On detective games; how to let the players draw conclusions

[Shane 'ProgZmax' Stevens]

Well, I haven't read anything beyond Andail's initial post, but here are some things I think are cool in some of the games I've played.

In the otherwise odious adventure game Sherlock Holmes: The Awakened, there were a few cool deductive sequences where you found a location by following visual clues.  In the two traditional 2d Sherlock Holmes adventures (which were great), you had a functioning chemistry set you would use at key points to study gathered evidence by, for example, using acid to erode some rust blotting out a name or a chemical mixture to reveal the hidden words in a letter.

Another thing I REALLY LIKED was the corpse investigation sequence in LA Noire.  While it wasn't as robust as I would like, and while the conclusions were largely made for you once you 'revealed' the information, I think it's a great step forward for immersion.  Instead of the character automatically interpreting the evidence, I'd rather you were allowed to do so via some kind of interactive response system that allowed you to piece clues or ideas together to form answers (kind of like the notebook in Blackwell Legacy).

I'd like to say the interrogations in LA Noire were likewise rewarding but they tended to be random and unbalanced and often relied on luck or spending an experience point just to get over with.  I'd prefer interrogations that let you work with your clues more dynamically without having sudden negative results.

The ability to closely examine and dismantle certain things coupled with the ability to combine and refine your own thoughts in a notebook-style interface would make for a pretty engaging investigative experience.

[miguel]

This is really a good topic, looks like everybody thought about this at least once since started to use AGS or other game engine.

Do you guys remember the game Captain Blood? The way the player could talk to aliens through a set of icons that meant something? I remember being hard but really rewarding. It was like learning a new language. I can see all that we've been talking about done through a similar system, where you would build sentences and thoughts in a sequence by arranging icons that could be ideas, places, objects... Any ideas on this?

[Victor6]

Do you guys remember the game Captain Blood? The way the player could talk to aliens through a set of icons that meant something? I remember being hard but really rewarding. It was like learning a new language. I can see all that we've been talking about done through a similar system, where you would build sentences and thoughts in a sequence by arranging icons that could be ideas, places, objects... Any ideas on this?

Not strictly an idea, but something that's worth a look;- The auto translators used by some MMO's \ Online games. Yes, the text versions are generally awful, and a massive troll magnet, however I think PSO used icon based communication with custom arrangements to get around the multilingual user base, and lack of console keyboards.

Same concept, only the 'aliens' are just on the other side of the planet.

[Ali]

Along similar lines is the unfinished but stunning game Rorschach in which the player's only inventory is 3 clue slots which can be filled up and used on different characters in conversations. Rather like Resonance's Short Term Memory, I imagine, and very effective.

I should also say, I'm a huge fan of notebooks in Discworld Noir and the Blackwell Games. With the 'Azile' puzzle in Noir I worked out the mystery, then had to get Lewton to work it out. I loved that because it was so unlike solving mysteries in other character-based adventures.

[Ryan Timothy B]

Rather like Resonance's Short Term Memory

I agree with Vince that his idea for the short term memory would have been much better suited with a detective game. But I dislike the limitations with his current system (not that I've played the game, just watched the videos). Looking at a broken door should automatically add the memory, then when you are talking to a repair man or something you can ask him about the broken door.

I shouldn't just be able to grab random items like a chair and such and drag it into the short term memory. At least that is how it looked on a few of the videos, perhaps I'm wrong and it automatically gets added to the long term? I do like the idea of being in control of what you add to your memory, but it leads to walking back and fourth going "damn, I forgot to add that broken door to my short term memory".

A little hand holding by automatically adding things that are or possibly are important to the case into your memory system would be the best solution, otherwise you're running back and forth to go back to the murder investigation to add the odd shaped blood pattern into your short term memory so you can compare it with something that belongs to the accused.

[Shane 'ProgZmax' Stevens]

Having it automatically add all relevant points of interest would take away 100% of the challenge of investigating.  Then you'd literally have spoilers in your ltm/stm inventory that you'd randomly use on people constantly until it worked, like a ring of keys and a locked door.  Either that or so much clutter that you can't make sense of anything.

No, this is not good.

Vince's method is the most sensible one because it relies on the player to choose what is a point of interest or a topic for conversation and then store it (like the image of a dead body, for instance).  While the game technically allows you to memorize virtually everything in a room, most of these items understandably are going to give replies like 'That doesn't sound important/useful' when discussed, but it's there for people who are obsessed with the potential of every detail. 

[Ryan Timothy B]

That's the difference between a casual detective game and a game that requires you to be completely alert and intelligent and terribly patient. Or persistent by going back and forth and adding every item to your memory until you find the right memory.

For a detective game it wouldn't clutter your memory. They are clues that become larger clues. Or clues that have no lead at all and become removed when you discover they have no connection to anything (how that's done, I'm not sure yet).

When you see the odd shaped blood pattern, it adds the memory. Then when you look at the man's gym bag with a red stained bottom, it does not add that in your memory; otherwise the player would automatically know the gym bag is important. The only way you can make the connection is to drag the blood stain memory onto the gym bag and voila, the blood stained gym bag is added into your memory. In following you lose the odd shaped blood stain because it isn't relevant anymore now that you know the connection.

If I see a stab wound on the dead body, I shouldn't have to drag that into my memory. I of course know as the player that a stab wound needs to be investigated to find out who has a knife. Not having me walk around and realize once I see a man with a knife that I actually needed to drag that stab wound into my memory or I can't talk to him.

Edit: I actually remember being stuck in Blackwell Legacy a few times with having the clues right there in my notebook. There were a few times that I didn't know which clue to combine on what, or which to ask about. Having the clues automatically added don't actually solve the puzzles for you. You need to make many connections that slowly lead you to the full picture.

Now I have no issue with being able to add things that don't appear to have any relation to the case/problem into your memory, but into a different miscellaneous category. For instance, you need to get into the victims safe. The combination is in his office. Three sport shirts hanging on the wall with different numbers on them. That's the combination. Now it automatically adding that to your memory would definitely give the clue away, since it isn't obvious that it's a clue.

But in an instance like that, would you even need to add them into your memory? Depends if the safe is automatically opened when you drag the shirts onto it OR if you have to manually turn the knob instead.

[miguel]

The difference I see is not it being a casual game or a more "serious" one. The difference is in what clever coding is there.
What if you see the gym bag first? Does the player gets that into his STM? Wouldn't the player be allowed to ask about it to the owner? Is the player never allowed to see the bag first? What "freedom" is given to the player?

We are all talking about a system here that can deal with all the objects/thoughts/suspects/locations in a way that would "feel" the least linear as possible.
There are two approaches, 1, you let the player collect all the evidence in a room (even if its a chair like someone mentioned) and therefore is able to combine them and create thoughts and ultimately deductions. This approach feels fair to me.
2, you carefully choose what the player is able to collect, you lead the player, sometimes even mislead the player for narrative sake, but the obvious skill is to trick the player into thinking he's in charge.

Both can work but the first one seems more fair to the player.   

[Ali]

If I see a stab wound on the dead body, I shouldn't have to drag that into my memory. I of course know as the player that a stab wound needs to be investigated to find out who has a knife. Not having me walk around and realize once I see a man with a knife that I actually needed to drag that stab wound into my memory or I can't talk to him.

Surely that's what the distinction between Short Term and Long Term Memory is for in Resonance? Not having played it, I can't say how well it's implemented. But I do think that you should be able to drag a nondescript chair into your memory, because in a mystery or detective story you never know what seemingly innocent objects may hold clues.

[Anian]

Having it automatically add all relevant points of interest would take away 100% of the challenge of investigating.  Then you'd literally have spoilers in your ltm/stm inventory that you'd randomly use on people constantly until it worked, like a ring of keys and a locked door.  Either that or so much clutter that you can't make sense of anything.

No, this is not good.

Vince's method is the most sensible one because it relies on the player to choose what is a point of interest or a topic for conversation and then store it (like the image of a dead body, for instance).  While the game technically allows you to memorize virtually everything in a room, most of these items understandably are going to give replies like 'That doesn't sound important/useful' when discussed, but it's there for people who are obsessed with the potential of every detail. 

Well actually, guessing by every combination may be true in usual puzzles, but in a detective game (the part of solving a crime, not some puzzles to widen the gameplay) it's not really like that. Automatically connecting information snippets to some degree won't make it that much easier if you have a lot of combinations. Let's say you need to also solve a crime that has multiple steps (they usually have) like getting poison, getting alibi that turns out to be false etc. - it's a lot of informations that need to be directed to the player in some form of another.

I do agree that the memory gameplay Vince uses seems interesting and would fit in a detective game, maybe add in Blackwell notes system combinations, a timeline set up that you can mark up (to help the player out and make it more of a game then writing on paper) and also a semi clue gathering system with a basic forensic kit or analysis (maybe also a coroner NPC to give some extra info). And then you have a nice combo of things. Actually might be easier to make the memory system and the combining clues all notes, with the amount of stuff you can analyze (I really don't know how they solved that in Resonance without a lot of work or limitations.
...that might be rather hard to fit in a game though.

What I do know is that regardless of how interesting something might be, I really hate to read (and especially if I have to stare at it and reread it a lot) on screen, especially in a game. I think it also shows lack of imagination in the design of gameplay.

[Ryan Timothy B]

Perhaps a memory system as Resonance has is the completely wrong approach then. If the 'room' was the crime scene and you need to interrogate someone about an object in the room. Maybe being able to pull up the whole room, similar to Resonance style, would be best?

If the dresser hasn't been examined yet, the room snapshot shows the dresser as a black and white blur. Because you haven't examined it enough to remember it. Etc.

[Yoke 2.0]

In a modern (official police) setting the crime technicians would go through the room and supply you with a list of interesting finds. With a rogue investigator or Sherlock Holmes, who are their own CSI, you would have to relate to the fact that you can't investigate every single item due to time and due to the fact that you might not even get access to them without breaking in.
So either you outsource the work of determining which items are interesting and wich are filler, or you have to justify what items you want to look closer at.
If you want the game to be really, really, REALLY hard I guess you could have the possibility of investigating everything but having to relate to time units of some kind.

Based on the assumption that it's a detective ADVENTURE game and not a puzzle game:
- You have an inventory where everything you have seen and heard are added automatically.
- You have to combine two, three or more items to get to a theory. With enough items in each combination it will be near impossible to solve it by trying random combinations. And as I mentioned earlier these theories could go on to be building blocks in constructing new theories.

I also like the idea of reenactment as a central point of solving a case. Maybe as a screen where each time you reach a theory that pertains to the case, that action is added to the screen. That way you will have a visual clue as to what information you have and what information you are missing.

Dang... I want to make this game now... 

[Vince Twelve]

I will freely admit that a drawback of Resonance's system is that once you realize that something is important, like meeting a locksmith and then realizing that he could help you with that locked door (note: this does NOT happen in Resonance), you sometimes have to travel through a couple screens to go get that door, add it to your short-term memory, and then walk back to the locksmith and use it on him.  (In practice, though, you're never that far from the map screen which takes you anywhere in the game, so it's not a long backtrack.)

However, the alternative is having those items automatically pop up in your notebook/inventory, thus letting you know that it's important before you know why it's important.  This also makes choosing all the possibilities (brute forcing the solution) much easier.  You've already got your collection of important clues, you just have to guess which one applies to your current situation, or try all of them until you get it.

A better system of handling might be having some way of the player remembering all the places he's been and being able to choose any item from those places at any time without backtracking to them.  Like a photo album of all the room backgrounds you've seen, allowing you to choose any hotspot within those rooms as a topic of conversation with any character.  Same effect as Resonance, without the backtracking.  But a potentially overwhelming interface as you flip through all the places you've been.

Also, Snarky, why haven't you made that game yet?!

[Ryan Timothy B]

A better system of handling might be having some way of the player remembering all the places he's been and being able to choose any item from those places at any time without backtracking to them.  Like a photo album of all the room backgrounds you've seen, allowing you to choose any hotspot within those rooms as a topic of conversation with any character.  Same effect as Resonance, without the backtracking.  But a potentially overwhelming interface as you flip through all the places you've been.

Just as I wrote above but worded much much nicer.   

Also I would like to think this method would be much easier to implement than the short term memory. You wouldn't need an icon for every object/hotspot in the room. It would draw the room and the objects in them as a smaller snapshot of the room. Only having blurry and grayed out images for the stuff that you haven't actually examined yourself yet (which can even be scripting by drawing it offset at a light opacity to create the blur - or just pre render it). Since how do you ask someone about something you haven't examined yet.

Like a poster that shows a hint. Or a safe that you have yet to know if it's even locked. Then once it's opened, the photo of that room would show an opened safe. That way you can drag the opened safe to someone for new clues.

[Shane 'ProgZmax' Stevens]

What I do know is that regardless of how interesting something might be, I really hate to read (and especially if I have to stare at it and reread it a lot) on screen, especially in a game. I think it also shows lack of imagination in the design of gameplay.

Are you actually saying that you hate reading?  You hate journal entries, descriptions, short in-game books, that sort of thing?

However, the alternative is having those items automatically pop up in your notebook/inventory, thus letting you know that it's important before you know why it's important.  This also makes choosing all the possibilities (brute forcing the solution) much easier.  You've already got your collection of important clues, you just have to guess which one applies to your current situation, or try all of them until you get it.

And this is why I stand by manually gathering what you as a player consider to be important clues to be the most human, the most realistic, method of play.  As the investigator you're not supplied with a perfect list of everything unless you have other people doing your work FOR you (and is that any fun?) and you are also not supplied with perfect understanding of what has happened, so perhaps you will collect something as a clue which seems valuable but is a red herring, forcing you to revisit the crime scene and dig deeper or visit the evidence locker at the station and re-examine the evidence from a new angle.  This is real detective work.  

It seems to me like some of you would prefer to be handheld like so many modern gamers so there's no chance of slip-ups, mistakes, backtracking or delays, like you'd rather play 'Andail's Detective Story: The Movie Videogame' instead of 'Andail's Detective Story:  The Adventure Game' where everything is streamlined to the point that everything flows at breakneak cinematic pace with nothing left to chance.

Frankly, I find games like that to be a waste of my resources and this is precisely why I revisit older adventure games rather than buy many new ones.  I cannot stand being hand-held and coddled through a game I paid for.  Don't misunderstand, I also despise deliberately convoluted and obtrusive gameplay elements but, if faced with a choice between the two, I will always side with a more challenging game that relies on my wits to solve it than a game that does everything but wipe your bottom.

To summarize:  clue gathering should be a manual process for an investigative game.  It should take some time and energy from the player, but reward them for that time and energy with insights, backstory, further clues, and so on.  A notebook, in-game computer, or other means of storing clues for correlation should be included, as well as a kit (where applicable) for studying hard evidence, whether it's using your computer to analyze voice patterns, thumbprints, or pull up criminal records and case files or the means to process certain evidence with reasonable realism and accuracy (like a portable lab kit the hero keeps in a suitcase in their car).  These more portable devices limit the delays and nagging sensation of constantly backtracking to the player's home/place of work for thorough lab testing while still requiring it on occasion for items that cannot be reasonably studied (like actually processing bloodwork/hair/skin samples for dna).  These major 'visits' to the home/police station/whatever could be conducted as an 'end of day' event where the current data is gathered and sent in to be processed.  Meanwhile, this frees up the player to use their personal toolkit during the course of the day to do most of the light lifting while the evidence begins to provide the player with clues to work with.  One could always include an option for the player's boss/partner/confidant/(or the player themselves) to offer some cryptic but helpful clues once they begin processing the evidence and clues if they've somehow stumbled onto the wrong track or missed something entirely at a crime scene.  Worst case scenario, a difficulty level could determine whether or not other investigators catch clues you miss and pass them on to you as the game progresses.  This to me would be a sensible and strong approach to an investigative type game that rewards insight and study while limiting a lot of senseless backtracking.

[Snarky]

I will freely admit that a drawback of Resonance's system is that once you realize that something is important, like meeting a locksmith and then realizing that he could help you with that locked door (note: this does NOT happen in Resonance), you sometimes have to travel through a couple screens to go get that door, add it to your short-term memory, and then walk back to the locksmith and use it on him.  (In practice, though, you're never that far from the map screen which takes you anywhere in the game, so it's not a long backtrack.)

The only point at which I thought it was tedious (in the part of the game I played) was when controlling multiple characters, needed a certain STM as well as a LTM only one of them had, and had to have them work together to enter the screen to pick up the STM. (I'm sure you know the bit I mean.) That took quite a bit of back-and-forth just to be able to mention something I knew I wanted to say.

Also, Snarky, why haven't you made that game yet?!

Because it would take effort. [yawn smiley not found]
Also, it turns out to be pretty hard to come up with a good, twisty and non-obvious locked room murder mystery. OSD is more my level.

ProgZmax, yes, an adventure game should generally feature mental challenges of some kind, but they don't have to be realistic representations of all the things that would actually be difficult about that situation if it happened. It's OK to have a game where you play a spaceship captain but don't have to compute trajectories and fuel consumption, analyze sensor signals etc., and it's OK to have a game where you play a detective but don't have to go through every step of a real investigation.

And unlike real life, in a game there should always be a solution, you should always have a chance of finding it, and there should always be hints about how to do it. Because the important thing about games is that they are fun, and real life isn't always fun (even if you are a detective or an astronaut).

That doesn't mean the game can't be difficult. There are other (and, I would argue, more creative) ways to make things difficult than force you to do everything manually and explicitly make every little decision yourself. Ultimately, the difficulty is determined by the tasks it does involve, not all the tasks it doesn't involve.

[Shane 'ProgZmax' Stevens]

I think my last paragraph pretty much said that.

[Snarky]

Really? You don't think what you said was more... the opposite?

clue gathering should be a manual process for an investigative game.  It should take some time and energy from the player, but reward them for that time and energy with insights, backstory, further clues, and so on.  A notebook, in-game computer, or other means of storing clues for correlation should be included, as well as a kit (where applicable) for studying hard evidence, whether it's using your computer to analyze voice patterns, thumbprints, or pull up criminal records and case files or the means to process certain evidence with reasonable realism and accuracy (like a portable lab kit the hero keeps in a suitcase in their car).

There are other (and, I would argue, more creative) ways to make things difficult than force you to do everything manually and explicitly make every little decision yourself. Ultimately, the difficulty is determined by the tasks it does involve, not all the tasks it doesn't involve.

You seem to be saying that the only good detective adventure game is something that offers a reasonably realistic simulation of a real investigation.

I'm saying that a game is a game, it doesn't have to represent all aspects of reality (the Indiana Jones games are really crappy archeology-simulators), and what matters is the gameplay it does include, not all the things the game designer could have had you do but decided not to make part of the game. LA Noire and Apollo Justice make you read the body language of suspects/witnesses during interrogation/cross-examination. Does that mean all other detective adventures need to do the same? WHY NOT? It's an important part of investigating, isn't it?

Having to manually filter relevant evidence from all the irrelevant information is the same. Yes, it's something investigators have to do. Yes, you can make it part of the gameplay. But you don't have to.

[SSH]

Of course, you could take the Cluedo approach and let the objects one finds be eliminatory, and only once you have eliminated everyone else can you know who the murderer (and weapon, place if you like ) is. Now, the key thing with Cluedo games in real life is that if you want to win, you need to calculate second-hand what clues OTHER people are seeing. Now, if you could work out a way to do something similar in a single-player game, it woudl start to get interesting. I think there was an adventure-ish game version of Cluedo that did the first part, actually.

[Dave Gilbert]

The main issue I've faced with systems like this is that often the player will come to the right conclusions before the character will, and then get very annoyed with the character. If, say, you know that Jenny could not have eaten that steak because she is a vegan, you can't just go up to Jenny and confront her with it. No, you have to go jump through a hoop (or in the case of Blackwell, use the "Jenny" clue with the "vegan" clue) so your character can come to that conclusion as well, and THEN you can confront her with it.

The first two Blackwell games suffered from this problem a lot, and is something you once complained about yourself, Andail.  "Painful", I think you called it.  I think I eased the problem in Deception by creating "clue combining" puzzles that could never be solved any other way. Say, your current case involves a pet shop. You don't know which one, but you've got a list of all the pet shops in the city. Then you uncover a set of initials.  Use the "pet shops" clue with the "initials" clue and you cross reference them, finding a pet shop that matches that set of initials. The player would never have found the name of the pet shop without using that system, so it becomes a help rather than a hindrance.  

I'm sure there are other ways to improve this, but this seems to be the proper direction. At least for Blackwell.