PDA

View Full Version : DM Help Mystery Post-Mortem



Mars Ultor
2017-08-17, 04:45 PM
I had posted a while ago asking for opinions on an Agatha Christie-style whodunit, here are the results. The mystery was described in depth in a previous post (http://www.giantitp.com/forums/showthread.php?517289-Test-Drive-My-D-amp-D-Mystery).

After escorting an Earl to his estate, the weather turns bad and the PCs invited to stay until the storms ends. Confined in the house during the storm, the mystery begins. The earl’s wife had discovered that the steward is connected to bandits that had killed members of her family, and was still acting in concert with them. The lady’s plan was to poison the steward, then arrange the scene of the crime to indicate that someone had come into the house, stabbed the steward, and escaped. The steward died from the wine, but while staging the room the lady was forced to hide on the balcony—in the rain—while the butler came in. The butler had intended to threaten the steward, who had been blackmailing him, but finding the steward already dead, he fled, leaving behind evidence implicating him for the murder. The lady then finished staging the room and went to bed.

The other members of the family were up that night doing various things that initially suggested malice, but if questioned they would clear themselves and implicate others. Comparing alibis would eliminate most of these suspects.


I’ve run the adventure and it was a success. It took four four-hour sessions to run nearly the complete mystery; I didn’t think it would take anywhere near that long. I’ll address what caused the delays during my recap.

They arrived at the earl’s house during a heavy rainstorm (tracking in mud). They were invited for dinner and this is where I inadvertently caused a big problem: I told them that this adventure would be a mystery and they should be prepared to keep notes.

This was a big mistake and added several hours to play time. The initial dinner, which was to serve as a brief introduction of the suspects and victim, took almost the entire session. It’s impossible to get anything done when an NPC says, “pass the salt,” and six people want to Sense Motive and use Spot to see if it’s really salt.


The first session had them choosing suspects for an unknown crime before it had taken place, then the murder actually taking place, the earl requesting the party investigate the murder, the party contaminating the crime scene, destroying and ignoring evidence, and arguing over whether they should search for evidence or interrogate the suspects first.

The second session resumed with their discussion of whether to investigate or interrogate, eventually they decided to do a more thorough search of the crime scene. Finding several clues, after some discussion, they decided to search elsewhere. The party was pretty good at detecting things overall, they were specific about how and what they would search, and only missed a few pieces of evidence. Since I included multiple clues leading them to the same evidence, this wasn’t a problem.

The entire second and third session were devoted to the search, much more time than I thought, but there were extensive delays when half the part wanted to investigate and half wanted to interrogate and they spent too much time debating what to do. I feel one entire session combined was wasted because of their fighting over nothing and trying to solve a crime that hadn’t actually taken place.

Out of the six players, there was one player who took extremely thorough notes, another who took good notes, and another who seemed to note only that information which confirmed his suspicions. The other three used an unreliable system of shouting, “I found a clue! Someone write it down!” Since there were several interactions with NPCs where only one or two PCs were present, their not taking notes eventually caught up with them.

As the third session was wrapping up, someone had the idea of using the spell Zone of Truth as an interrogation aid. I had anticipated this and had prepared evasive replies and rambling answers intending to defeat the spell. The fourth session began with them choosing questions and deciding on the order of the suspects. I told them I was going to time the questioning to conform to the duration of the spell, in this case, three minutes. I set a timer on my computer, and paused the timer when they talked out of character to ask me a question or Sense Motive.

I rolled the Will saves in plain sight, both times they cast the spell. When the first Zone spell expired, the cleric cast it again, and I repeated the previous procedure. This time they saw that five out of eight suspects had rolled high enough to save versus the spell’s effects. Coincidentally, the lady—the actual killer—made her save both times.

Also, I had explained to them the parameters of the spell and suggested that they ask specific questions; vague questions would lead to vague answers. They asked several people, “Did you kill the steward with the knife?” before recalling that he was dead from poisoning before he was stabbed.

They decided that they could have asked better questions and debated confining everyone while the cleric rested and prayed for new spells. They eventually decided to just interrogate everyone. During one of the interrogations, a Sense Motive check convinced them that that one the suspects was being completely dishonest. The party decided a better course of action was to confront them with the evidence while asking the questions. “If you didn’t leave your room, how do you explain this?”

They also tried to intimidate, I explained they were still all dressed in nightshirts and sleeping caps, but they took the penalty and persisted.

After the questioning they made a timeline and eliminated everyone but the lady and the daughter, and possibly the jester. They weren’t sure on the motive for the daughter, but thought the bandit connection was the motive for the lady. There was some disagreement over whether the jester was uninvolved, an accomplice, or a werewolf. Then they seemed stumped for a while and didn’t know what to do. I suggested that they reread their notes and make connections between things they had noticed, not necessarily just physical evidence, and what they had learned. Among other things, this led them to realizing that the lady always covered her hands and suspected she had burns from the fires set by the bandits. This seemed to convince most of them the lady was the killer.

They weren’t completely sure and felt they had missed something. I thought of “The curious incident of the dog in the night-time,” (from a Sherlock Holmes story where he deduces it was an inside job because the guard dog didn’t bark) and told them there were two types of clues: The things that happened and the things that didn’t happen. They considered this and realized that part of the lady’s alibi—her dress was wet because she had gone outside to look for the jester—was false. They double-checked and confirmed that everyone else who had been outside had tracked mud in the house, this was mentioned several times, but the lady’s dress, though wet, was free of mud and so were her shoes and cloak. The ranger, who had investigated the balcony and who had also searched outside, came to the (correct) conclusion that the lady was on the balcony, hiding there while the butler interrupted her arranging the murder scene.

I feel they got a little stuck at the end, they had all the information but it needed to be drawn out of them. I had to prompt them to think about what some evidence actually meant. One player said she thought about the mystery during the week, reviewed notes, texted the other players, and finally began to put it all together. I suggested that they do a Hercule Poirot-style walkthrough of everything, and she did this (sadly, she refused to do the accent), the other players offered explanations for the gaps in her solution. They finally came up with a unified story, confronted the lady, and solved the crime.

They argued about what should happen to the lady, and some were disappointed that she wouldn’t be put on trial (the earl was also the judge). The earl said he would go public with everything and see what the people decided. The party was enthusiastic about this for the most part, only player was unhappy with the resolution when he heard the earl speak. The earl announced that his wife had discovered a spy in the household, she took matters into her own hands and killed him. He chided her for not following the law, but explained that she was eager to end the threat to her family and the kingdom, who could blame her for that?


The Good

• Everyone enjoyed themselves.
• Most of the players want to do other mysteries in the future.
• At first they figured the butler did it, then realized they had solved one crime, but not the one they were expected to solve.
• Zone of Truth didn’t ruin things.
• They all had different suspects initially, some for no particular reason.
• There were eight suspects and the players were able to keep track of them and eliminate a few pretty quickly.


The Bad

• Knowing in advance it was a mystery made them try to solve a crime that hadn’t yet been committed.
• They didn’t know if they should first interrogate or find clues, there was a long argument over how to start.
• They kept attempting to split the party and search the house independently.
• If a clue was found in one bed, they fought over who would search the next bed.
• They were inconsistent in their interrogations; sometimes they would use Intimidate and Sense Motive, other times they forgot completely.
• I couldn’t seem to recall whether the murder weapon was on the wall (yes) or in the body (no). This led to a bit of confusion as I had described it both ways and their notes were contradictory.
• Early on a player realized that an NPC kept talking about something and thought it might be important because I had repeated it. For some reason they didn’t make a note of it, couldn’t remember it, and it took a while going over things before someone recalled it. I’m not sure how this could be fixed. If the players disregard a clue over and over, after recognizing it might be a clue, there’s nothing to be done.


The Ugly

My hard drive crashed the day before the first session and I used an old computer to recreate the murder from memory and what I had posted here. It was impossible to find things sometimes and some things were forgotten or ignored.

After that session I typed out a bullet list of every clue grouped by room, and a list of suspects with their alibis, excuses, and confessions. I printed this out and could check off clues as they found them or make notes. This worked very well and I wish I had thought of it initially.

I would definitely run another mystery, but I’m going to make it less complex. It would be a slightly different format, more of a police procedural. The players would encounter the crime, be able to eliminate two of the three suspects with minimal effort, and then spend their time investigating the actual criminal in an effort to prove that he’s the guilty party.

I’m considering eventually putting everything together as a module and posting it online. If you were an initial contributor to the “test drive,” I’ll send you a copy if you wish.

Beneath
2017-08-17, 09:00 PM
Interesting! Glad to hear that it worked, and that they managed to solve it despite their bungling and despite any confusion with your hard-drive crash

Pacing, how much material can be expected to be covered in one session, is one of the hardest parts of GMing, so I'm not surprised it took way longer than estimated.

One thing I'd try is to make using Sense Motive an action; it's not a passive thing you just get when someone's lying to you, it's an active thing you do when you think someone's hiding something, and remind people that wasting sense motive checks wasting goodwill, especially if they roll poorly. Also, since you involved yourself in giving hints about how to think about it, I'd be willing to tell them, out of character, to settle down and that the crime won't be any harder to solve if they wait until it's happened to start investigating

Mendicant
2017-08-17, 09:19 PM
Nice! If you do more of these run them through here again. (And I'd definitely want a module.)

daniel_ream
2017-08-17, 10:15 PM
The first session had them choosing suspects for an unknown crime before it had taken place, then the murder actually taking place, the earl requesting the party investigate the murder, the party contaminating the crime scene, destroying and ignoring evidence, and arguing over whether they should search for evidence or interrogate the suspects first.

Here's the problem with most mystery or investigation scenarios (and that includes things like hacking, long cons, or anything that requires research or technical knowledge) in a simulationist system.


In the real world, investigating anything is tedious as ****
Most players don't know **** about this stuff
A lot of real-world investigations that aren't dead simple and obvious never get resolved


I'm steadfastly of the opinion that it's practically impossible to run a mystery scenario effectively in anything but a narrative system where the goal is to tell a good mystery story rather than for the players to solve a mystery. Dirty Secrets does this well, but it helps that in the hard-boiled P.I. noir genre the story is about the dirty secrets that the P.I. uncovers, not who actually committed the crime or even whether they get caught.

For an Agatha Christie-esque whodunit, RPG medium has serious limitations (for instance, a GM telling you things about the scene is a lousy interface between the players and the physical world and virtually guarantees lost and garbled information). Add to that that investigating a crime is a technical skill that most players don't have and aren't interested in developing, and the fact that D&D magic can both create unsolvable mysteries and solve mundane ones trivially, and sixteen hours to resolve a drawing-room murder with only eight suspects isn't too surprising.

Having said all that:



The Good
• Everyone enjoyed themselves.
• Most of the players want to do other mysteries in the future.


This is all that matters.

Some suggestions for making things go faster in future:


Be very clear with yourself and the players about what kind of mystery you're running. A forensic procedural is very different from a drawing room mystery or Columbo episode (specifically, the former is about gathering physical evidence, the latter is about interrogation and tripping up suspects with psychological warfare).
Look at systems like GUMSHOE and consider just giving the players the clues instead of making them roll. Yes, you can mitigate this by just flooding them with clues, but rolling until you find one of the clues takes up time. You know they're going to get the clue, so do they. Just give it to them.
Teach your players the basics of investigation: establish motive, means and opportunity for each suspect, create a timeline for the crime and each suspect, look for contradictions and coincidences. This includes taking proper notes.
Keep your mysteries simple at first. Bait-and-switch mysteries like the one you ran or lots of red herrings slow things down unless your players are very good at timelines and alibis.
Don't allow bickering. Point 3 will help with this; you want them arguing about who did it, not what to do next. It should always be obvious what to do next.
Don't be afraid to tell them what is and what is not relevant information. If they're wasting hours rolling Sense Motive on the salt, tell them there's nothing unusual about the salt and move on. This goes with point 2 - don't make them struggle to figure out whether something's a clue. They should be working to figure out what the clue means.
The players will assume that anything you tell them is important, so use that. Don't junk up your descriptions or interactions with a lot of pointless colour or detail. Only give them information that's relevant. Just the facts, ma'am.
Be very clear about what magic can do and what it can't, and how that affects the investigation. In a D&D world that assumes every other peasant is a 3rd level mage, this can be difficult; but the more mysteries you can run where you can say up front "none of these people can cast magic" the less exception cases you'll have to deal with. Similarly, the spells the players have can be ruinous if they start using them cleverly. It's unreasonable to just declare that every murder takes place in a null-magic zone, but be aware of what magic the players have access to and how that slots in to your mystery. In particular, have a plan for what to do if they solve the mystery in five minutes with a spell (or for that matter, brilliant deduction). Simply nullifying that is unsatisfying, but you can always run with the natural consequences of the crime. Somebody's dead, after all, and that means power vacuums, recriminations, revenge, secrets coming to light, all manner of nasty - and adventuresome - things.

Mordar
2017-08-18, 12:01 PM
Congratulations on the successful adventure! I'm actually glad it consumed some time and was enjoyed - not often you can get 4 sessions out of something so "simple" as a dinner party!

I do think that it also opens up all kinds of potential follow-up adventures. This story could be a great beginning to a mini-campaign or campaign arc.

Good job!

- M

MintyNinja
2017-08-18, 01:52 PM
Well done.

As a side note, the message you sent linking to this thread didn't contain an actual link. I'd recommend you resend the messages with this Link (which you can copy and paste):

http://www.giantitp.com/forums/showthread.php?533702-Mystery-Post-Mortem

Mars Ultor
2017-08-18, 05:11 PM
Here's the problem with most mystery or investigation scenarios (and that includes things like hacking, long cons, or anything that requires research or technical knowledge) in a simulationist system.

. . .

I'm steadfastly of the opinion that it's practically impossible to run a mystery scenario effectively in anything but a narrative system where the goal is to tell a good mystery story rather than for the players to solve a mystery. Dirty Secrets does this well, but it helps that in the hard-boiled P.I. noir genre the story is about the dirty secrets that the P.I. uncovers, not who actually committed the crime or even whether they get caught.

For an Agatha Christie-esque whodunit, RPG medium has serious limitations (for instance, a GM telling you things about the scene is a lousy interface between the players and the physical world and virtually guarantees lost and garbled information). Add to that that investigating a crime is a technical skill that most players don't have and aren't interested in developing, and the fact that D&D magic can both create unsolvable mysteries and solve mundane ones trivially, and sixteen hours to resolve a drawing-room murder with only eight suspects isn't too surprising.

Having said all that:

[List of suggestions]


You insist it's practically impossible to run a mystery without narrating the story, but I just did it. Most of the problems I encountered were created because I provided the setting and the crime and let the players take free rein. I didn't just have them sit on their hand while I told them a story and burdened them with extraneous details or too much background and description. To a large degree, I told them there's been a murder--Go. My complaints about wasting time and pointless debates were indicative of the problems with handing over the adventure to the players. They could search everywhere and question everyone, there was no linear trail to follow. I had included an NPC to assist them, they chose to put him in circumstance where he couldn't contribute. If I had to do it again, I'd make sure he stayed with the party and was able to stifle unhelpful debates or provide limited direction when necessary.


The salt was illustrative of the problems I encountered by letting them know in advance a crime was going to occur. It wasn't literally an hour asking about the salt, it was symptomatic of what happened overall. They were suspicious of everything and giving the third degree to people at dinner before the murder happened. I have a (potential) solution for it now, but it wasn't something I or others anticipated.

For anyone else running this or a similar mystery, I'd begin with the mystery and let the players handle the rest. "You have dinner, you retire for the evening. A scream rings out. When you rush into the hall you see all the people from last night's dinner gathered around a body." Let them take it from there. They'll want descriptions of the other guests and begin to feel them out, don't include it in advance. All the information and descriptions provided at dinner could have been handled at the players' request. Introducing the other guests as a way to get the party to know everyone wasn't helpful. It only served to make them treat everyone as a suspect ahead of time. Elvis warned me about Suspicious Minds, but I didn't listen.

Your perception of a mystery and mine are vastly different. Handing them all the clues and telling them only what's absolutely relevant turn it into a puzzle. I'm not against puzzles, but that's not a mystery. Their discoveries of the different color wax seals on the wine bottles, seeing the rope was too short, realizing that no one had heard the ransacking and nothing was actually broken were all big moments of excitement. It would have defeated the purpose of a mystery to hand them the clues and then let them accuse someone straightaway.

Regarding red herrings and irrelevant information; it was necessary to interrupt them when they fought over which room to search first, that contributed nothing. It wasn't necessary to interrupt their debates over the suspects and their guilt. Everyone enjoyed arguing why someone else's favorite son was the real killer while defending their own choice. The fact that they had no information didn't seem to slow them down. In fact, their investment in red herrings and irrelevant information led to emotional drama. One of the players had decided early on that the earl's daughter was too perfect, thus she must be guilty. Another had taken a liking to the same daughter and defended her vociferously. When "Lisa" had discovered that the daughter wasn't the killer, but had been lying to them the whole time, she was actually quite upset--she felt betrayed by the NPC. I'd take that any day over another Monty Python reference (not that I mind them).

No red herrings, no extraneous information, no open-ended lines of questioning would have meant that never occurred. The fact that "Lisa" felt the NPC had taken advantage of her trust led to her being considered a suspect long after I expected her to be eliminated. Without that reaction the party would have realized that the daughter had no real motive. Her actions and the killer's actions were somewhat entwined, they were the two most likely suspects, but "She lied to me" was enough to keep her under consideration even when the motive pointed to the earl's wife.

I don't understand your criticism of the world being lousy with mages since that's not my world, and I anticipated the use of magic. Magic wasn't an issue, the mystery was designed to minimize the use of magic in finding a solution. I didn't eliminate the use of magic, but I didn't let it shortcut the mystery. There's no mystery if you "Speak with Dead" and the corpse tells you the butler did it, but it's not unreasonable to make sure the victim was killed without the killer in sight, or in this case, not even in the room.

Mars Ultor
2017-08-18, 11:43 PM
Well done.

As a side note, the message you sent linking to this thread didn't contain an actual link. I'd recommend you resend the messages with this Link (which you can copy and paste):

http://www.giantitp.com/forums/showthread.php?533702-Mystery-Post-Mortem

The link works here, but when I paste it in a message it gets an error message. You found it, hopefully others will as well.

Mars Ultor
2017-08-19, 12:01 AM
Pacing, how much material can be expected to be covered in one session, is one of the hardest parts of GMing, so I'm not surprised it took way longer than estimated.

One thing I'd try is to make using Sense Motive an action; it's not a passive thing you just get when someone's lying to you, it's an active thing you do when you think someone's hiding something, and remind people that wasting sense motive checks wasting goodwill, especially if they roll poorly. Also, since you involved yourself in giving hints about how to think about it, I'd be willing to tell them, out of character, to settle down and that the crime won't be any harder to solve if they wait until it's happened to start investigating


There's a player in my group who doesn't like to accept any idea which isn't his. He finds a way to oppose everyone, he resists going with the group consensus. This has created a situation where most of the other players won't compromise with him even if his idea is better.

I've had to deal with the intra-party issues and he has improved his behavior, but it's a constant struggle. Several of the debate issues were caused by the party dynamic that's been created and it's something that's been difficult to resolve.

He'll be starting in a new school soon and I expect that he'll be participating rarely if ever in the future. I suspect things would have moved along more quickly without him, particularly since he never plays his character, he plays the game.

daniel_ream
2017-08-19, 12:08 AM
You insist it's practically impossible to run a mystery without narrating the story

No, I said


I'm steadfastly of the opinion that it's practically impossible to run a mystery scenario effectively in anything but a narrative system

It took your group significantly longer than real-time to resolve a one-night murder scenario. I don't consider that "effective" and by your own admission, the group wasted a great deal of time on things that added nothing to the experience.


I didn't just have them sit on their hand while I told them a story and burdened them with extraneous details or too much background and description.

That isn't what a narrative system does. If you don't understand what I'm describing here, look at some of the narrative mystery-focused games out there, like Seth Ben-Ezra's Dirty Secrets, Bubblegumshoe or A Taste for Murder.

Mars Ultor
2017-08-19, 11:17 AM
It took your group significantly longer than real-time to resolve a one-night murder scenario. I don't consider that "effective" and by your own admission, the group wasted a great deal of time on things that added nothing to the experience.

That isn't what a narrative system does. If you don't understand what I'm describing here, look at some of the narrative mystery-focused games out there, like Seth Ben-Ezra's Dirty Secrets, Bubblegumshoe or A Taste for Murder.


My expectation was that investing the crime, learning about eight suspects and the victim, and solving the mystery would take just less than three complete sessions, it was never a one-night adventure. My group wasting time had more to do with my group than with the encounter itself. My inexperience in running a mystery also contributed, that's why I posted this originally. Several people helped me work out the details, I wanted to let them know what had happened in practice.

After play testing the mystery, I'm reasonably certain I could run the same adventure in a shorter amount of time. Dispensing with the introductions and the problems that created would reduce the time significantly. Having the evidence for each room made into a bullet list instead of contained in the room description made things run easier, and including an NPC to make suggestions to keep things moving if the unusual format caused indecisiveness amongst the players. The adventure wasn't a race, it wasn't the time itself that was an issue, it was that it was wasted time. Specific, repairable things caused those delays, that's what I'm addressing.

I just ran a specific mystery, it's not some generic "Rules for Mysteries." It's a summarized campaign log, it's not me pontificating about mystery adventures and telling people they're not doing it right. D&D is not like those games, and my mystery was not a narrative. People can decide for themselves whether or not this adventure was a success. If they're interested in running a mystery in D&D my experience running an actual adventure might be helpful. I provided possible solutions for the problems I encountered, some might be applicable to other mysteries and some were unique to this particular murder investigation.

Based on the fact that five out six players want to play another mystery (the sixth liked it but prefers more combat), and that they were able to solve this one, I consider it a success. I expect that things would go much differently next time now that we've all had experience with a mystery adventure, and that the same issues probably wouldn't be repeated.

Mendicant
2017-08-25, 08:34 PM
One mechanical idea that just popped into my head that might help speed things up: giving any clue gated with a skill check a passive detection threshold. In other words, any clue that requires a roll to discover is automatically uncovered if a character meets the right conditions.

For instance, if you've got a bullet-pointed sheet with all the clues on it, you can mark clues that were missed on the first pass through a room. If that room is visited again by a character who would normally find the clue with a roll of 8 or better, they notice it. You could rejigger the necessary conditions as you saw fit depending on the clue or the scenario, but the basic idea would be to smoothly feed them relevant information without it feeling like deus-ex-dungeonmaster and without removing the value of actively investigating and using their character abilities.

You could also use this to include some more obscure skills that people might not think to roll--"the letter is fairly anodyne, but your training in linguistics causes something to jump out at you: the baron used the informal "ты" pronoun, which would be highly improper in this case unless the vizier is a close friend."

This would require you to do more work upfront and have your PC's relevant skills at hand, but done right it could make things run a bit smoother and faster.

Jaeda
2017-08-25, 10:14 PM
I'm glad you had a good time with it. It sounds like you were generally following the Three Clue rule (there should be multiple clues pointing to each revelation), which is good. It ensures that at least one of them is found and picked up on.

I'll echo DR in that I think you shouldn't be afraid of giving away information. If they can find something with their passive perception (or taking 10 or whatever), then just give it to them. This will save you some time rolling and you don't have to worry about a terrible roll causing something to be lost forever.

Even if you feel that they've learned too much, the end result is what they do with the information. They still have to catch the killer and make everyone else believe that she did it.

I'll also repeat that I think you should keep the number of red herrings to a minimum, especially when all of you are new to mysteries. The players are going to come up with some of their own. Any that you do intentionally insert should probably lead to something interesting; so-and-so isn't the killer, but she's having an affair that she's trying to cover up. It can be like a Cardassian mystery; everyone is guilty of something; the real mystery is what is everyone guilty of?

I do concur that you probably didn't need the dinner scene, but you probably do still want something to introduce the NPCs before the real mystery unfolds. You should communicate to the players though that this scene is primarily to introduce them and that the mystery hasn't happened yet.

Beneath
2017-08-25, 10:34 PM
I'll also repeat that I think you should keep the number of red herrings to a minimum, especially when all of you are new to mysteries. The players are going to come up with some of their own. Any that you do intentionally insert should probably lead to something interesting; so-and-so isn't the killer, but she's having an affair that she's trying to cover up. It can be like a Cardassian mystery; everyone is guilty of something; the real mystery is what is everyone guilty of?

This reminds me of the Neverwinter Nights module Black Thorn (https://neverwintervault.org/project/nwn1/module/black-thorn). In it, you're tracking down a devil lord's lieutenant before he can open a portal home to warn his master of a scheme against him, and he's in disguise as... someone. In one of the scenarios (it has eight) the intended solution involves you eliminating someone by proving that she's
a jewel thief trying to put some distance between herself and her latest heist

Black Thorn is a good template to follow for building this sort of thing, I think. There are very few, if any, cases where skill checks are necessary to get clues; I think there might be some in dialogue (there's a dialogue option to make up a noble family to ask if a purported noblewoman is friends with them that I think requires a high Lore skill and might have a Lore check behind the scenes to fool her if she isn't who she says she is, but you can't tell the difference between failing to fool her and her being who she says she is), and there's a chest you can pick the lock on to get access to books that may or may not have relevant info about how Gorson's powers work, but the mystery doesn't hinge on your skill checks, it hinges on you paying attention to what the NPCs say and what the physical evidence is.

Thrudd
2017-08-25, 10:42 PM
I'm glad you had a good time with it. It sounds like you were generally following the Three Clue rule (there should be multiple clues pointing to each revelation), which is good. It ensures that at least one of them is found and picked up on.

I'll echo DR in that I think you shouldn't be afraid of giving away information. If they can find something with their passive perception (or taking 10 or whatever), then just give it to them. This will save you some time rolling and you don't have to worry about a terrible roll causing something to be lost forever.

Even if you feel that they've learned too much, the end result is what they do with the information. They still have to catch the killer and make everyone else believe that she did it.

I'll also repeat that I think you should keep the number of red herrings to a minimum, especially when all of you are new to mysteries. The players are going to come up with some of their own. Any that you do intentionally insert should probably lead to something interesting; so-and-so isn't the killer, but she's having an affair that she's trying to cover up. It can be like a Cardassian mystery; everyone is guilty of something; the real mystery is what is everyone guilty of?

I do concur that you probably didn't need the dinner scene, but you probably do still want something to introduce the NPCs before the real mystery unfolds. You should communicate to the players though that this scene is primarily to introduce them and that the mystery hasn't happened yet.

I disagree with advice that suggests telling the players when the mystery is starting. What the OP learned was that it was a mistake to tell them there was going to be a mystery at all, and I agree. There should be no visible seams in the game's fabric. Introduce the characters in the same way you would introduce any NPC. Treat the lead-up to the mystery event exactly like any other day in D&D. The players should pay attention to everything you say, just as they always do. If there is something they want to know that their characters ought to have seen, you can tell them, but only if they ask.

Of course, it is weird to suddenly role play a fancy dinner involving chit chat between a bunch of NPCs when normally it would be brushed over as "you have dinner and go to bed. Now it's morning." So I would say, progress with the same level of detail as you normally do, until the players decide to start investigating. Then you can fill them in on details about the NPCs that they would have learned at dinner and afterward, only once the players decide that they need that information.

Beneath
2017-08-26, 12:24 AM
Seamless is an ideal, but it's not the only way to play, and once the cat was out of the bag that a mystery adventure was coming (it's a good idea to telegraph your adventure style anyway) then saying "we're just introducing characters here. When you need to investigate, you'll know" isn't bad.

Glossing over the cast introduction scene and then flashing back to it as needed also seems like a really bad structure for a mystery. Like, I can see why you would do it if you want to surprise your players, but if there's some level of trust then it's better to do the introduction up front.