Topic: On detective games; how to let the players draw conclusions

[Andail]

[CaptainD]

It's a very good question.  Unfortunately at the moment, I'm not entirely sure what the answer is.  In KOTOR there was a section where you had to uncover the answer to a certain thing (think it was on Tatooine?) but I'm not sure that method would work for a full-length game.

The only way I've really seen done, which more or less works, is by for instance dialogue options coming up only when you have found the relevant clue.

For instance:

Q: How was Alice poisoned?
Possibility: She ate a T-bone steak with poison on it
?? other options

When you click on one without evidence:

"I don't have anything to back that up" etc

When you've found out that Alice is a vegan:

"That can't be right since Alice was a vegan"

Obviously that doesn't completely eliminate the trial and error element.  Without having free text input that could somehow be interpreted, I don't know how you could achieve what you're aiming for.

[HandsFree]

This was handled very effectively in Sherlock Holmes vs Jack the Ripper.
A bit hard to explain how it worked in detail but they used a deduction board and a timeline that you could manipulate.
Check the images on mobygames for instance (they don't allow linking I believe), or better check the game itself.

[Anian]

The Barn runner Bake sale game seemed to have a pretty tangled "who dunnit" plot...though I don't really like it when I have to have 10 pages of notes to keep up with what's going on, I'd like it more if there was some GUI that I can control and make notes (or a version of notes).

That's not really true, I mean even if you use a basic combo of WHO, WEAPON (or let's say HOW but somehow designed to be more interesting and less obvious) and TIME (or lack of alibi) and you put that in combo with say 5 suspects, 5 weapons and 5 times of day - that's kind of a lot to just guess, isn't it?

I'm really leaning more to traditional game solutions.
Giving the player freedom to make conclusions, requires you to give tools - timeline that can be marked (or is marked when and info is obtained). Also something like all suspects get a sheet with info you gather, in one or few sentences per info. And you mark the timeline yourself if there's some extra info.
Then a tool to inspect the body and clues (again just 4-5 would probably give enough complexity to make the guessing tedious), this of course depends on the technology that will be available depending on the plot setting.
Then you connect different information that can give you an animated insight (or just a few lines of text explaining the conclusion, it is still the player making the decision but they get rewarded for it. Of course you have to be carefull that the player figures out or knows something but cannot achieve a conclusion in game. That kills the imersion pretty quickly and adds frustration.

And also what CD mentioned, conversation opens up when you have new info or evidence.

[miguel]

Hi Andail, I've been reading a lot on the internet about writing for adventure games, there are even blogs and forums about it although they are more RPG oriented.

I liked the way Diskworld Noir allowed the notes you'd write on the notepad to be used on characters. GK3 used icons instead of notes that would be available or not depending on what the player knew about the plot.

In my opinion, the worst option is having 10 dialogue options like :1) What do you think about John Doe? 2) What do you think about Ronald Doe? 3) What do you think... and so on... I get really frustrated having to go all the options because I'll do it in a sequence and that means that for the next 10 minutes or so I'll be reading a lot of text and not exploring the game.

For what I've been reading, it all goes on the quality of the writing, if it's well done almost any way of doing it - works! Using NPC to work with the main player and feed him knowledge about puzzle situations is the most common way that game makers use.

So, for me:

a) relevant notes (notepad) that can be used with NPC's,objects, other notes;
b) team-work between the player and a friend(s) NPC;
c) do not put the player having to read 10 pieces of dialogue straight! Sometimes it even happens that you're given the chance to ask something about a character that you haven't really met yet;
d) the plot and quality of writing is 90% of a adventure game;

This said I also believe that the Detective Genre is the best suited for adventure games, and because I can't write a good interesting plot is the reason I don't try to make one myself on a more serious approach.

[Victor6]

How about letting the player screw up? That'll cut down on the trail and error. You don't have to make it fatal, but allowing the player to combine two pieces of contradictory evidence in order to support their personal 'hunch' removes the classic adventure game safety net ('can't do that' 'I don't think this will work' 'I shouldn't do this now').

This forces the player to think, or waste hours of their time.

Of course, in order to be fair you'd need to let the player break up the combinations again, so they can start from scratch.

[Andail]

I'm also not very fond of notes and memories and stuff, mainly because it's been used so much, and also because  it doesn't really feel intuitively right to me that the game should make notes for the player. I mean, either the player finds the information interesting, and notes it, or he doesn't, and forgets about it.
If the player isn't prepared to commit to that sort of mental labour, there are plenty of other game types
One option is an in-game note-writing function, like you say Anian, that the player is in control of.

Apart from this, I'm not really into a brand new game system or GUI, I'm rather into finding story-elements, situations and puzzles that rely on the player to solve, using their own intellect.

While the Sherlock Holmes game looks innovative and clever, it's not at all what I want for my own game. The problem is I guess I don't really know what I want

[miguel]

I know what you're saying here, but the player's possible thought about a certain situation has to be programmed before.

Let's say you have the following note on Alice's (she's one of your suspects):
  Alice said she went to Tony's Pizza on Tuesday;

On John Carpenter notes you have:
  John Carpenter heard a discussion between Alice and a waiter, she was upset that the sauce tasted like meat;

You could use JC note on Alice's note to conclude that she is a vegan.

That is some kind of detective work, right? I can't see it done in any other way.

[Snarky]

A return to game making, eh? Very cool. Good luck!

I mentioned this some time ago as an idea I had, where you use (imaginary) staged recreations of the crime as the game mechanic to make use of the clues you've found. So at various points in the game you'll be challenged to re-enact the crime, and you'll flash back and step into the shoes of the killer (represented, in my mind, as an unidentifiable shadow). You'll have to use the evidence you've gathered to prove what happened (e.g. scratches around a keyhole to show that a lock was picked). Or a snippet of testimony can be used to show e.g. that a window was actually open, not closed as is being assumed (so in the flashback, 'use testimony on window' would open the window).

It still relies on having most clues as inventory items in some form (though you can use dialog mazes to challenge the player on their understanding of things from time to time, like in Phoenix Wright), but I think it makes it a bit more dynamic. I think the Phoenix Wright games in general are a good example of how to represent detection and deduction in adventure games, though of course they use a very specific system.

It's always a question how much to make the act of deduction explicit in the gameplay, and how much to let it happen in the player's mind, and just have the character's actions reflect the conclusions of those deductions. The Vacuum is a great example of the latter, where you don't have to do anything in the game to indicate who you suspect of being a killer, you just keep it in mind as you play and as you decide who to tell what, and whether to hand someone a gun. But if the mystery is quite complex, I think you do need to represent the steps of logic within the game (either have the player perform them, or have the character explain them), because otherwise players might get completely lost so that nothing makes sense any more.

[Khris]

This a really difficult topic. One problem I see right from the start is that if you pulled it off, only an actual Sherlock Holmes could solve it
Breaking it down, Sherlock has to gain knowledge in some way. The only question is, does the game keep track of that knowledge or does it accumulate exclusively in the player?

One idea I had was that there's less of a puzzle tree and more of a huge map of clues. Finding them increases confidence that suspect X or Y is the culprit.
At the end, allow the player to accuse anybody, but only show the good ending if they managed to accuse the actual culprit with >90% confidence or something.
Specifically, the player is supposed to take notes, and the confidence values are stored internally, not being displayed.

[Anian]

This a really difficult topic. One problem I see right from the start is that if you pulled it off, only an actual Sherlock Holmes could solve it
Breaking it down, Sherlock has to gain knowledge in some way. The only question is, does the game keep track of that knowledge or does it accumulate exclusively in the player?

That is the problem, if you give the player complete choice of which notes to take or not and which conclusions you can make you get a complete open gameplay that you have no control of pace or the story really. The other thing is the player is lost and if a complete opus of info is not available, they'll feel uninformed and not sure if they have everything to solve a crime.

That's where the whole you are "insert famous detective name here" superpowers come in - whatever you put as part of the GUI is what makes the detective awesome, you limit the player so you can guide them, but also because you cannot really put everything in the game. Making things easier for the player is not lowering the difficulty, it's using a gameplay issue to make them believe they're a better detective.
Maybe a good parallel would be 2d fighting games, like Street Fighter or similar - you don't really make the player learn exactly how to lift a foot, get energy from or whatever, they learn how a button does something so a skill/move that someone must learn their whole life gets done - so in a detective game with a detective that solved a couple of cases before, somethings get concluded, and those things are not available to the player in real life (or most players).
Making the game difficult in spite of the help you give out, that's a challenge. I'm just saying that basically handicaping a player (especially when you want them to think they're Batman, Sherlock Holmes etc.) in order to keep the game from being too easy, is wrong.

[Sane Co.]

On a totally unrelated topic, although it has to do with the beginning post.
I find that, in most detective games, what you say to the character doesn't influence your relationship with them. So I could say something totally negative, they'd be angry for a moment, and then they'd be their old selves again. Also in some detective games you're allowed to ask a question only once, or in a few others again and again and again. Laying off what I said earlier, you should allow the player to ask the question as many times as he wants, but the more he asks the question, the more irritated the witness becomes.

[Vince Twelve]

Great topic, Andail!  After five years of futzing with such a system, I'm intimately familiar with the shortcomings of the "game records memories, player draws conclusions from those memories" format of mystery solving.  I do plenty of this in my game, where at times you have to realize that you need to bring up a certain memory in dialog.  I also have a lot of Phoenix Wright-inspired moments where the player is directly asked to figure out which memory connects x to y, like what Snarky was talking about.

A few ways I've found to improve this system:
-Make memories that aren't relevant to puzzles.  They also serve as a "the story so far" to remind players what's going on in the plot.  Granted this only serves to increase the number of memories that need to be randomly tried when the player is stuck.
-Obfuscate important information within the memories.  In my game you can click a memory to watch or replay the event surrounding the memory.  The memory might summarize an important part of the game, but have an important detail hidden in the background or in a seemingly innocent comment by a character.
-Punish the player for random choosing.  I actually don't do a lot of that, but you might be able to find a way to do so.
-Make every item in the game a potential memory.  That makes random combinations of everything in the game much harder!

However, I think what you're talking about, making the player demonstrate that he has made a mental connection or figured out a clue by just acting out that suspicion (by going to and searching a suspicious characters' home, or by slapping the cuffs on a person who you know to be lying about their alibi) is much more compelling from a gameplay standpoint.  Just take out all interface-driven mental work (or have an open note-taking system of some sort, but don't make this into a game-play element) and instead put extra work into broadening the number of ways the player can interact with the game world. 

Of course, giving the game the depth needed to make this work is a tall order.  And the tallest of the tall orders is getting dialog working in a way that doesn't telegraph answers to the player by having a small set of dialog options displayed to the player.  The Short-term memory system in Resonance where you can take any item in the game world, make it a short-term memory, and then use that in dialog with any character would help with this a lot.  In fact, I've realized that this system would have worked a lot better in a more intimate crime-scene investigation game rather than in an epic adventure.

I do agree with the notion that such a system takes away the designer's control over pacing and plot.  Having a system where the player can act out their suspicions could lead to a very short case for the observant player.  Such a game would be great if it could randomly generate cases, though!

[Ilyich]

One of the problems with making the players 'think for themselves' in games is that while in real life there is a single universal set of rules that every human being is more or less familiar with, every game has it's own inner logic. And figuring out how each particular game works is an additional challange. You're not only interacting with the enviroments, but with the designer of the game. So it's sometimes necessary to push the player in the right direction and hold his hand - not to make the game easier, but for the player and the designer to be on the same wavelength.

I have no idea how to implement deductive reasoning without the use of notes or keywords, but to make the player think for himself more, you have to give him some freedom within the game. Here are some of the ways to do it:

- Text input - googling, searching through phonebooks [for the address of the hotel from where the matchbox comes from] or even answering questions like "Who is the murderer?" is a nice way to make the player pay more attention and maybe even write notes. It also makes trial-and-error a bit tougher.

- Non-linearity, choices and time limit (The Last Express did that quite interestingly) - for example, after inspecting a crimescene you have to make a choice - going to the house of suspect A or suspect B. If you've made the right decision you'll get some additional clues. This obviously requires a lot more work and can lead to dead-end situations, though.

- Allowing to pick up useless items. I remember collecting a full inventory of mugs and plates in the beginning of "Dreamweb". I think it was a very interesting, although slightly crazy, approach, especially given the bad case of kleptomania that every point-n-click gamer suffers from.   But that way you have to decide for yourself what's useful and what's not, what is a clue, and what is just a red herring. It's a bit extreme, but can be a fun gimmick for a small game.

I love Snarky's idea of reenacting the crime - it's pretty unconventional and as such will force you to think at least slightly differently to the usual "What do I use on what?" scheme.

Also, some RPGs made good use of trials (I think Neverwinter Nights 2 had a great courtroom sequence), which rely on long strings of arguments and the fact that you have only one attempt to prove that someone's guilty or not.

[Narushima]

To me what has worked the best is one of the old Tex Murphy games. You'd go around talking to people, noting things down to make connections and then you'd go and ask somebody else about something you had noted.
That way it was really all about your reasoning power and it felt quite rewarding when you finally got something right and could confront somebody with solid evidence.

So giving the player thr opportunity to input any text seems to me like a good idea. To test that you could always play some text adventures (or interactive fiction, as some call it), which are all about that. Make it Good is very good, for example.

[Darth Mandarb]

One of the things I miss about games from back in the day was they were freakin' hard!!

Games today seem to practically walk you through them.

I was, a few years back, working on a murder/mystery game (on a train) set in the late 1940s.  I gave the player a notepad and pencil and it was up to them to take their own notes to solve the murder.  I didn't hold their hand at all.  They had to draw their own conclusions, discover objects/clues, take their own notes, follow their own gut.

This mechanic works very well until you get to interacting with NPCs.   Because there's just no way for the player to just "say whatever" and have the computer reply to it (well there is, but I didn't feel like programming the AI for it!).

So I solved that by using this method: There were countless items scattered through-out the train; some were clues, some were objects that could you make use of, and some were just fluff added in to give the game more substance.  If the player character picked up something that was flagged as a "clue" it would then be added to the "talking points" (but they'd have no idea it was a clue flag, that was invisible to the player).

For example; if you talked to the conductor but hadn't interacted in any way with the wrench on the floor, you don't have any questions about the wrench in the dialog tree.  However, if you notice the wrench (look at, pick up, etc. basically you just need to "notice" it in some way), and then talk to the conductor, you will be able to ask about it.  (this has been done before I'm sure)

It makes the game incredibly difficult, yes.  But that's what I was going for.  No "hints" or anything like that.  I loved the concept/idea that you could play for an hour and have accomplished nothing if you didn't pay attention and make an effort to actually use your brain.

I also added in an invisible "friendship" meter to all characters you will interact with and based on how you treat them it may help/hurt you later in the game.  For example if you treat Mr. Smith rudely, later on he won't be so willing to help you (but some characters respond differently, some might see politeness as weakness, etc). 

I also made the murder victim, culprit and weapon different every time you play to make the game VERY replay-able. 

I digress...

[Yoke 2.0]

I've given this some thought in connection with other projects and my favourite solution so far is "the logic train".
The player has several pieces of evidence, a person and a goal (this does not have to be the whodunnit. it could be a smaller goal) Then the player needs to arrange a chain of information tidbits to form a chain from the person to the goal. If achieved that chain will become a new object that can be used in new chains.

I don't have a problem with "spoon feeding" the important pieces of information to the player, but if you want to make that even more interactive without the player actually resorting to pen and paper I think it's hard to get away from a transcript of sorts. This will enable the player to go back andreview previous conversations and discover clues that might not have seemed important at first. If you add an "HTML" function to the transcript where each sentence is a link that you can "bookmark" you will certainly get a bigger challenge for the player.

However making a single sentence a link will impose new limitations as it requires the characters to give information more or less outright instead of letting it bleed through in a more subtle way.

It's more of a question of what limitations you can live with and how much writing and programming you are willing to do.

[Snarky]

miguel said he thinks the detective genre is the best suited for adventure games. I wonder about that.

If we're thinking about the classic "whodunit" mysteries (police procedurals and other, more psychological/sociological sub-genres are a bit different), I think the big hook is the surprising revelation of the killer (or whatever type of criminal is being chased); that moment of insight when you suddenly see everything differently. It isn't (always) quite a twist ending, but it has a lot of the same appeal. This stuff is like crack for our brains, it's the same principle psychologists invoke to explain why we find jokes funny.

The flipside is that if you see the revelation coming, the experience is kind of underwhelming. That's not to say that you should never be able to guess the solution (it wouldn't be an interesting challenge if you have no chance of seeing through the author's misdirection; it'd just become an arbitrary flip-a-coin-to-pick-the-killer moment), but if you do it should also come in a flash of insight, not as the long-foreseen outcome of a steadily building pile of evidence. And once you know, the story is essentially over; it's just a matter of wrapping up all the loose ends as efficiently as possible.

That works in books and film because of the separation between the detective and the audience. In a classic mystery, where a genius detective cracks the case, we never have access to the sleuth's thoughts about the investigation, because then the reveal wouldn't come as a surprise.

But if you're playing an adventure game where you are the detective, you'll generally work out the solution at the same pace as the character (and when either of you falls behind, the effect is usually frustrating). So either the whole case has to rest on a single "a-ha!" clue (which is hard to engineer and may make the rest of the game seem pointless), or you'll work it out gradually and not get that sweet, sweet eureka moment.

What I think is the real secret to the success of the Phoenix Wright games isn't the clever game mechanic, the comedy or any of that. It's that the cases have a pretty deep structure, with several twists to be unraveled, each one (mostly) non-obvious enough to give you one of those revelatory jolts when the pieces fall into place. Gemini Rue, the Blackwell games, and I - would guess - most other successful detective mystery adventures do this to some extent as well.

So to sum up a long post: adventure game mysteries have to be structured a bit differently than in books or films, to account for the player's active involvement in the investigation and solution. Relying on one "big question" to drive the story may not work, because player's actually have to solve it, but  still have to be surprised for the thing to be any fun. A series of smaller mysteries that each function as a puzzle (i.e. you have to make one particular mental connection) is more promising, and the challenge is to weave them together in such a way as to create a coherent story and overall case.

I also made the murder victim, culprit and weapon different every time you play to make the game VERY replay-able. 

I have a hard time seeing how you'd avoid the whole thing becoming arbitrary this way. As outlined above, I think the appeal of solving the mystery is realizing "Of course that's how it has to be! It couldn't have been any other way. It all makes sense now!" If you're designing the game so that any solution could be the right one, how do you make it seem like the actual solution is the only one possible? (I realize you change the clues and details a bit depending on the randomly chosen configuration, but story-wise most of it has to be outcome-independent, right?)

[Ryan Timothy]

AGS isn't exactly robust enough for something this complex. Don't get me wrong, it can be done, it's just not as elegant. You'd have to create a struct for each character in the game that addresses all the possible clues, hints, current knowledge etc. Then with a universal struct for common things that are shared with every character, like stress, mood, etc.

Then depending on the hints and such that you can gather from looking at a photograph, or a doorknob, etc. you may need to make individual structs for each of those as well.

I would definitely prefer to tackle a game like this with C# style scripting instead of AGS script. But doing so would require writing all your own functions that AGS already takes care of for you.

[RickJ]

Wow, what an interesting discussion and many great ideas.

Snarky's idea about enactment sounds really interesting.  I guess it would all be taking place in th PC's mind so the reenactment world would only be composed of elements he experienced or investigated in the real world. 

I thought about making a Fact Based Dialog system a couple of years ago.  It's has come things in common to ideas mentioned by Darth, Khris, and Vince.   The basic idea is that each character has a collection of knowledge represented by one or more "facts".  The player can't ask about things which he doesn't know about.   For example if the coroner hasn't determined that the murder weapon was a knife the the PC can't ask a suspect if he owns a knife.  So until the coroner tells the PC "it was a knife" then the PC doesn't know to ask about it.   This is very much like what Darth describes.

I also thought of having confidence levels as Khris describes where each fact could have a confidence level of +/- 100%.   This could be used to trigger events or enable actions in the game world. The player wouldn't necessarily see this value but could possibly be made aware of it through comments of other characters such  as "Are you sure about that?" or "... sounds like you can take that to the bank!).

It would also be possible for NPC's to exchange facts via dialog (either on or off screen).  So when the NPC asks an informant about a knife the first time he might not know anything.  Now the informant knows about a knife and so could ask around about it.  When asked about it later in the game he may have new information.   

I imagined that it would be possible to instruct two characters to have a conversation about a specific topic or a random topic.  Options could be selected randomly or by keyword and facts would be exchanged.  Having this capability would allow the game world to operate in a sort of autonomous or chaotic fashion where it wouldn't be predetermined which facts the characters would have or what they would talk about at any given time. 

It would be also be possible to tell NPC's falsehoods or disinformation that they could share with other NPCs  would would share or react to.   For example PC could tell informant "Put the word out on the street that the police know where the knife is hidden...".  When the killer finds this out he goes to retrieve the knife.  If the PC is following the right suspect he will be led to the weapon.