On the Design of Good Investigations

Find detailed help from the AAO community, or write your own tutorials.

Modérateur: EN - Forum Moderators

On the Design of Good Investigations 

Message par Ferdielance » Jeu Fév 07, 2013 6:39 am

Investigation segments demand close attention to design. Not all players investigate in the same way, so a design that satisfies the designer may be torturous to the player.

Here, I'll discuss principles and pitfalls of investigation design in a guide punctuated with exercises to try out. Feel free to post your answers below!

What Makes a Good Investigation?

---

Exercise 1: What do you consider the top three most important aspects of a good investigation?

-----

My answer, which is far from the only possible one, is that classical AA investigations require:

Spoiler : Ferdie's answer :
    * A sense of exploration
    * Engaging storytelling
    * Well-paced gameplay


Wildly non-traditional investigations can break the rules, but usually, failing at even one dramatically reduces the fun of an investigation. Let's start with the sense of exploration.

The sense of exploration:

Traditionally, AA investigations give the player little real choice. Every dialogue option must be chosen and every piece of evidence noted, ensuring that the player sees all relevant clues. This makes it easier to pace the development of the plot and characters.

---

Exercise 2: Note some of the pros and cons of giving the player choices in an investigation. Is it possible to do this and still have a fair, winnable game? What challenges does it present?

---

Even in this model, careful design can give the player the impression that they're freely exploring, to a point. Filling every location with curiosities to examine, letting them decide whether or not to examine irrelevant scenery, appeals to those of us who like to poke around. The same goes for optional "present" dialogue; at the very least, people should respond to their OWN profile.

There are two kinds of AA players: those who show the badge to everyone, and those who only show it when necessary.

---

Exercise 3: Choose one person from group A and one from group B:

    Group A: Phoenix, Apollo, Mia, Edgeworth

    Group B: Manfred von Karma, Chris Rock, Queen Elizabeth I, Superman

Write a present conversation for when the first person presents their badge to the second.

---

Exploration is more compelling when there's a genuine sense of place. By a sense of place, I mean that the locations seem to exist in physical space, connected to each other by more than a "Move" button. The AA games have a hit-and-miss record in this regard. In Trials and Tribulations, Hazakurain is organized like a real location with real (if somewhat compressed) distances, but the scattered locations in the restaurant poisoning case are transparent bits of programming to click through.

(Some of the locked room fancases perfect this aspect of investigation design, probably because classical mysteries place a major emphasis on how locations are laid out in space. Turnabout Pairs hinges on it.)

The other way a location can give a sense of place is through consistent atmosphere and tone. If you pull backgrounds from Ceres' background pack, try to pick ones from the same game if possible and use consistent color schemes. I haven't always managed this myself, as it's hard with limited resources, but do try. Nothing breaks immersion more quickly than when the bedroom of a Western mansion looks like a small Japanese apartment.

Another trick to make places seem real is to show them at more than one time of day, or under different weather conditions. If done well, this brings out the way nighttime dulls the colors and distorts the shadows of a place, making even known locations suddenly unfamiliar. Of course, this is only possible if competent day and night backgrounds are available, which is not usually the case.

Finally, and most critically, locations that are not sterile and institutional should reveal something about the characters who occupy them. They certainly shouldn't contain objects that don't belong there; if you use a pre-made background, you should either edit out anything that is blatantly out-of-place for your story or come up with a convincing Examine message to explain it.

---

Exercise 4:
Here are the rooms of two people in the same business, managing goofy entertainment:

Room A:
Image

Room B:
Image

Both rooms are decorated with entertainment-related trinkets, but do they tell similar stories? Imagine you were playing these games for the first time. What impression would these locations give you of the people who inhabit them? What do they have in common? How do they differ?

---

However, no amount of geographical logic can make an investigation immersive if the player dawdles for too long. When this happens, the gears and wires that underlie the game logic get exposed.

For example, suppose the player wanders around desperately presenting everything to everyone, then discovers that presenting a case file to Gumshoe causes him to get a phone call from Franziska. Upon arriving in a location which had been empty the last five times they visited it, they see that Furio Tigre has arrived. From this player's point of view, it takes much suspension of disbelief not to consciously think, "Presenting the case file to Gumshoe caused Franziska to call him. These events also caused Furio to arrive." They are, of course, entirely correct, as this is how the investigation was programmed.

Under these circumstances, investigation becomes a dull guessing game. If the player is trying to figure out what triggers the next event, rather than where the next discovery would logically be made, the exploration is over and the metagaming has begun. The more naturally the events flow from the player's actions and movements, the better.

This doesn't apply as strongly if a well-defined obstacle bars the player's progress, such as a Psyche-Lock. In this case, if the designer has done their job well, the player will think, "How can I break this lock?", not "What event do I need to trigger?" But even that situation can break immersion if the player ends up re-searching every location for missed evidence.

----

Exercise 5:

Consider the following two design schemes. In both, the player can go to six locations, labeled A-F. Three events can be triggered by the player.

SCHEME ONE:
    * Player moves to Location C to trigger event 1 by examining an object.
    * Event 1 occurs and unlocks event 2.
    * Player moves to Location D to trigger event 2 by talking to a character.
    * Event 2 occurs and unlocks event 3.
    * Player moves to Location A to trigger event 3 by presenting an object.
    * This segment of the investigation ends.

SCHEME TWO:
    * All three events are unlocked from the beginning.
    * Once the player has triggered events 1, 2, and 3, in locations C, D, and A, in any order, this segment of the investigation ends.

Which of these schemes seems more like a player-directed exploration? Which seems more like a linear story? Under what circumstances might scheme one work better? Under what circumstances might scheme 2 work better?

---

Mummifying your player in endless wrappings of non-interactive text will also kill the sense of exploration. In the text adventure community, this is known as "tying the player to a chair and shouting the plot at them." Don't do it; research and experience show that a bored player won't even remember those critical details you're telling them. Instead, make them question the suspects and find physical evidence that tells the story more vividly.

After writing this guide, I edited it to remove and tighten text. Do likewise.

---

Exercise 6: Why are there exercises in this guide?

---

Summary:

What's GOOD for the sense of exploration:
    * Sensible location design
    * Coordinated backgrounds
    * Optional Examine and Present dialogues
    * Player interaction that logically drives events

BAD for the sense of exploration:
    * Unconvincing responses to evidence presentation
    * Flatly incorrect examine dialogues ("There is nothing there" when the player looks at an object.)
    * Plot monologues
    * Arbitrary event triggers

-------------

I'll continue to the discussion of storytelling and pacing later, with a focus on the design rather than writing. Any questions or comments?
"A slow sort of country!" said the Queen. "Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!"
Avatar de l’utilisateur
Ferdielance
 
Message(s) : 762
Inscription : Dim Mars 09, 2008 12:46 am
Genre: Masculin
Langues parlées: English

Re: On the Design of Good Investigations 

Message par Bad Player » Jeu Fév 07, 2013 7:50 am

I think one of the big difficulties of making investigations is just the fact that it's spread out and non-linear; it makes scripting a bit harder. You don't know exactly what the player has seen so far, you need to make sure all your variables and multiple copies of the same location are working right to ensure the player doesn't get weird bugs and every location works properly at all times, you want to get on with the plot but you can't because you have to go script 3 other locations first. Even just copying examine convos when making the copy of a location can be dull and relatively time-consuming. In a trial, you get linear progression and always see the story continuing to develop; in a investigation, it's a lot slower.

Arbitrary event triggers also tend to be pretty hard to avoid. There usually isn't any convincing action the player can do to cause a lot of necessary events to occur, so event triggers usually need to be unrelated. There probably isn't anything the main character can do to get Furio to show up, but it might be vitally important to the story that we find out the contents of Franzy's call before talking with Furio. I think it's less important that the event triggers aren't arbitrary, and more that the triggers are something the player will naturally fulfill. Even if the trigger is arbitrary, as long as the player isn't forced into brute forcing its activation...
Avatar de l’utilisateur
Bad Player
 
Message(s) : 7211
Inscription : Mer Mai 20, 2009 10:53 pm
Localisation : Under a bridge
Genre: Masculin
Langues parlées: American

Re: On the Design of Good Investigations 

Message par Ferdielance » Jeu Fév 07, 2013 10:00 am

think one of the big difficulties of making investigations is just the fact that it's spread out and non-linear; it makes scripting a bit harder. You don't know exactly what the player has seen so far, you need to make sure all your variables and multiple copies of the same location are working right to ensure the player doesn't get weird bugs and every location works properly at all times, you want to get on with the plot but you can't because you have to go script 3 other locations first.


They're extremely challenging to design, but I'm going to be presenting tricks to make them much less of a chore. Believe me, I'm interested in doing really over-the-top programming in my investigations, so I know that it requires great care and patience. At the same time, I think that an investigation should be fun to code, on some level, and there are some good practices that can keep them from being torture.

Even just copying examine convos when making the copy of a location can be dull and relatively time-consuming. In a trial, you get linear progression and always see the story continuing to develop; in a investigation, it's a lot slower.


I never copy-paste examines; it's too easy to make a typo or error in one, then not notice it in testing. In fact, I have a general rule that if I'm copy-pasting, I'm likely doing something that I should be using redirects for.

I prefer to use the following trick:

In the first copy (investigation frame) of a location, I write the full examine scripts, all of them, and set the value of a variable called "locCopy" to 1. However, at the end of each examine, there's a frame redirect. This redirect goes to that location's "copy handler," which is just a "Read a variable's value" action that reads the value of "locCopy." If it's one, nothing special needs to be done. Otherwise, it jumps to a different investigation frame.

In other copies of the location, I set "locCopy" to a value representing which copy of the location the player is at. I then add a single frame for each examine going back to the examine frames in the original location. This allows me to make the initial handling of the examine quite sophisticated without making it a pain in the neck to re-copy all of that programming later. Thus, I can check whether the player has examined an object before and give a short version of the examine dialogue on subsequent clicks.

At first, this seems like more of a mess than copy-pasting, but once you have three or more copies of a location, it's a huge time-saver.

(This is pretty close to how I did it in AtBaS chapter 1, where it worked quite smoothly with a bit of debugging. However, wherever possible, I do try to avoid re-making the same location over and over unless there's a graphical change to it. Instead, I use the intro conversation to check whether or not a new discovery needs to be made or a new convo shown. That way, every time the player visits, something different can happen, if necessary.)

Arbitrary event triggers also tend to be pretty hard to avoid.


Below, I'm going to list several strategies to make event triggers that flow more naturally from the gameplay and don't strike the player as contrived. If the player naturally trips it, that's fine. But it's best to have a contingency plan in place in case they don't!

The key thing is to think carefully about the passage of time and take advantage of the way stories work in people's heads. (The original title of this tutorial was going to be "Time and Exploration in Investigations," until I realized that there wasn't a general "how to design good investigations" tutorial.)
"A slow sort of country!" said the Queen. "Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!"
Avatar de l’utilisateur
Ferdielance
 
Message(s) : 762
Inscription : Dim Mars 09, 2008 12:46 am
Genre: Masculin
Langues parlées: English

Re: On the Design of Good Investigations 

Message par AP-Master » Dim Avr 21, 2013 7:42 pm

You should really continue this :sawhit: It's just great, without a doubt; and the exercise dynamic gives an interaction that other guides usually don't give :)

I would like to say that I favor non-linear investigations:
SCHEME TWO:
* All three events are unlocked from the beginning.
* Once the player has triggered events 1, 2, and 3, in locations C, D, and A, in any order, this segment of the investigation ends.


Since I have practiced a bit in the design of investigations, I can see how this is hard to accomplish though... It requires proper knowledge of variables and evaluating conditions in order to pull it off :grossberg: And also, there are times, most in the end, in which I think it's better to do something similar to Scheme One:
SCHEME ONE:
* Player moves to Location C to trigger event 1 by examining an object.
* Event 1 occurs and unlocks event 2.
* Player moves to Location D to trigger event 2 by talking to a character.
* Event 2 occurs and unlocks event 3.
* Player moves to Location A to trigger event 3 by presenting an object.
* This segment of the investigation ends


In a sense, I think that a dramatic moment which is triggered only after doing something generates some good emotion in the trial. Easy example: examining something in a dark place suddenly makes someone start running near you, making noise and such. Something this critical (since you would think he is the culprit) can often only be made by unlocking it after having investigated in all the normal places, because I think an event like that one would make the investigation take a whole new turn. I hope I'm explaining myself :?
English Section - Visit my Sprite Castle!: viewtopic.php?f=30&t=5716
Sección Española - Visita mi Castillo de Sprites! [Peticiones ABIERTAS]: viewtopic.php?f=31&t=7792
¡Aprende a Spritear! - viewtopic.php?f=31&t=7838
Avatar de l’utilisateur
AP-Master
 
Message(s) : 286
Inscription : Lun Nov 16, 2009 10:59 pm
Genre: Masculin
Langues parlées: Spanish and English

Re: On the Design of Good Investigations 

Message par Ferdielance » Jeu Jan 05, 2017 9:50 am

Inspired by a comment on this month's discussion thread, I'm going to continue this tutorial.

NOTE: This is the "draft version" of this post. I'm putting it up to get feedback.

Engaging Storytelling:

Engaging a player means making them care about:

    * What happens next.
    * Why it happens next.
    * How it happens.
    * Who it happens to.

Let's look at each of these.

What happens next:

When a player is immersed in their exploration (see prior section), they will care about what happens next. Always make 'what happens next' something worth caring about!

What defines an event worth caring about? To oversimplify a little, an event can engage players if it:

    * Raises tension...
    * Or resolves tension.

How do we create tension? In a game, tension rises when:

    * The player wants something, but they don't have it (yet).
    * A character wants something, they don't have it, and the player cares about that character.

---
Exercise 7:

Imagine you're playing a (well-written) investigation sequence that is relatively early in the game. What do you want to happen in the game? How does that differ from what the characters want? Consider the following things that could happen in an investigation. For each list, rank the events, starting with the events you as a player would least want to happen and ending with events that you as a player would most want to happen.

Events (rank from least to most desirable):

    A. The defense finds a piece of evidence at the scene that is quite bizarre, doesn't seem to belong there, but is clearly linked to the murder. Its significance is unclear and rather worrying.
    B. The prosecutor (not the player character) suddenly works out who the real killer is and drops all charges against the defendant. Justice is done without the defense's interference, but they still get a decent paycheck for their investigation work.
    C. A new complication occurs that makes the defendant's situation even worse. Panic ensues for the defense.
    D. A witness gleefully insults the defense attorney using an astounding variety of witticisms, facial expressions, and animated gestures.
    E. The detective announces that there's going to be a new prosecutor in court - one the defense has never met before.
    F. The defense marries the love of their life.

Now repeat this exercise, but instead rank the events starting with those that a defense attorney protagonist would least want to happen and ending with events that a defense attorney protagonist would most want to happen.

Do you see any trends?

---

There are two kinds of things that players want:

    * Things they want to get right now.
    * Things they want, but not just yet.

At any given time, players generally want the music to sound appropriate, the graphics to be pleasant-looking, and all the other niceties of a polished game. They want that right now.

But although they want to win the game, find the truth, see the characters reach dramatically appropriate fates, and so on, they do not want those things immediately.

If you're doing your storytelling job well, you'll give the player many things to want, but not get right away. The same goes for your characters. You will understand their wants inside and out, and stoke them, and make the player share in them. This is tension.

Tension can be released by:

    1. Lightening the mood: making what is wanted less deadly serious for a moment. (Goofy jokes during a murder investigation)
    2. Lowering the stakes: making the character or player want something less.
    3. The player/character earning what they wanted: this is the most obvious one, and often what the trial is geared towards.
    4. Stopping the search for what is wanted: A character realizes that what they want is impossible, or not what they really needed. This is a powerful technique. Often, it reveals deeper character development.

The last one is often reserved for endings, but the AA series has examples of each of these happening during investigations. The best cases raise and lower tension in surprising and illuminating ways. Tension should rise AND fall, but tend to rise overall.

The Furio/Viola subplot in AA3-3 exemplifies great tension management. At first, Viola menaces Phoenix and Maya. She's a Mafia loan shark who makes disturbing threats of arson and poisoning, and she's very attached to Furio Tigre, the villain. This raises tension in many forms: fear for Phoenix's safety, uncertainty at what this unpredictable woman will do, and, more subtly, curiosity about her motives. Why is she so loyal to a goofus like Tigre? Why would she help him commit a murder?

This tension meets a surprising emotional resolution. We, and Viola, learn that Furio's love for her is a lie. Phoenix, once fearful of her, now feels sympathy.

The things these characters thought they wanted - Tigre's love (for Viola), to get away from a frightening person and take her down for her role in a murder plot (for Phoenix and Maya) - are replaced by different wants. Phoenix wants to stand up for her. And Viola wants...

...well, that's better saved for a later part of this post. Suffice to say that the mystery and tension there are never fully released even through the end credits.

This is an extreme example of an entire emotional arc that gets settled outside of the courtroom, but there are subtler releases of tension during investigations. Lotta Hart's search for Gourdy ends in disappointment. Matt Engarde reveals his true face, removing one source of tension ("can we trust this guy?") and replacing it with a new source of tension ("oh no, he really is that bad!")

Fancases often neglect to raise and lower tension, and to seek multiple types of tension. It's not enough to assume that giving clues to a mystery will create enough tension to carry the case. Remember that you as the author may know why a piece of evidence is important, and why you think the player ought to care about it. But sadly, the player may not share your opinion. They certainly don't share your perspective.

--

Exercise 8:

The protagonist and their assistant go to a party. A murder occurs, as murders will at parties. A partygoer is accused.

State an event that could raise even more tension at this point... in such a way that solving the murder would NOT be sufficient to release the tension!

Then devise four ways to release that tension (partially or entirely), using each of the four techniques above.


(Note: What sort of answer you give for part 8 says something about the kind of stories you like to tell. A person who answers "The accused is therefore searched and no weapon is found, but several stolen rings fall from their pockets" is likely to write a very different kind of investigation from someone who answers "The accused reveals to the defense that they were in love with the victim and about to present them a wedding ring.")

--

(TO BE CONTINUED)
"A slow sort of country!" said the Queen. "Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!"
Avatar de l’utilisateur
Ferdielance
 
Message(s) : 762
Inscription : Dim Mars 09, 2008 12:46 am
Genre: Masculin
Langues parlées: English

Re: On the Design of Good Investigations 

Message par Gav » Sam Jan 07, 2017 3:58 pm

I like this, though I feel more time could be given to the different ways of releasing tension. As it stands, I'm not exactly sure what you want to express with your reference to the relationship between Viola and Tigre, for example.

Spoiler : My answer to exercise 8 :
The accused turns out to be the protagonist's runaway brother, and the accused reveals that the protagonist drove him to run away because of relentless, dehumanizing bullying.

1) The protagonist's assistant relieves tension by trying to mediate between the two, causing a comic scene.
2) The protagonist reveals to the assistant the terrible things the brother did to the protagonist when they were younger.
3) The protagonist and accused have a heart-to-heart and begin to reconcile.
4) The protagonist decides that a good relationship with the accused cannot happen, and cuts off all ties with the accused after the trial.
gotMLK7 a écrit :This is a list where NBA Jam beats Mega Man 2.

ANYTHING IS POSSIBLE.
Avatar de l’utilisateur
Gav
 
Message(s) : 2782
Inscription : Jeu Juil 15, 2010 2:10 am
Localisation : I'll get back to you.
Genre: Masculin
Langues parlées: English, Chez Swedish


Retour vers Tutorials

Qui est en ligne ?

Utilisateur(s) parcourant ce forum : Aucun utilisateur inscrit et 1 invité