Teaching Players To Be Better Detectives

[Blade]

Many detective stories (books, films, series, etc.) are made with the idea that the reader should be able to find the explanation before the reveal, even though it shouldn't be too easy.

That's actually what makes a detective story good for some people. You'll see many bad reviews/opinions when the detective comes up with the solution because of a detail that was never revealed to the readers, or because there are too many possible explanations and you'd have to be in the writer's head to know which one is the right one.

So it IS possible - though difficult - to design a mystery that will be solvable and also challenging. It is a bit more difficult in the case of a RPG than in the case of a book, since players can walk past the clue without seeing it, or can do something that would prevent them from being able to solve the mystery.
A good move is to make it "too easy". You'll be surprised how difficult it will turn out to be.
Another one is to prepare a list of possible extra clues and extra complications you can use to adjust difficulty on the fly, if the players seem to be lost or too find it not challenging enough.

And as Chandler wrote, "when in doubt have a man come through a door with a gun in his hand."

[Lago PARANOIA]

I'm with a lot of posters in this thread. If you want your players to become better detectives then you need to have a good framework for them to act in. TTRPGs in general have the eternal problem of 'what my character would do if they were independent, autonomous characters in a fully realized universe is different from what they would do if I was dictating their every command' but in no genre is that as troublesome as with detective stories.

To that end, I would start by reading the various 'rules of writing a detective story'. I don't agree with everything on the Knox/Van Dyne lists, but they make a good starting point because they eliminate objectively plausible but narratively unsatisfying possibilities. And that's really, in my opinion, the key to making a good story -- and thus making better detectives out of your players. The best way to do this in a TTRPG of course is to make it so that all roads, while varying in difficulty, can still lead to victory. And a big part of making that otherwise foregone process satisfying is by making players nonetheless feel that there are roads that won't lead to victory.

So in addition to the standard Van Dyne list, here are my own suggestions. They come from playing or watching plenty of Ace Attorney / Scooby Doo / TV police procedurals / actual homebrew TTRPG mysteries and are informed by my own subjective experiences. Nonetheless, I hope that they'll help:

• Obeying the Law of Conservation of Detail is always a good idea for any kind of story, but it's absolutely vital for a mystery. Most of the time when a 'fair play' detective story stalls out it's because the writer or GM ran afoul of this rule somewhere along the line. If you're writing a mystery and you haven't at least had one reverse-brainstorming session where you had to cut cruft and simplify plot points, it's doubtful that you put any effort into it at all.
• A fruitful investigation must never be allowed to be permanently aborted or undone because of a lack of authority. If there's a potential for an investigation to be halted due to the actions of a government / criminal conspiracy / etc. then the detectives must account for this roadblock and have a way to continue and resolve the investigation. TTRPG EXCEPTION: Doing this is less problematic in a TTRPG in terms of player satisfaction, though be aware that the typical methods of subverting hostile authority may cause an unwanted genre shift.
• Anyone connected to the main culprit by one or two degrees of separation must always have a personal stake in the mystery. As improbable as it may seem, the bartender should ideally not be merely a bartender nor can the security guard just be merely a security guard. If you must use one of these utilitarian characters then their involvement should be absolutely minimal. Even if the character is a career or professional criminal their involvement must still be personal in some fashion i.e. killing this particular target is a matter of family honor for the assassin or this impartial burglar knew the victim as a child. TTRPG EXCEPTION: This rule is probably impossible to hold in a TTRPG with sufficiently motivated or curious players; you may be expecting the players to investigate at the workplace or family home and they might decide to focus their investigative efforts at the restaurant or the church. Nonetheless, you should still plan for a web of 'deep in' people.
• The mystery should not be able to be solved exclusively with existing clues and scientific investigation, with no need for interviews or interrogations. At the absolute minimum, the detective must interact at least one person for insight into the case. TTRPG EXCEPTION: This will nonetheless sometimes happen in an actual game. Don't cockblock your players if this happens, just take that as a silent admonishment to plan your mysteries better.
• If your story must use secret societies, mafias, government conspiracies, etc. in such a way that the organization provides the culprit assistance, it is only allowable if the detective is directly gunning for the ringleader. In those cases, the resources of the organization become a knowable (if formidable) obstacle in solving the mystery rather than a Swords of Damocles-ish reverse deus ex machina. In other words, while the detective is allowed to target the don in a murder mystery, it is literary malpractice for them to target a mere capo. If you are going after someone lower in the food chain, their crime must either not be connected to the organizational structure at large (i.e. a professional assassin killed a hitchhiker for their winning lottery ticket) or the detective must be allowed to work their way up the food chain until the leader goes down.
• A story is allowed at most one unforeshadowed complication not related to the detective's investigative skills after a significant number of leads are gathered, preferably as far away from the denouement as possible. If your story requires multiple complications such as a blackout AND a riot, they must occur before the investigation starts making serious progress. TTRPG EXCEPTION: If the players cause the complication by going totally out into left field and/or just being screwed by the RNG, that's okay. However, preplanned complications should be kept at a minimum.
• Any red herrings or false/misleading evidence must be resolved by the summation. A reader should not be forced to guess at the moment of truth which contradictory clues are relevant and which are not, even if the weight towards one conclusion is overwhelming. It's fair, if feeble writing to have unexplained clues that point in the direction of the true solution if there are enough explained clues to solve the mystery.
• Real-world investigative anachronisms, such as the use of chemical analysis or radios before they became an official part of then-contemporary popular culture, are not to be used by the detective if the culprit is unaware of these techniques. It's fair for a criminal to be anachronistic in this way as long as the mystery is solvable with this advantage and/or the detective figures out the technique. A 19th-century genius criminal who uses polonium to kill is fair game as long as the deed can still be traced. A 17th-century genius prosecutor who springs the surprise of fingerprinting on a culprit is cheating. TTRPG EXCEPTION: This is okay to do with actual players in a game as a reward for cleverness or creative use of abilities, as long as they're not metagaming. In fact the absolute worst thing you can do is break the fourth wall to punish players for being too clever.
• A mystery does not need to have a corpse at the end. However, the mystery must have a payoff at least as interesting as a murder. If the crime was particularly interesting or ingenious even a petty theft can be suitable fodder for a mystery. To this end, a murder is always safe as a payoff and the writer should never be ashamed to fall back on a murder. In fact if the story is serialized it's preferable for the majority of mysteries to be murders so that the occasional saboteur ring or hostile corporate takeover stands out.
• A murder turning out to be a suicide or an accident is fine as long as the collateral damage is equal or greater to that of an ordinary murder. If two feuding families destroy each other in a cycle of revenge because they believe that a patriarch was murdered when they really died of a messy heart attack it can make for a moving and ironic finale.
• If a serious crime must have multiple culprits, it behooves you to give them as different motivations as possible. Three men who kill a relative as part of an honor killing will never be as interesting as one man who does it as an honor killing, another one does it to escape the victim's blackmail, and another who does it to advance a political agenda. TTRPG EXCEPTION: This is an especially good idea for open-ended games as it makes the route to solving the mystery feel less ad-hoc and instead feels more like a reward for cleverness.

[hyzmarca]

Clues are stupid. The very word clue suggests a mindset which is anathema to actual detecting. You do not want to give you're players clues. That way lies madness.

What you should do is give them evidence.

The difference is quite important.

A clue is something that is supposed to point the players in a specific direction. But they might not interpret it correctly. Evidence just is. It isn't meant to be interpreted a certain way, and is just detritus left behind by what events actually happened.

A bad mystery GM builds his ending first, and then puts in clues, and then builds his beginning.

A good mystery GM starts with the crime, goes over the details with a fine toothed comb, and figures out what evidence is left behind. And he understands that he's not omniscient and so is willing to generate new evidence that he hadn't thought of if it's pointed out to him.


Second, actual detective work is boring as hell. 99% of it is the shotgun aproach. You go over everything with a fine tooth comb, without exception. You take to everyone, without exception. These are things that players should not have to endure in a game.

It is perfectly alright for a detective player to just declare that he orders the uniformed cops to canvas the neighborhood and for the CSIs to check the entire room and for the GM to come back with every witness and every piece of evidence.


The problem isn't that the players are shitty detectives. The problem is that your stories are linear adventure paths in which the players are not told where they're supposed to go and have to guess.

But a detective game absolutely can't be linear. In a linear mystery path real detective skills don't matter at all.

[kzt]

When I realized that important characters almost never die in movies from that genre, I did not suddenly magically find the whole genre boring.

Though it did shock the hell out me in "To Live and Die in LA".

[8d8]

I feel Schwartzy's pain, even if he gave up and left in a huffy sulk. In a recent game one of the players was looking for Mark's favorite whore. He overheard a conversation in which someone mentioned that Mark's brother knows his every move and manages his life for him. The player then decides that, instead of talking to Mark's brother or finding a place where whores might frequent, he'll go down to family day at the old farmer's market and ask everyone there if they've seen a whore with Mark and how to find her.

Some people are either stupid, or oblivious to the obvious. I would like to know more about how to train players to think through a situation better, allowing clues and cues and evidence and facts to inform their situation and how they deal with it.

I probably didn't start trying to think about an rpg situation as my character until I saw someone else do it and realized I'd been treating it all like a video game. And I didn't start trying to solve mysterious situations until I saw someone solve one just by reasoning through it. This is the totality of what I "know" about this topic and I want more understanding.

[hyzmarca]

Another thing to remember is that the characters are often smarter than the player is. If you're playing a game of D&D, you don't expect the wizard's player to cast an actual working teleportation spell at the table every time he wants to teleport somewhere. The player's actual magical skills have no bearing on the character's success. If you're playing Riddle of Steel you don't expect Conan's player to pick up a greatsword and physically duel a real live necromancer every game. The player's sword skills have no bearing on the character's success. If you're playing Sherlock Holmes, why the hell would you expect the players to be as smart as Sherlock Holmes?.

[Occluded Sun]

Even Conan Doyle wasn't as smart as Holmes was supposed to be, resulting in the Great Detective having some rather peculiar screw-ups. Given that CD believed in Spiritualism and fairies...

Holmes' analytic abilities are more informed than demonstrated. He has long chains of 'reasoning' which often either don't make any sense (but work anyway because the author creates the world they're in) or rely on evidence-gathering the reader can't match, such as reading an old friend's facial expressions and guessing what the things he's looking at are making him think of.

[MGuy]

When I realized that important characters almost never die in movies from that genre, I did not suddenly magically find the whole genre boring.

Though it did shock the hell out me in "To Live and Die in LA".

I find stories where I don't know if a character will live or die more compelling than ones where I know where they don't.

[Username17]

I feel Schwartzy's pain, even if he gave up and left in a huffy sulk.

I was on the fence about whether to be sympathetic or not. I mean, running a detective game in a table top RPG format is legitimately difficult. You're running a cooperative storytelling game where the contributors are also the audience and the intended feel is one of mystery. There's a narrow line between the audience not knowing what's going to happen and the contributors not knowing what to add. But he also closed with this aside:

Oh, generally, I'm not a fan of the "make a perception roll to see if you look in the right place" approach to investigating a crime scene. If I know something is under the bed, for instance, I'd rather award that clue to the PC that specifically says "I look under the bed" than the one with the best perception roll who isn't especially specific about where in the bedroom they're searching.

...and that's terrible. I mean, it's really terrible. Let's break that down. He is not a fan of a character's actual abilities contributing to their ability to advance the story, but he is in favor of asking players to guess the secret numbers he is thinking of as a gateway to story advancement. Holy crap. That alone could account for absolutely all of his players' purported tendency to stall out, because that's totally unfair bullshit that stonewalls people for no reason other than being on a different wavelength than the MC.

And of course, when people called him on specifically that, he took his ball and went home. Which makes me think that the condemnation was exactly right. You don't get that kind of butthurt because someone misinterpreted something you said.

-Username17

[kzt]

Yeah, it's like running a game in which the player is playing the world's greatest sniper and then asking the player how they take a shot at a guy 1500 meters away. Then after they roll a solidly successful hit you tell them "You miss by 6 inches, because you didn't say you accounted for the Coriolis effect!"

Fuck that.

[erik]

Clues are stupid. The very word clue suggests a mindset which is anathema to actual detecting. You do not want to give you're players clues. That way lies madness.

Unless you make it an abstracted minigame where your party accumulates clue tokens and can trade them in for knowledge to eliminate suspects or confirm suspects and the like.

Like you get a clue indicating crime was within a time period which some people have solid alibis. Or turn up clues on motives. Get enough to create a profile. I suppose you could call them evidence tokens or pips, but clue sounds better to me. A clue is just evidence in context.

It is perfectly alright for a detective player to just declare that he orders the uniformed cops to canvas the neighborhood and for the CSIs to check the entire room and for the GM to come back with every witness and every piece of evidence.

I was with you until the part about the GM coming back with evidence. That's yuck. Ignore evidence. Just gloss over that shit and say, "you achieve 1 clue from your efforts".

Go around doing gather reports, crime scene forensics, squeeze informants, autopsy, follow-up surveillance footage. Once you narrow it down or want something more solid on an individual you could set tail, apply pressure on suspect or friends, investigate their finances. Can even have illegal or questionable tactics which can backfire if inappropriate or just botched.

This gives actions that are fairly abstract; you don't have a stupid guessing game, and you don't actually have to role play the interrogation of countless witnesses or going through oodles of pages of credit statements.

I'd be tempted to use this except it has no place in my Rifts project (mysteries are solvable via psychics and magic; bad guys are served with violence), nor in my slower-moving sci-fantasy heart breaker project.

[hyzmarca]

I was with you until the part about the GM coming back with evidence. That's yuck. Ignore evidence. Just gloss over that shit and say, "you achieve 1 clue from your efforts".

Solving the mystery is part of the fun, though. Full abstraction takes away from that. The important thing is to abstract the boring parts. Stupid guessing games can be avoided by giving the players solid evidence and not making the mystery stupidly "clever."

[violence in the media]

It is perfectly alright for a detective player to just declare that he orders the uniformed cops to canvas the neighborhood and for the CSIs to check the entire room and for the GM to come back with every witness and every piece of evidence.

I'm sort of with you on this, but I'm curious as to how many hours or game sessions a mystery is supposed to occupy in this scenario. Cop shows mine everything you just described for like 20-30 minutes of air time every week, and use various failure points in those processes to create drama--e.g. "We can't find this witness we want to question." However, once all the evidence is processed and witnesses interrogated, the mystery is pretty much done and they just need to round up the bad guy. So, unless every crime scene is spotlessly clean and there are no witnesses to talk to, what are the players actually supposed to be doing?

Feel free to expand this conversation to a general discussion about what actions in a game should take table time and what ones shouldn't.

[RobbyPants]

In a recent game one of the players was looking for Mark's favorite whore. He overheard a conversation in which someone mentioned that Mark's brother knows his every move and manages his life for him. The player then decides that, instead of talking to Mark's brother or finding a place where whores might frequent, he'll go down to family day at the old farmer's market and ask everyone there if they've seen a whore with Mark and how to find her.

Some people are either stupid, or oblivious to the obvious. I would like to know more about how to train players to think through a situation better, allowing clues and cues and evidence and facts to inform their situation and how they deal with it.

Make the clues increasingly obvious?

Player: I'll go down to family day and question random people.

MC: None of them know anything about this whore, but several people say that they know Mark's brother knows his every move and manages his life for him.

Player: Hmmmm. I'll try down by the docks. Sailors know lots of shit!

MC: The sailors say that Marks brother can almost certainly tell you what you need to know, because he manages his whole life, and would be privy to this information.

Player: I'll go to the temple and see if anyone confessed anything to the clerics.

MC: All of the clerics are busy, and there is a line to talk to them. Standing right in front of you in the line is Mark's brother. You probably have an hour or more to kill waiting in line.

Player: I don't want to pester him with small talk. I stare at the pattern on the floor.

MC: Mark's brother starts chatting with you. He says "Man, you know what's a pain? Managing all the details of my brother's life! He recently got involved with this whore and it's been making things difficult. I told her to leave him alone, and she went to her place, that I happen to have the address to.".

[TheFlatline]

If you want detectiving to work in an RPG format, you have to have the whole whodunnit nonsense be totally fluid. The players find clues wherever they think to go, many of which are made up on the spot. And then you listen to the players hypothesize based on those clues and adapt your fucking storyline to fit the coolest direction the players suggest.

This is *exactly* what I talked about in my "how to run a conspiracy" essay.

Although when I ran it in Dark Heresy (CoC with rocket launchers and Orks) I came up with an actual conspiracy and had the basics of it flushed out so I could pick appropriate clues to give to the players. But a lot of it did change based on arguments/conversations the players had and ideas they came up with that were too good *not* to use.

You have to change your basic, underlying approach to a mystery. You don't want to say "no" to the players, you want to say "yes" to anything creative.

[TheFlatline]

In a recent game one of the players was looking for Mark's favorite whore. He overheard a conversation in which someone mentioned that Mark's brother knows his every move and manages his life for him. The player then decides that, instead of talking to Mark's brother or finding a place where whores might frequent, he'll go down to family day at the old farmer's market and ask everyone there if they've seen a whore with Mark and how to find her.

Some people are either stupid, or oblivious to the obvious. I would like to know more about how to train players to think through a situation better, allowing clues and cues and evidence and facts to inform their situation and how they deal with it.

Make the clues increasingly obvious?

Player: I'll go down to family day and question random people.

MC: None of them know anything about this whore, but several people say that they know Mark's brother knows his every move and manages his life for him.

Player: Hmmmm. I'll try down by the docks. Sailors know lots of shit!

MC: The sailors say that Marks brother can almost certainly tell you what you need to know, because he manages his whole life, and would be privy to this information.

Player: I'll go to the temple and see if anyone confessed anything to the clerics.

MC: All of the clerics are busy, and there is a line to talk to them. Standing right in front of you in the line is Mark's brother. You probably have an hour or more to kill waiting in line.

Player: I don't want to pester him with small talk. I stare at the pattern on the floor.

MC: Mark's brother starts chatting with you. He says "Man, you know what's a pain? Managing all the details of my brother's life! He recently got involved with this whore and it's been making things difficult. I told her to leave him alone, and she went to her place, that I happen to have the address to.".

Why go through all that? You've lost the party by the second scene. It's obvious they don't want to follow Mark's Brother The Clue. That doesn't interest them. Sure Mark's Brother could give them context to the mystery, but by skipping that clue and doing something else, the plot will have holes in it looking back. And that's not a bad thing. Just don't make Mark the only source of game-dependent information.

Have them go to market day and the perfume maker remember making a special perfume for the rather pretty lass glued to Mark's arm that smelled of dragonsblood and elderberry. Give a basic description. Maybe suggest where the whores hang out or where the local house of ill repute is. Maybe they'll go to the whorehouse. Maybe the bard will start flirting with all the working girls trying to get a good sniff without spending (too much) money. That'd be a great montage scene.

The key is to just roll with it.

[TheFlatline]

I feel Schwartzy's pain, even if he gave up and left in a huffy sulk. In a recent game one of the players was looking for Mark's favorite whore. He overheard a conversation in which someone mentioned that Mark's brother knows his every move and manages his life for him. The player then decides that, instead of talking to Mark's brother or finding a place where whores might frequent, he'll go down to family day at the old farmer's market and ask everyone there if they've seen a whore with Mark and how to find her.

Some people are either stupid, or oblivious to the obvious. I would like to know more about how to train players to think through a situation better, allowing clues and cues and evidence and facts to inform their situation and how they deal with it.

I probably didn't start trying to think about an rpg situation as my character until I saw someone else do it and realized I'd been treating it all like a video game. And I didn't start trying to solve mysterious situations until I saw someone solve one just by reasoning through it. This is the totality of what I "know" about this topic and I want more understanding.

The lesson here is that for the party, Mark's Brother is boring. And they're probably expecting him to warn Mark or have a GM fuckery somehow, and that railroading them to Mark's Brother is a dick move.

Throw out a couple different leads. I mentioned the perfume above, the location of the whorehouse is another, a third could be the jilted sweetheart that Mark left when he became infatuated with this working girl. They had a bad public fight at the last farmer's market. It's like he's changed overnight ever since he met this hussy (cue dramatic music). See which lead they follow and run with that one.

[8d8]

The way it played out is not that Mark's brother was boring or no one wanted to talk to him. The player just didn't put two and two together. "Mark's brother manages his life" didn't seem like relevant information because the guy wasn't listening at all. He got it in his head that talking to everyone at the market was The Way You Get Information (he later said that was because "that's what you do in Zelda").

However, everyone's suggestions are probably better than the way our DM pulled it off. I loved that he had most of the people at the market totally skeeved out by being asked about a whore, and even the ones who might have known something just wouldn't talk about it in public. But he let it become a dead end until one of the rest of us suggested he look for someone shady, then he had to pay for the information. Robby's idea would have been better, or at least a lot faster.

[ishy]

or finding a place where whores might frequent, he'll go down to family day at the old farmer's market and ask everyone there [ ] and how to find her.

Seems like the player doesn't know where whores might frequent and is literally asking others where to look.

[RobbyPants]

Why go through all that? You've lost the party by the second scene. It's obvious they don't want to follow Mark's Brother The Clue. That doesn't interest them...

Have them go to market day and the perfume maker remember making a special perfume for the rather pretty lass glued to Mark's arm that smelled of dragonsblood and elderberry.

No, you're absolutely right. Giving them ad hoc leads like that makes them feel like they're better at finding clues.

That rather lengthy exchange I posted was out of jest. I can't imagine anyone wanting to actually play in a game that ham-fisted.