordinary or listen at the door, then they're going to catch anything a perception check would catch.

BTW, I do encourage people to take a more narrative approach, but much like the DMG talks about style when doing RP I don't care how a player states their actions as long as it's clear. I'd say it's clear at least 80% of the time and if it's not I ask for a clarification.

But this is not a new opinion for me. I've never liked the "tell me how you..." even back in ye olden AD&D days. I'm not a thief, my PC is. My PC knows how to safely check a door for traps, I don't.

Yes, you very clearly indicated that a failure for insight would be no information. This is not a consequence, it's just the status quo. Nothing has changed, so therefore no consequence.

I've said this before, you ignored it last time as well. This evasive answering is very indicative of less than earnest engagement.

You're the one so dead set that there must be a setback before a PC can attempt to do something, not me. If someone tries to climb the wall that can't be climbed, they may not realize it couldn't be climbed until they tried. If an NPC is telling the truth, they'll probably get a "they seem to be telling the truth". When it comes to traps and whatnot, there definitely can be serious setbacks. Or if they need to climb a cliff and falling could be deadly of course.

Didn't say gotchas, said I dislike using mechanics to instill paranoia. You, clearly, enjoy using mechanics to engender paranoia, via asking for approach just to instill it or answering success with not resolving the uncertainty, just smearing it around a bit and leaving it uncertain. Personally, I think a door that tries to eat someone to be damn cool. My players would, too, and they'd be able to clearly see how their approach led to getting eaten by it.

Depending on the campaign and group, yes I promote a sense of paranoia. In other campaigns, not so much. In a campaign of political intrigue, not knowing who you can trust, paranoia is just one of many tools in the toolkit.

Sometimes goals are obvious. That doesn't mean that a statement of approach is no longer needed, or that the goal doesn't exist. Sometimes, though, the goal isn't obvious. If a player is sniffing a door for perfume in my game it would be because I've already established that the scent of perfume is a marker for a thing the character cares about. If your assuming that I'd have this happen in my game at a random door, you're off base by a large margin. That declaration in my game would be a specific set of circumstances that had a clear line traced through previously established fiction to the present moment, and it would be very important that perfume is or is not on the door. You keep assuming we play as you do, with vague traps that might be on any given door, but this is not the case. You cannot evaluate it as if it is.

My point is this: If someone says "I'm a bit paranoid about this alley, I look cautiously for an ambush before I enter. Can I get a perception check?" I'm going to tell them about the trap on the ground even though it's not an ambush. If someone is using insight (however invoked) because they believe someone is lying, they're paying close attention to the person. They may pick up emotions other then deception such as fear, love, envy, any number of hints and clues not related to lying. In those cases I think the goal is kind of meaningless. If the goal is not obvious I ask for clarification.

Sure they do, to highly competent tomb raiders. There's holes in odd places, discolorations, mismatched tiles, etc. If you actually treat traps as utterly invisible until the player guesses it's time to make a check and succeeds (or invests in a high passive score to avoid secret traps), then you should take it as given we play very different games. To me, traps aren't gotchas that players can stumble into, but are encounters all their own -- they have tells and signals because I want the players to interact with them, not hide them.

As I've stated before, if people think their PCs would be paranoid about a trap they get passive checks in exchange for moving more slowly because they're being cautious. If they're particularly paranoid because they want to open a chest or jewelry box because they're objects that would be logically trapped then they can roll and use the higher of their passive or the roll. But yeah, in my games there's not going to be a neon sign. Then again doors that get used all the time aren't going to be trapped either because that would just be dumb IMHO. Obviously using passive values does mean that there will be times when someone's passive is so high they detect every trap in which case I'll just narrate it.

 

[pemerton]

I specifically called out that I would have not allowed that roll to fail.

A bit of a tangent - but what do you mean by I, the GM, would not have allowed that roll by a player to fail?

 

[pemerton]

Ok, fair point. I was scanning for the pattern I use [approach] to accomplish [goal], so in each case Perception sounded like the approach, and I read, for example, "listen for an ambush beyond the door" as a goal. If he had just said, "I listen for an ambush beyond the door" I can see that being a goal and an approach.

However, I will say that in each case the approach is awfully vague. Especially "use my senses".

I personally think this issue of vagueness vs adequate precision is a matter of taste - not arbitrary taste, but still a wide range of table variation based on local expectations, context, experience with adventure design and adjudication, etc.

Which to some extent relates to [MENTION=22779]Hussar[/MENTION]'s "pixel bitching" complaint: one table's sufficiently precise action declaration is another's overly detailed pedantry. This is also why some posters keep coming back to notions of "sufficiently detailed" decriptions, despite being told it's not about quantity of detail, nor its luridnesss, but about precision.

Think about other contexts, too. Is "I attack the orc with my sword!" enough? That's about as vague as "I listen at the door." Or does the player have to declare a more detailed combat move? Does it make a difference if the player wants to feed in a Battlemaster manouevre - eg does the player have to describe how s/he is wrongfooting the orc, or whatever?

Is it enough for the player of a bard to declare "I flirt with him to see what information he might drop"? Or does the player have to perform the actual words, mention the placing of the hand on the arm, etc?

The last time this sort of thing - ie flirting with a NPC to learn information - came up in a game I was refereeing (it was Classic Traveller), the player of the diplomat/spy, in character as his PC, issued an invitation to come back to his room. We used the Reaction mechanics (which is the closest Classic Traveller has to a CHA (Persuade) check) to determine if the NPC accepted the invitation, and when she did I then called for an Interrogation check (the closes thing in Traveller to a WIS (Insight) check) to see what the PC (and player) learnrd. The result of that check was that he had to share some info in order to get some info - ie some back-and-forth of conversation - but we certainly didn't play out the detail of the events in the hotel room or the real time of the pillow talk.

My own view, as referee, was that I had enough information - in the player's stated intent for his PC, in the issuing of the invitation, and in the (not only obvious but also expressly stated) invocation of James Bond tropes - to adjudicate the action and extrapolate the fiction in a way that generated fair consequences.

In a system or fictional context laden with earseekers, trapped doors, issues of doffing and donning head armour, etc, then "I listen at the door" may not be precise enough. But in most games that I run it certainly would be.

As far as this is concerned . . .

Now the players can say, "Ok, everybody be really quiet...I'm going to listen for deep breathing behind the doors" or maybe "I'm going to get down at the bottom of the door and take a deep whiff and see if I can catch a trace of that perfume we found."

In 35-odd years of GMing I don't recall ever having a Perception-type check declared in this degree of detail. (Maybe there's something I'm not remembering, but the fact that I can't remember it is enough to show that it's not common.) I would perhaps expect that sort of detail in a very Sherlock Holmes-y game, but I've never run such a thing.

The most recent high-stakes Perception check I can recall was about two years ago now: a PC shaman was dominated by a dark naga, with instructions to bring the mage Joachim to the naga so that his blood could be spilled as a sacrifice to the spirits. Joachim had been badly hurt and was recuperating in a room in another mage's tower; the PC was rushing to that place to try and get there before an assassin who was determined to kill Joachim. The assassin got there first, and as the PC rushed in he saw Joachim being decapitated. At about this point in play, we switched adjudication from a somewhat abstracted time scale (which had finished with the PC failing the opposed Speed check to beat the assassin to Joachim's room) to the melee combat resolution framework. The player of the PC's first action declaration was "I look around the room for something to catch the blood in, lilke a chamber pot." The player spent the appropriate resources within the action economy, succeeded on the check (which was set at a fairly low difficulty given the likelihood of there being some sort of vessel in a bedroom in the tower of a well-to-do mage) and was able to grab the chamber pot and start catching blood.

I look around the room for a vessel to catch the blood was enough to establish adjudicable fiction.

And what exactly is gained when you suspect someone might be on the other side of the door that is not covered under "I make a perception check to see if there's anything?" I might clarify in some cases along the lines of "Do you put your ear to the door?" if it matters or I just want to set a mood that's a little paranoid. Because if they don't hear an ambush, they might hear a party. Do you not tell them about the party? Because after all that's not what they were seeking. If the room reeks of alcohol are you going to tell them they don't smell it because they didn't mention they were smelling?

I don't see a reason to state a goal in most cases.

For my approach to adjudication, much more important than the details of how the PC is trying to notice things, is a clear grasp of what the player is hoping that his/her PC will achieve. So the player having a goal, which normally needs to be stated (occasinally it's implicit but quite evident), is crucial.

If the relevant fiction is not yet established then success on the check establishes the fiction in a way that conforms to the player's goal. The chamber pot example is one illustration. Here's another which, in the past, some posters on these boards have thought was unfair "gotcha" GMing, but which was fun at the time, and was done in a system (4e) which is very robust for acommodating this sort of thing:

I drew up my map similiarly, including with the side tunnel (behind the tiefling) which on my version ran down into the chasm, and the columns, stalactites, etc.

I didn't use four beholders, only 2 - an eye tyrant (MV version) and an eye of flame advanced to 17th level and MM3-ed for damage. And also a 15th level roper from MV, introduced on a whim when the player of the wizard asked, before taking cover behind a column, if it looked suspicious. (Response to result of 28 on the Perception check before adding the +2 bonus for knowing what he is looking for - "Yes, yes it does!")

On the other hand, if relevant fiction is already established, then I will have it express or implicit in the framing, so that the players can incorporate it into their goals. So no completely acontextual ambushes. (I think this would be an instance of what [MENTION=16814]Ovinomancer[/MENTION] means by "telegraphing", although the mode of telegraphing may be different - see further below on this difference.)

As far as noticing a party in lieu of an ambush, that would either be narration of a failure (because the player wanted to get the drop on the waiting assassins on the other side of the door, but instead there's a crowd of revellers who are now an obstacle to finding the assassins and getting the drop on them) or in some contexts just a bit of colour.

I think in many ways this comes back to a recurring theme in this thread: one side keeps assuming a model of "listen randomly at every door and check it for traps" or "check every NPC to see if they are lying" of play. And if that's the case, then yeah it gets a little unwieldy to describe approach in detail every time, and it also would feel like mere embellishment...a charge Oofta, for example, keep leveling.

But if the players know they don't have to listen at every door, then when they get to the door they do suspect they really need to listen to, they might have a more specific approach.

The vibe I'm getting from the "goal and approach" posts is very much one of classic, Gygaxian "skilled play" - a high degree of engagement with the fiction, where the fiction is understood primarily in terms of "engineering" or "mechanical" details (mechanical in the physics sense) - to do with locations of things, placement of things, numbers of things, amount of sweat on a NPC relative to room temperature, etc.

I would be surprised, though, if the 5e non-combat resolution system couldn't handle a game in which the engagement with the fiction is less about those sorts of details, and more about it's dramatic potential and significance. Which would mean running it more like Dungeon World, Burning Wheel, 4e, etc. In which case, to elaborate on what I said above, "telegraphing" takes the form not of "engineering"-type clues (like mismatched tiles to indicate a trap) but dramatic/thematic-type cues (like the PCs have defeated the sentries and broken into the enemy outpost, and are now moving through it, so one would expect an ambush around any corner!)

Sure. You’re more than welcome to run 5e the way you ran 3e. I’ve heard plenty of people who do, who attest that it works just fine and I believe them. I also agree with Iserith that it works better if you run it the way it says to run it. You may disagree, and that’s perfectly fine, have fun with the game the way you like to run it. But when people complain of problems running the game that I do not experience when I run the game, I don’t think it’s unfair to say, “I run the game this way, as per the advice in this part of the book, and this issue does not occur for me when I do so.”

I don't run 5e, and have no intention to do so. And I've already posted in this thread about how I would run it, which would be roughly the same as 4e, which is neither like you and iserith not like [MENTION=6801228]Chaosmancer[/MENTION], but rather is closer to Burning Wheel, Dungeon World, Dying Earth and other systems that use dice rolls to determine outcomes when the fiction arrives at a moment of decision.

Nor have I run more than a handful of sessions of 3E which, at the time, I ran roughly in the same manner as Rolemaster.

But the problem raised in this thread is equally one that could be asked about 3E (only it would be Sense Motive rather than Insight); and if your answer is a good answer for 5e, I don't see any fundamental difference between 3E and 5e in this particular respect that would make your answer not a good one for 3E. The structure of both games, in this respect, is largely the same: the GM has almost total authority over framing and adjudication; but the player has a high degree of authority over action declaration within the context of the GM's framing; and when dice rolls are involved, the method of resolution (single d20 roll modified by a bonus derived primarily from the PC sheet but also factoring in possible circimstantial modifiers) is the same.

I think it's more useful to explain the merits of an approach within this context, rather than to point to rules. Others may have read the rules and believe themselves to be conforming to them; may treat them as broad advice rather than a strict statement of play procedure (there's certainly a long history of doing this in D&D, actively encouraged by canonical game texts); or may be RPG anarchists who don't care about rules but do care about play experience!

 

[Charlaquin]

D

Reading through all your negatives, it really seems like you are trying to be ultra-precise. You will not move forward with resolution until you have an action verb that is tied to a specific sense that could be used to perceive something.

I don’t think “listen” is particularly precise. It’s about as vague as one can be while still technically stating at thing your character is doing.

I'm not that precise, and the only thing you seem intent on avoiding that this may help with is avoiding upsetting your players by resolving something based on an action they might have done. Which is a valid concern, but is not my primary concern when running the game.

It’s not my primary concern either, it is one benefit among many of the technique. The main reason I do it that way is to keep the game focused on the narrative.

I think we are getting mixed up with our "uncertainties" again.

A cleric can always attempt to heal a soul with channel divinity. It might not work because the damage was caused by a specific threat that left wounds you aren't powerful enough to heal. It might not work because you are standing in the dominion of a being opposed to your deity and your power for miracles is diminished here. Maybe it doesn't work because you couldn't handle the strain of channel so much divine power.

The world (at least as well as I can make it) isn't inconsistent. But it isn't spelled out all the time. A paladin is immune to disease, but a magical disease contracted by the paladin almost breaking their oath in the usage of an artifact of Orcus... maybe they aren't immune to THAT.

Demons are immune to poison, but are they immune to the Cosmic Serpent's venom?

It doesn't make for an inconsistent world, just one larger than the books laid out.

Ok... I’ve lost the thread of this part of the conversation.

That wasn't what I was saying when this line of discussion came up.

I just said not all skill checks need to be rolled, because sometimes a player's abilities would change the roll or negate the roll.

No, all skill checks need to be rolled because rolling a d20 and adding an ability modifier (and potentially a proficiency bonus) and trying to beat a target number is the definition of a skill check. If you’re not rolling, then a skill check is not what you’re doing.

If a rogue has reliable talent they cannot roll below a 10. That means their minimum for a check is X. If X is higher than the DC you might not call for a roll. But, Reliable Talent only works for skill checks. If you refuse to call an action that doesn't require a roll a skill check (sorry, ability check, skill checks don't exist per the rules either) then Reliable Talent can't guarentee success.

If we want to get pedantic, a Rogue with Reliable Talent still rolls the die, they just change the result to (10 + Ability + Prof) If the die comes up less than 10.

And also, as a side of that conversation, you seemed very concerned that I might know that certain likely actions a player may take would require an ability check. For example, just this morning I had someone roll athletics to tear down a steel vault door. The barbarian player likes destruction like that, so when I put the door their I figured they would try and break it down and it was solid enough they might not be able to. We also just got a rogue, it was a well made door, so trying to pick the lock would be a skill check as well. Knowing those very likely scenarios exist, and that they would likely require skill checks isn't a problem. Yet, it seemed to bother you and this line of discussion spun off.

I have no idea what you’re talking about.

No, like I said, we've moved far beyond the premise of this thread. But, if you are going to get upset at the assumption that a skill check might be called for in a thread that started off about how to resolve skill checks... maybe you should double check your assumptions.

No one got upset at the assumption that a skill check might be called for. A few of us answered the question, saying “I wouldn’t call for a check at all, so no DC,” and then a couple of people got upset that we said that had some honest questions about why we choose to run the game that way. In answer to which the rules were quoted, and... well, here we are.

Yeah, which is why in that example I specifically called out that fact, and said this was only an example of what I would do if for some bizarre reason we decided to call for a roll anyways. I incorrectly figured that would be enough to make it clear what I was doing.

”Obviously I wouldn’t put mustard on my hot dog, but if for some buzzard reason we decided to put mustard on them anyway, I would put Dijon on mine.”

Who cares what kind of mustard we would use “if” we put mustard on our fries, if we all agree we don’t want to do that?!

I know what the phrase means.

The point I'm trying to defend... do you even know what it is?

I specifically called out that I would have not allowed that roll to fail. So the point I am defending is I DIDN'T SAY THAT.

You kinda did, though. Again, you claim you wouldn’t call for a roll in that situation, but you are advocating hard for the proper way to narrate the failure “if someone did, for some reason.” If a roll shouldn’t be called for, than all ways to narrate the failure are improper, because it is not proper for the action to fail in the first place.

Actually, dijon mustard is great on fries.

-.-

So, your entire problem with my statement is that I said "What you did wasn't right, but if we did it, we would at least do this part better"

Strange hills indeed.

Not exactly. My problem was with the way of narrating the failure you claimed was better. It wasn’t better, it was flawed for exactly the same reasons that the ruling in the example was flawed - namely, that arriving at it would still have required calling for a roll in a situation where the outcome was not uncertain. It was no better a call, it was just a more flowery way of making the same bad call.

 

[Charlaquin]

I think it would have been better if you have written some thing like, 'I/We believe for my/our playstyle that it works better..."


Just saying 'it works better...'


See this way it comes across as your opinion instead of statement of fact that the game is better when run 'by the book'.


But it could be that I miss read what you were trying to convey.

Alright, that’s fair.

 

[iserith]

Alright, that’s fair.

Sit there and patiently fend off absurd assertions and purposeful obfuscations about the manner in which you play the game for weeks on end... but don't you dare use words that someone can take to mean you think the way you play is better than some other way, even though that's obviously why anyone would choose to play the way they do.

Sounds fair.

 

[Guest 6801328]

Gawdz....I was about to start doing some mass replying to posts in this thread, but I'm not sure I have the energy for any more.

 

[Ovinomancer]

You're the one so dead set that there must be a setback before a PC can attempt to do something, not me. If someone tries to climb the wall that can't be climbed, they may not realize it couldn't be climbed until they tried. If an NPC is telling the truth, they'll probably get a "they seem to be telling the truth". When it comes to traps and whatnot, there definitely can be serious setbacks. Or if they need to climb a cliff and falling could be deadly of course.

Exactly! Your play is different. You have a different set of assumptions as to what an action entails, and a different way of adjudicating them. This is perfectly fine, but it's not the same way that a goal and approach method uses. You really need to accept that this is so and stop trying to judge the method from how you play and instead try to understand how it's actually used.

Depending on the campaign and group, yes I promote a sense of paranoia. In other campaigns, not so much. In a campaign of political intrigue, not knowing who you can trust, paranoia is just one of many tools in the toolkit.

I promote a sense of paranoia, but I don't use the mechanics as the means to do so, I use the fiction in play. I don't need to be vague about a success, with and answer of 'they seem to be telling the truth' because that's not needed -- I have plenty of other tools to inflict paranoia on my players.

This is my point, the difference in our play is that I do not see the mechanics, either in success of failure, as a place to make the player uncertain of outcomes. Those are where outcomes become certain. I get to play with the before and after.

So, again, our play is different -- we're prioritizing different things, and this means that you cannot judge my play by situations that occur in your method because they're not the same situations as in my method.

My point is this: If someone says "I'm a bit paranoid about this alley, I look cautiously for an ambush before I enter. Can I get a perception check?" I'm going to tell them about the trap on the ground even though it's not an ambush.

In my game, this player would be doing themselves a disservice because asking for a check is asking for a chance to fail, and failure has consequences that are not the status quo. That's my point -- if you do not add consequence to every check, and, indeed, only ask for checks when there is a consequence (and a chance for success/failure), then asking for a check makes perfect sense -- it's the only way to get the GM to divulge their hidden story to you. I do not play this way. My method does not work in your method of play. This should be obvious, but I keep having to say it.

If someone is using insight (however invoked) because they believe someone is lying, they're paying close attention to the person. They may pick up emotions other then deception such as fear, love, envy, any number of hints and clues not related to lying. In those cases I think the goal is kind of meaningless. If the goal is not obvious I ask for clarification.

This is because your point of conflict is "is this NPC lying to me." That's, frankly, utterly boring to me.

If I present a lying NPC, figuring out the NPC is lying will not resolve whatever the actual issue is. It will just lead to a new point of contention. Why did the NPC lie? What do we do know that we know the NPC lied?

To go back to the shopkeep example you proposed, determining that the shopkeep lied would never be a check in my game. I'd never need to prevaricate to preserve uncertainty so that my plot continues. Instead, discovering the lie is just one more means to advance the plot and do something different. You'd need evidence, and could then brace the shopkeep with it to expose the lie and get the truth (which leads to more adventure), or maybe you engage in discussion, discover something about the shopkeep, like that he loves his little girls, and use that to get him to confess to the lie. Or, maybe, you do not, and have to come at the problem a completely different way. To me, discovering a lie is just like opening a door -- something you have to do to move the game along. As such, if it's uncertain, there will be a consequence to failure that will change how the fiction sits -- the status quo will not hold. On the other hand, a success is a success -- the character reaps the reward and I don't try to diminish the success. Why would I? The character just took a risk I'd hammer home on a failure, so a success deserves nothing less than actual success at the intended goal. Or, for complex goals, a solid step forward.

As I've stated before, if people think their PCs would be paranoid about a trap they get passive checks in exchange for moving more slowly because they're being cautious. If they're particularly paranoid because they want to open a chest or jewelry box because they're objects that would be logically trapped then they can roll and use the higher of their passive or the roll. But yeah, in my games there's not going to be a neon sign. Then again doors that get used all the time aren't going to be trapped either because that would just be dumb IMHO. Obviously using passive values does mean that there will be times when someone's passive is so high they detect every trap in which case I'll just narrate it.

That's fine, but it's also why you need to have the "shortcut" of letting players ask for rolls and why you don't seek approaches -- there's no change if they fail and they can only benefit (maybe vaguely) on a success. You've built your game around the idea that asking for checks is what's what, so that behavior is prioritized. This is not a shortcut, or even a good idea in my game, because rolls will change the fiction -- for the better on a success and for the worse on a failure -- so it's better to seek to not to roll. This is accomplished by providing an approach and goal so the GM has the best information possible to determine you might automatically succeed or, if it's going to be a roll, that you get the best possible chance by leveraging your character's abilities to the maximum extent possible. And, a good approach might net you advantage!

This difference -- rolls change fiction -- is absolutely a huge difference in our game. If you ever say, "nope, you don't find any traps," and nothing else on a check looking for traps, then this is a huge difference in our games, and, indeed, why our methods differ. This is not something that is ever said in my games. Instead, it's, "<sharp intake of breath> ooh, that's not going to be good."

 

[Ovinomancer]

The vibe I'm getting from the "goal and approach" posts is very much one of classic, Gygaxian "skilled play" - a high degree of engagement with the fiction, where the fiction is understood primarily in terms of "engineering" or "mechanical" details (mechanical in the physics sense) - to do with locations of things, placement of things, numbers of things, amount of sweat on a NPC relative to room temperature, etc.

I'd strongly disagree with this. It's just making sure that the character is concretely doing something and having the player tell me what that is. Gygaxian skilled play is about players beating the puzzle. I run much more of a narrative game, where player declarations will change the fiction in their favor on a success or against them on a failure. However, some declarations are so obviously fitting that they succeed. We can start looking at this by way of a declaration to walk across a room absent of any dangers. This is not a point at which anyone would ever call for a check of any kind to see if it succeeds. The goal and approach method just requires that the player put their character into the action in a concrete way -- that they actually declare an action, not a mechanic. This is contrasted by the method of allowing players to request mechanics without concrete actions. The actual action declared just has to be reasonable for the goal stated, not precisely tuned to the GM's notes. Any reasonable action will at least get a check. Blatantly unreasonable actions might automatically fail. Very apt actions might automatically succeed.

The example of the poisoned doorknob, for instance. It's been established in the fiction previous that the doorknob is coated in contact poison. Many possible actions could be taken, and any reasonable action will get at least a check. But, if you leverage the established fiction (not the GM's secret fiction) well, you might autosucceed. For instance, if your answer to the poisoned doorknob is to use an axe to chop down the door, there's no check to bypass the poison (although one might be needed to chop down the door, depending). If you use a cloth to wipe the door down, and your character has training in poisons, then this might also be an automatic success. If you do not have training in poisons, I'd definitely call for a check. Or, maybe the player comes up with something very different, like using a decanter of endless water to wash the poison off of the doorknob. Engaging the established fiction is the key, which Gygaxian skilled play was more about predicting the GM's secret notes by experience of how the GM prepares traps.

I write problems down in my prep, not solutions, and rarely even mechanics, preferring to determine these things in play according to checks. I can't predict how the fiction will go, so I don't, and instead prep more generic problems that can be quickly adapted to fit the current fiction when and if needed. This isn't illusionism because I'm not always going to put my prep in front of the players -- but I will use it if it's warranted. A more generic prep helps here. Of course, 5e is a game that expects prep, and many of it's systems do not work well if you try to push too far to the story now methods of play, so it's always a balancing act. I find that, to run a more reactive game, the level of system mastery must be very high. You must understand how the system balances things over time and be able to place challenges that fit this balance. This was MUCH easier to do in 4e because it's balance point was the encounter whereas 5e's balance point is the adventuring day. So, running a game more narrativistly while having to balance across multiple challenges requires a very strong grasp of the system and how it works. Gygaxian skilled play wasn't concerned with any kind of balance like this.

 

[iserith]

This difference -- rolls change fiction -- is absolutely a huge difference in our game. If you ever say, "nope, you don't find any traps," and nothing else on a check looking for traps, then this is a huge difference in our games, and, indeed, why our methods differ. This is not something that is ever said in my games. Instead, it's, "<sharp intake of breath> ooh, that's not going to be good."

That's so true. Last night, my Eberron session kicked off with an important social interaction challenge that was pivotal to their success on the overall adventure. They needed to team up with NPCs who were hostile toward them without letting them know their true motives for helping (which would undermine a kingdom's plot to overthrow the government of another nation). I prepared the central NPC's ideal, bond, flaw, and agenda, and fleshed out the personality traits of the other four NPCs. I then presented a series of objections to what the players were putting forward. Overcoming those objections would slowly improve the NPCs' attitudes. Once the "final attitude" was resolved, then came the ask - could they work together?

They overcame some objections handily, gaining automatic success. But some weren't so clear cut and I called for rolls. Every single one was filled with tension as the mostly un-charismatic PCs considered what arguments to make and what limited resources to apply to improve their odds. You'd think they were trying to avoid disintegration traps.

In the end, they succeeded in convincing the NPCs to team up. It was as tense and interesting as any combat. And being an alliance built upon a lie, we'll see how this goes awry soon, I'm sure!