Investigation in RPGs---Phonelobster just don't fucking bother.

[Prak]

Decided to split this off from the Play/Run Dichotomy thread

I think an investigation mini game is probably more important than a social mini game. Honestly, we wring our hands about a social mini game, and how it can break down over people doing things they might realistically do in an actual social "contest" (the nuke in the briefcase thing), but, honestly, I think... you should just roll with that, and let people socially interact in game they do irl, and let the dice and skills be there to fall back to if players or the GM want to use them, and to determine whether your character can convince another character of something that isn't pretty reasonable (let me go in this locked and guarded room, as opposed to "hey, can you watch my briefcase while I go piss? Don't worry about what's in it")

But investigation doesn't really work the same way. At least outside of grilling suspects, I suppose. You need to know whether the PC can find a hidden thing, how well it's hidden, whether they can recognize that it's a clue, whether they can make connections, and so on.

Harshax mentioned Gumshoe, and I have looked at it a bit. I ultimately decided I didn't want to use it when I was working on an investigation game, but I do think it has some good ideas. IIRC, the "you can use multiple skills to search for things relevant to those skills" and "you always find something| come from Gumshoe.

Last year, I decided I wanted to run a game set in the ALIEN universe as a Halloween one-shot, and decided I wanted it to be pretty investigation focused. I wasn't really ready to go all in on the actual ALIEN RPG, especially since I would have to learn it to write something up, and then teach it to my players. Instead, I used FATE, which I think is a very good system for one-shots. And I know it, at least well enough. You can pretty easily write up an adventure and a set of pregen characters if you want to do that, and so on in FATE. And the fact that it's pretty light means that it can be easily taught to others.

But I did have to figure out how to make an investigation happen in FATE. As well as how to write a mystery adventure at all. I did draw on the idea mentioned above with Gumshoe, where you can use different skills, and you always find something. I would have handled looking for clues by just saying "tell me what skill you're using to investigate. Say the right skill, you get at least this much, and some clues you can roll to get more." Investigation that involved talking to people was definitely going to care about rolls, though. One tip I found when googling how to write a mystery for an RPG was to have three clues at each scene where players investigate, and those clues should lead to the other scenes.

That's all I really did when I wrote an investigation adventure, though I can't say how well it worked since the game didn't happen. But I think those are three important parts of an investigation mini-game:

• People can use multiple skills to look for clues--and if you're making pregen characters, make sure each character has at least one investigation skill. If you're not Pregen-ing characters, the players should be guided to have at least one specific Investigation skill, and if you're writing an Investigation Heartbreaker, you can do this by having a list of such skills and a character gets n, where n is at least 1.
  • Corollary, you should probably not have a skill called Investigate. If a clue is found simply by saying "I snoop around the physical location," maybe it should just be Notice rather than Investigate.
• When investigating, a PC always gets something, and you should probably do this by saying "your job is to investigate. As long as you are in the right place and say you're investigating, you will get at least one clue, with no dice/DC involved."
• Each investigation scene should have at least three clues specifically called out, and at least two of those clues should lead to other scenes for investigation.

On the first point, I failed a little with the Pregen characters. There were five archetype-based characters, The Detective, The Courier, The CI, The Medic and the The Driver, each with two +3 skills, three +2 skills, and four +1 skills.
Each archetype other than The Driver had a +3 skill with a clue in at least three of the five scenes where you had to specifically say "I'm using X to investigate" (there was also a scene where you could just talk openly with someone who had no reason to lie, obfuscate or misdirect, and that needed nothing more than asking him about some of the clues you previously found), as well as at least two other clue-relevant skills between the +2 and +1 ranks. The Courier had Comtech at +1, but being an android, they had a special ability that basically upgraded it to +3 if they used their integrated datalink, as well as also having +3 Rapport, useful for the couple of interview investigation scenes where you did actually need to get info out of someone who had things to hide.
The Driver, however, only had two investigation skills (Notice and Provoke), both in their +1 track, so they would have been of somewhat less use in investigation. I think the idea was that they would get the spotlight when combat happened, which it was going to, and would have involved a xenomorph, but, still, I should have specifically planned for the Driver to be more involved in investigation.

Thinking forward, more, if I were writing an Investigation focused game, I think I would also try to make clues more relevant than just in solving the puzzle. Maybe having found a clue could give you bonuses to other stuff in the game, like, as an example off the top of my head with the above game, if you found the picture on the victim's desk of his daughter, you could invoke that when talking to the estranged daughter to put her off her guard and get a bonus in the interview.

[OgreBattle]

Just as there's rounds in combat with the enemy(s) moving as players move, some sort of structure like that can create time pressure. Like thoroughly collect clues and learn advantages vs the killer strikes again.

Here's an old post I've been revisiting, Hunter's Moon has players vs a supernatural monster to hunt down, players can take one action to prepare for the showdown, track it, try to prevent people from being eaten, and so on.

Character Creation
You start with character motivation that gives a general motive of why your PC is hunting monsters. An 'avenger' lost something precious to a monster and is out for revenge, a thrill seeker just wants to test himself and get the adrenaline flowing, a magician belongs to an order with a duty to hunt monsters, and so on. You determine if your character 'hates' or 'fears' the monster, hate means when your hitpoints are low your attack power rises, while fear means when your hitpoints are low your defense rises.

You then pick from a list of skills that you're good at. Different attack powers are tied to skills, and you can have skills disabled from damage. This also applies to monsters.

The game has set phases called "Sundown", "Night", and "Sunrise", so it represents one night of hunting for one gameplay session. Sundown, Night and Sunrise are broken down into further phases of...


Pursuit ("Legwork"): The phase starts off with the GM determining what the monster is up to, and then the PC's react to that.
-Monster roll: The GM rolls for what the monster is up to such as targeting specific people or going on a rampage, if the monster succeeds they get a benefit like a new ability to use, more morale (ablative hitpoints). and PC’s have to roll against that skill to stop it
-Location roll: Determines where the upcoming battle takes place, locations have unique effects like “Warehouse: players receive a free item”, “subway: anyone who rolls a fumble takes damage from being hit by a train”.

PC’s get 1 action to spend on things like researching the monster's weakness, stopping the monster from eating people, and so on:
-Weakness investigation: Finding out the monster’s weakness, making it easier to kill the monster or figuring out what body part you have to damage to disable its powers ("By piercing its chest it can't breath fire anymore!")
-Behavior Investigation: Finding out what the monster’s capabilities are, what triggers its super powers. ("The monster feeds off of anger and gets stronger when it senses it")
-Location Change: Change location of battle ("This monster flies, lets not fight it on the cliffside")
-Practice: Training, prepping weapons and so on, gives a bonus to the next battle.
-Support: Lower emotion, change status condition of allies (healing them), change emotion from fear/hate

After that's done you move on to...


Battle: PC’s run into the monster and fight
Battles are pretty short, with everyone getting maybe 1-3 turns before one side has to flee or dies. Hunter's Moon has a 'ablative hp and real hp' system with 'morale' (restores after battle) and 'hp' (when you lose hp you roll to see if you take injuries that disable skills and abilities).

This cycle of "Legwork-> Battle" is then repeated for "midnight" and "sunrise".

That above game is based around an expected final showdown with the monster, but I think that can be adjusted to different lose conditions based on what's at risk that PC's care about.

So for your ALIEN based game, in one round of 'non combat time' every PC can do one thing (prepare for combat, analyze its acid blood, predict its movement, gather NPC's to safeguard them etc) before the Xenomorph strikes again.

[Neo Phonelobster Prime]

But investigation doesn't really work the same way. At least outside of grilling suspects, I suppose. You need to know whether the PC can find a hidden thing, how well it's hidden, whether they can recognize that it's a clue, whether they can make connections, and so on.

I'm just going to use this as a tag in so you know I'm talking to you about your post. I picked it because I find it amusing that you describe hiding and discovery of information, declare that to not be how investigation works then sort of immediately confusedly back away from that statement (good idea to distance yourself from that statement there, even if you just said it yourself).

Anyway. The three clues per scene advice thing isn't entirely bad but it's a reductive gimmick. Your planning should center around detailing timelines of past events and the involved people, places, and things. Then extending that into a projected time line of future events that will (probably) happen during play.

It is more important (as GM, in your preparation work) to very clearly know what your "mystery" is and all it's details than it is to specifically place hand crafted blood stained candelabra in every room. You can make the noose on the spur of the moment as long as you know what captain plum recently did in the conservatory.

And that focus on time lines ties into the other thing I have to say about this.

You do not need an investigation mini-game for your mystery game. In fact it is probably the very last thing you want.

You probably DO want a social minigame or at least minimal mechanic. Because you care a lot more about information, compared to a dungeon crawl and a primary source of that is characters interacting with each other. So you WILL need to cover some minimal bases there.

You definitely do want a stealth minigame/mechanic extending to hidden objects and their discovery.

You absolutely want to definitely do that thing you mentioned where sometimes the information is just there and by being in the right place doing the right thing there is some minimum necessary discovery.

You probably need to add to that some advice on how to GM such that being in the WRONG place at the wrong time should also do SOMETHING that at least progresses the story interestingly.'

But most of all what you probably want, and what some of your later text suggests to me you are touching on is a minigame for...

...Out of combat time management. Your suggestions (rightly) don't seem to super care about the investigation success or failure (like a true investigation minigame would). They seem to care more about just determining a time interval outside of combats or other fast paced events/encounters and determining what characters are doing and where they are doing it.

And that sort of mechanic is very useful, specifically for mystery games with ambush events, roaming monster cults and count downs to events on the road to doomsday. And extra especially because it helps you interact with that time line preparation you definitely should already be doing.

If you have your ongoing future time line plotted out in your time management rules, hell, if your monster isn't even plotted out, it just uses those time management rules so it is always SOMEWHERE in town doing SOMETHING, you don't have to worry as much about prep work on fiddly little clues and player spontaneity can be more interesting.

Because players might make a random decision to go somewhere and just by chance walk in on the monster, because that is where it was this time management interval. OR you know that that location is where it was, eating some guy, one time management interval ago, and you also know it's still two time management intervals more before the police cultists turn up to clean the scene, something the players might ALSO encounter in progress.

If I had to somehow build an investigation game out of single game mechanical theme, it would not be an investigation minigame about clue discovery and accumulation towards an abstracted moment of mystery conclusion at critical clue mass. It would be event timelines, their progression, and interaction. Even basic discovery and actual monster fighting is absolutely secondary to the core of the game play for that genre of play.

[Thaluikhain]

Can't remember where, but someone here posted a link to an article which had a lot to say on this topic.

Notably, that rolling dice to determine if the game can progress or not, such as if the PCs find and interpret the clues correctly, isn't a great idea. Which, obvious when it's pointed out, but something that could be missed in the creating of stats things.

I should finish my Blacksand! review, their (rather long) scenario was in part a mystery. Long, IIRC, because they made sure to stick a lot of pointless waffle the players might want to do instead of following the clues correctly, which seems fair enough. You'll miss most of the sections, I guess, but then you can stick them into the next mystery I also guess.

[Prak]

But investigation doesn't really work the same way. At least outside of grilling suspects, I suppose. You need to know whether the PC can find a hidden thing, how well it's hidden, whether they can recognize that it's a clue, whether they can make connections, and so on.

I'm just going to use this as a tag in so you know I'm talking to you about your post. I picked it because I find it amusing that you describe hiding and discovery of information, declare that to not be how investigation works then sort of immediately confusedly back away from that statement (good idea to distance yourself from that statement there, even if you just said it yourself).

Phone, how in the fuck do you always manage to jump in with some bullshit misrepresentation of what's being said, and then go on to kinda sorta make something approaching a valuable point, but never actually saying a single worthwhile thing?

Is it a skill? Is it on purpose? Because statistically, by now, you would have said something meaningful at least once.

Just for fucking posterity, in case you genuinely could not comprehend the point I was making because you left your brain in a barrel of cocks-

In social interaction, people will ask for big favors that they need to convince the other person to do, and small favors, that generally someone will just say "sure, whatever." But people can hide a big favor as a small favor. And sometimes, someone is just an asshole, and won't tell you what fucking time it is, let alone let you through the door they're guarding. And trying to statistically model this shit is a fucking fool's errand. Just let shit happen, and if someone really wants to let the dice decide shit, then, fine, just roll a diplomacy skill. Who gives a shit, everyone here has better shit to fucking do than wring their hands about "But what if a player puts a bomb in a briefcase?" Bitch, who the hell gives a straining shit. Either don't let them get a bomb that fits in a suitcase, or reward them for having two goddamned brain cells to rub together and get on with your game.

But investigation does need some kind of model, even if it's written to facilitate success of the PCs more than other systems in your game, and that is something you can build a system around without wasting your time, in contrast to social interaction where you might as well just run it like actually socially interacting. Yes, there is an exception in interviewing people as investigation, and there is some kind of point to having an actual system to model those interviews. But even then, I think that you should mostly model it like you model any other process of investigating and looking for clues, rather than trying to create a fucking vacuum-tight mechanical representation of how people socially interact. Because you can model whether someone finds a clue or not, or has to dive into a dumpster to get it, but social minigames fall apart the minute someone thinks to just fucking lie.

[Neo Phonelobster Prime]

Are you seriously going to read that post, act like this, name ban me and THEN directly address me for 400 words?

I'm not the problem here.

[Harshax]

This is a long post and I'm just throwing ideas out. Full disclosure, these ideas are bad or problematic. I'm sharing my failure now because there might be a gem there that I'm not seeing.

Can't remember where, but someone here posted a link to an article which had a lot to say on this topic.

Notably, that rolling dice to determine if the game can progress or not, such as if the PCs find and interpret the clues correctly, isn't a great idea. Which, obvious when it's pointed out, but something that could be missed in the creating of stats things.

I don't know what article you're referring to, but that's sounds a lot like Gumshoe. And, I recommended it, not for it's rules but for it's understanding of constructing and running an effective mystery/investigation scenario. To Prak's original question in the other post regarding the Axeman investigation, you could build this scenario in D&D using the advice from the Gumshoe SRD, but there's a lot of mechanics that get in the way.

I've been thinking about what one would need to do to run effective mysteries in D&D while keeping the game intact and there's some obvious fixes and then some not so obvious solutions to lingering questions. Like, it's obvious that you have to remove/alter divination and alignment magic/effects so that paladins, clerics or spellcasters don't walking around spamming detect evil/zone of truth until they cast it on the right suspect.

But what does an actual murder mechanic look like in D&D? A single bloody murder weapon is a trope that's hard to fit into a game where people have gobs of hit points. Is there any satisfying way of turning off some of the most problematic D&D mechanics in order to broaden the types of scenarios Prak described in their original post? Specifically, can there be a 'Fight Music' mechanic where characters work exactly as described in D&D. But, when the music is off, characters are reduced to their bare stats, skills and possibly cantrips? Would such a mechanic just make the whole group rage quit for fear of being murdered?

broad proposal:
1. D&D characters represent their 'Fight Music' selves. When characters are properly prepared for combat are actively on an adventure, in high alert or just have their game-face on.
2. This state is not something that can be maintained indefinitely, which is why characters need downtime. Let's just call this civilian-mode. This downtime is often spent in cities or places where the safety and security of the populace is performed by guards and militia who take turns being in 'fight-music' mode.
3. When not in fight-music mode, Characters are reduced to base stats.
4. If a scene escalates from a downtime activity into melee, the fight-music starts. This shouldn't be immediate. There should be a surprise factor or reaction roll or something.
5. Damage is incapacitation is abstracted to a Saving Throw.

I fully realize this proposal is incredibly problematic because if someone is out to murder you while you're in civilian-mode, they will have a huge advantage and their full arsenal at their disposal because they're in fight mode. I don't know how to address that yet. What I do know is that this setup fits some scenes in fiction where a character believes they're someplace safe and let their guard down. A would be assassin may take brief advantage of this lowered state of alertness, but then the protagonist rallies and gets back into full fight mode.

Maybe 'Fight-Music' mode can only be maintained for a specific length of time and that's just built into the campaign on day one. This would force players to take down time, due city type adventures and plan for their next adventure.

I've run out of steam here. Hope some of this was food for thought. bbl.

[EDIT} - Hit Points as a concept in D&D is probably why I prefer to play games derived from Basic Role Play and then shoe-horn D&D-esque concepts into them. Games like those seem far more capable of producing a broader range of scenario possibilities but also suffer inversely for lacking truly over the top dungeon-crawl high-adventure anticts.

[Harshax]

So I just wanted to add one more thing here. I've been more and more intrigued by an idea that goes WAY back to the origins of RPGs: The scenario determines the game. I've read in some places that D&D used an Avalon Hill hex crawling game in the earliest days for outdoor travel and adventure, before TSR came up with its own mechanics. I've seen some games work really well for some types of scenarios, but not others. And, I think there's some room here to grab the best set of mechanics for a scenario while maintaining the same characters.

Like, if I wanted to run a hard-core dungeon crawl, I'd use torchbearer.
A heist/investigation: blades in the dark or gumshoe.
A war: Battletech
A political intrigue game: ???

So, if I'm playing Harshax the Adventurer, there would be a broad premise and description of my character. The dungeon-crawling Harshax uses stats and abilities from torchbearer and from a player perspective, I know what mechanics are relevant to crawling dungeons. When the MC takes us back to town and as a party we decide to get into a city/intrigue/heist adventure, it's my job to translate Harshax into meaningful mechanics to whichever system we're using (BiTD or Gumshoe). And as Harshax levels-up and starts dealing in world-politics, what the metadata about his character translates to an intrigue game.

Experience tells me that even the best efforts of RPG designers can't cover every conceivable scenario. Experience also tells me that players will often reinvent their favorite characters in the system being used for a campaign. So I'm kind of curious if there is a way to sort of smooth out the friction between switching between systems based on scenarios. Not necessarily a 1-1 conversion of a character from one set of mechanics to another, but more in keeping with fiction, a set of defining characteristics. Ultimately, I think this type of play would require every player to mind their place and establish a persona. And, maybe a group mechanic to call vote bullshit on a play that changes their character from a raving lunatic in the dungeon into a refined and persuasive aristocrat in a game of intrigue.

[Prak]

The person who writes/draws Kill 6 Billion Demons is working on an RPG called Icon that does something like that "different modes of play have different rules" thing, but in Icon, it's just Tactical Combat mode and Narrative Mode.

[Foxwarrior]

What if you look at it in the other direction, roleplaying in your investigation game. Imagine you have an investigation game, where the players try to work together to pick apart a mystery, trying to find out what happened and who did it. I think there are a couple such games in existence, actually... Many of them even involve talking to the other players and trying to socially determine what they know and if they're the murderer. If you start from an investigation game (look through some, I dare you) and want to make it more RPG like, what do you actually have to change?

[Prak]

Well, I've primarily been thinking in terms of Investigation RPGs, like Gumshoe. So... what do you mean when you say "Investigation game" that isn't an RPG? Like werewolf/mafia/secret hitler?

[angelfromanotherpin]

Or like Sherlock Holmes: Consulting Detective, or more recently, Detective: City of Angels.

[Dean]

The difficulty with mysteries and investigations is not the mechanical structure with which they are resolved but the immense difficulties in constructing satisfying mysteries to investigate. The rules of the combat minigame would be the least of your concerns if every time you wanted to fight a monster in game the DM had to invent a monster whole cloth, including their ecology, history, and society. AND if the invented ecology, backstory, and society elements didn't logically make sense then the monster fight would just implode and not work at all.

A good little mystery with some twists and turns will be satisfying to the players as long as your mechanics don't fuck it up unsalvageably. The question then is how can you get those good little mysteries without tasking the DM to spend 20 hours this week planning for just that element. A "Mystery Manual"? With basic mysteries like a monster manual perhaps?

You could also make the mystery generate itself to some degree. Clue and other board game "mysteries" do this. Where the mystery is just a combination of one thing from the A, B, C, and D columns put together. If you made a city sized setting with X Nobles, X organizations, X known criminals, and X targets then the mysteries you were solving could literally be randomized before the session and left for the DM to piece together the most logical thread of events they could before the session. This would allow a mystery game that was relatively quick to run (though it would still probably take an hour to fill in the details) though it would be relatively shallow.

TLDR: The rules, times, and tokens of your mystery minigame are a truly minor concern if the way you are going to make your mystery is "DM free form writes a good mystery". Requiring a DM to write something that often takes actual authors months or years to make is an unusable foundation for any system. How you're getting your mystery is the foundational question, once you know the mechanics of how you're getting your mystery you can make interacting mechanics for solving those mysteries.

[OgreBattle]

When running mystery investigation stuff with my students I've found having multiple "Crisis" situations to deal with adds to the fun. Standard Hollywood writing guides say "Have an A plot, then a B plot, and they meet at some point" so...

A mystery game with
A) Arson targeting stuff PC's are invested in
B) Police force arresting folks PC's care about

Player Characters can focus on one and the other gets worse, alternate or divide between the two.
It's like how in combat having two big strong monsters can be more dynamic than one. The monsters are expected to be defeated, the PC's are expected to solve the mysteries, but the order they're handled in and how they're handled uses PC abilities.

Side Story C that derails A & B) Burned kitten PC can roll healing and socialization on

[Harshax]

Interestingly, there is a Pathfinder/Gumshoe book out there by alégrame Press:

https://www.drivethrurpg.com/m/product/109418

[Harshax]

TLDR: The rules, times, and tokens of your mystery minigame are a truly minor concern if the way you are going to make your mystery is "DM free form writes a good mystery". Requiring a DM to write something that often takes actual authors months or years to make is an unusable foundation for any system. How you're getting your mystery is the foundational question, once you know the mechanics of how you're getting your mystery you can make interacting mechanics for solving those mysteries.

I believe you're overthinking this. A mystery involves someone doing an activity and a trail of decisions that someone did to perform that action. An investigator finds clues about those decisions to lead them to the someone who did that. I can make a mystery this fast:

I needs money.
I knocks over a casino for money.
I know a guy that knows a casino that is easy to hit.
I hire muscle to help me rob a casino.
I buy gear to help me rob a casino.
I rob a casino.

In reverse, as suggested by the gumshoe SRD the players need these clues:

A place where muscle can be hired to rob a casino.
A place to sell gear to make the robbery successful.
A person that needs money.

The game requires clues:

A witness that recognized muscle or maybe ID'd a person.
A transaction for gear.
A transaction for muscle.
A place that muscle could be gathered.
A fired employee with a grudge or debt.

For the record, I'm not suggesting that this format is anything like deep conspiracies of deceit and betrayal. I'm not suggesting this is a story that is so Byzantine as to require deep investigations into individuals. But it is indeed a mystery and if you're running an investigation there's enough here to point a to b and b to c. You can keep repeating this pattern until you reveal a grand Byzantine conspiracy, where the Bad Guy is revealed.

It doesn't require months or years, because you're not writing prose.

EDIT: I think the crux of my disagreement with you is that, as MC, you're not responsible for carrying the entire story. You're not writing a novel where every character is under your control. You're simply responsible for setting a premise for conspiracy and then admitting to the players that the steps toward making that conspiracy happen are available and actionable. Because why would you, as the MC, present an impossible scenario to the players? Wouldn't that be a waste of everyone's time?

[Dogbert]

You don't need to be Sir Arthur Conan Doyle to do a mystery (I'm sure as hell not). All you need to plan a mystery is answering the "5 w's"...

1) What
2) Who
3) When
4) Where
5) Why

Get your culprit, his motivation, and build a crime scene. You're done. The crime scene will spawn the clues your PCs will spot. The way gumshoe does it is that all clues are given away for free as long as you have the right skill to notice the right clue, but then it's all on the player (not the PC) to put the puzzle pieces together.

While not my favorite method (I hate puzzles), it's the best available in the market so far. Personally I'd let players spend an additional point from their PCs' investigation skills to have the GM interpret a clue correctly for them.

[OgreBattle]

The "Why" is often the payoff for capturing or defeating the culprit and they give their tragic or sadistic motive. Most Detective Conan mysteries end that way

Things that work and I enjoy are...
- Dungeon crawls
- Set piece battles

So I'm thinking of tabletop mysteries sorta in those terms and pacing. Does that make sense to y'all? Like solving the mystery is the treasure in the dungeon and getting there involves seeing traps, solving puzzles, encounters that deplete resources, and not taking so long that your Rival already took the treasure.

[JonSetanta]

My current take, an anti-Gygaxian opposite, is that rather than "roll your skill for pass/fail" all non-combat actions are instead based on time investment to a challenge.

To investigate a small room or single page would be 1 minute.
To do the same to an entire floor of rooms or chapter of a book is 10 minutes.
To do the same for a small building or entire book (skimmed) is 1 hour.

The better at detect/investigate/intellect or whatever ability or trait a character is, the less time it takes, to the point where a "Genius" or "Operations Director" or whatever could speed-read a whole book in a minute for exactly what they need to find, or order a team of lesser-able allies to spread out and search then report back in a fraction of time.

I was reluctant to include supernatural abilities such as "ask the spirit that died here for info" or "touch an object to learn it's past" but it would be the same time investment, different approach, yet same result.

Combat-focused characters could likewise interrogate and intimidate locals or suspects to cough up info. Same timescale.

[deaddmwalking]

My current take, an anti-Gygaxian opposite, is that rather than "roll your skill for pass/fail" all non-combat actions are instead based on time investment to a challenge.

Obviously NOT finding something that is findable can be frustrating. If it's the ONLY CLUE and it gets missed, that can derail an adventure. But that's really a failure of adventure design. If the PCs investigate and don't find something, the villain might THINK they know, and that would trigger a new plot development that opens up new avenues of investigation.

I think you can imagine a character that finds nearly impossible to find clues. The difference between Sherlock Holmes and Inspector Lestrade isn't that Lestrade needs 2 hours to do what Holmes does in 2 minutes; it is that Holmes notices things that Lestrade would never deem important enough to warrant notice. Whether Lestrade spends two minutes, two hours, or two days, he's going to overlook a clue that Holmes would notice.

Making a scene 'impossible to fail' is a design choice, but it comes with costs, too. If the only difference between running through the same adventure twice is how long it takes, it's pretty clear that it's a railroad. Maybe you decide that people will DEFINITELY find a clue, but you randomize which clue(s) they find in some way to give different investigative paths.