Mystery Games, Questions Without Giving Away Answer

[Westeford]

So I was watching this video on Youtube, "What Makes a Good Detective Game?" If you have the time, I recommend giving it a watch.
https://youtu.be/gwV_mA2cv_0
After watching, it got me thinking about how I can apply these concepts to my current project. I think we can all agree that one of the biggest problems with detective/mystery games is that the questions can very easily give away the answer. It's pretty much the problem with multiple choice questions in general.
For example: 2 + 2 = ?
A.) 4.
B.) 40.
C.) 400.
Ignoring the fact that most people already know the answer, the choices gives the answer away away. The choices show that the correct answer contains a 4. And we can also assume that 2+2 is not greater than 10. So the answer is A.
More effective answers would be
A.) 4
B.) 5
C.) 7
Here the answerer has to either know how to add or make a lucky guess.

Anyway, another concept the video explains is that the question prompts the player for an answer. Giving away that there was a question that needed to be answered. Now classic point-and-click adventure games are perfect for this. Areas/evidence/suspects can be added when the player proves relevancy via a search engine like in the The Shivah.

But sometimes you'll have to resort to using questions to prompt the player to make an answer. This is especially true for courtroom mysteries like Ace Attorney or Danganronpa.

For instance in Ace Attorney, when cross examining witnesses, the player knows that they have to make an objection somehow. The game won't continue unless they do. Ace attorney does subvert this expectation with testimonies that require you to press all statements only to find that there is no contradiction. The game doesn't let the player say "No objections" and end the cross examination. Because it would either require a lot of branching paths or a lot of paths to bad ends because you didn't find the contradiction that would've led to evidence that would've been helpful later. This could be done and make a challenging mystery game, but it doesn't make for an engaging and engrossing story.

So for these instances where questions must be uses, how do we avoid giving away the answers?
One solution is by giving many more answers. This helps make the questions harder to brute force. Not that it will stop players from doing that, but that's why we include penalties to dissuade players from doing that.

The way I try to do this is by making every possible answer a choice. If the question asks for a place, I give every place possible.
For example: Where did the murder weapon come from?
A.) Kitchen. B.) Living Room. C.) Dining Room. D.) Family Room. E.) Arcade. F.) Parent's Room. G.) Sister's Room.
You get the idea.
If the answer is a person, then the player chooses from every person.
If the answer is a piece of evidence, then the player chooses from the evidence.

Though sometimes you have questions with pretty specific answers. The questions mostly comes down to framing. Like if the game asks about how the victim was killed. The answer the game will give probably wouldn't be "Stabbing in the stomach" as an answer, instead giving "Stabbing" and if the placement of the wound is important then a follow-up question would ask and have the answer be "Stomach."

But sometimes this doesn't work and another problem comes up. How do we frame a question with an answer that's not in the evidence? For instance, the victim was said to have been killed by stabbing, but the MC figured out that the victim was killed by falling instead.
Danganronpa's approach to this problem is through it's minigames like for example Hangman's Gambit. You're given a blank word and have to fill in the letters.
Ace Attorney's approach is to either just ask when the story needs the answer or just have the character say it without asking.
It's a challenge for sure.

Evidence is another thing the Ace Attorney games are known for. Since every piece of evidence is used at some point in their case, it gives away that it's important somehow. Like if you get a hamburger then it's going to be important evidence somehow. Red herrings could be used, but they still are important because the player must discover what the red herring is so...yeah.
So obviously the way to avoid this is by removing the evidence list and make the player type the answer, right? Well...NO. Just no...
Typing answers is very tricky. Because of the fact that many words can mean the same thing. So you either account for every possible answer, or you make the player do a lot of trial and error. I'd reserve typing for passwords or for numbers. It's not impossible to have the player type answers, but it's risky. Honestly this is something I don't see going away, but it can be worked with.
For instance: I had pancakes for breakfast.
Somehow this witness statement is important. So how will this information be stored?
1. It could be by itself in a list of notes
2. it could be hidden among a piece of evidence like the witnesses notebook or their witness testimony records.
One is an obvious clue, while the other is an odd detail.

Finally there's something I want to touch on and that's the end. Whether it's an accusation like in Carmen Sandiego games or a case summary like in Danganronpa. These are used to have the player prove they know the case.
Now these aren't required. Ace Attorney doesn't really go into it and at best you'd get dialogue summarizing the case without any viewer input. But this risks the player completing the case, but still confused and not understanding how anything happened.

The way Carmen Sandiego games handle accusations is basically this: You're traveling to follow the V.I.L.E agent. During your travels you find clues to where they went next, and what their appearance is. When you arrive to where the agent is, you have to accuse the correct person there who looks like how the clues described the agent. Pick the right perp, then you win. Pick wrong, you lose. There are no second chances. You'd continue this process until you are tasked to arrest Carmen Sandiego herself.
The way Danganronpa does this is by filling out pages of a comic book with pictures. After you're done the comic book plays out and it's very stylish and memorable. By the time the summary ends, you pretty much understand the entire case.
In my current project, the way I'm considering doing this is by using a set of quick questions. The player goes through the case in chronological order. Questions like "How did they do this? What did they do? How did they know this?" Basically outlining the MC's thoughts before their closing argument. It's probably not the best way, but it's what I thought of for now.

Good mysteries allow the player to figure out the solution before the story tells you, but without feeling like it gives away the solution right away. Done right, the game makes you feel like you were solving the mystery and making deductions/inductions and whatnot. And even if you were incorrect, the game does a good job showing and explaining why you were wrong. like "Oh that's why!" or "I should've seen that." Done wrong, the game makes you feel more like Watson just tagging along while Sherlock plays the game for you. Or even worse it makes you feel like the game cheated.

Of course there are subsets to this. Like the mystery plots that show you the culprit, but you have to figure out how they did it like in Columbo, but that's another can of worms.

I mostly typed this out to sort out my own thoughts, so this might won't make too much sense.
I'd still like to hear what you think. What are the qualities of a good mystery game? What games made you feel like a detective. What did those games do right? How did they challenge you? What's your general thoughts?
Thank you for your time.

[Katy133]

I'm working on my own detective game, so I've been thinking a lot about this myself.

I also saw Mark Brown's "What Makes a Good Detective Game" video, and as you say, hiding the question as well as the answer is key. And one of the answers Mark Brown gave as a solution to solve this is the Search Bar (or text parser).

I've started doing niche livestreams where I play different detective games and see how much they make me "feel like a detective". Most of them don't, because they're too linear or hold the player's hand too much for them to be able to come up with answers themself (The Wolf Among Us, Agatha's Christie's ABC Murders). Or, usually if they're older games, the answers are way to obscure/specific and I reach for a walkthrough out of frustration (Monty Python's Complete Waste of Time, Sherlock Holmes: The Silver Earring).

But I have managed to find some gems that really did let me feel that "Eureka!" moment of making a deduction myself (that "Oh! Of course! Why didn't I think of that before?!" feeling). Return of the Obra Dinn and Hypnospace Outlaw were released after Mark Brown's video, and they're very good case studies on how to use the Search Engine concept in a game. For a visual novel example, He Beat Her also uses a text parser to let the player character make choices.

The one big downside to using a search bar is that it makes it harder to set up a pre-written "Aha" or "Reveal" moment in the story, because instead of the protagonist making the deduction, it's the player. I still think it can be accomplished however.

Here's another Mark Brown video on Return of the Obra Dinn:

[RicharDann]

This was really, really helpful, I watched the video and read through your thoughts and I realized that, more or less, I've thought of all of these very same things during the time I've been planning my game, which, while it's not really a detective mystery videogame, it does have the player gather clues and figure out how they are related.

Personally I love mystery games, but I don't feel the need to be the detective or that I have to figure it out everything by myself, I just go with the flow, enjoying the story and letting the game lead my hand if needed. I think it's actually good that some games are easier than others, because there are people like me that are just interested in seeing how the crime is solved rather than solving it themselves, but still want to have a hand in it.

This is pretty much what I like about Ace Attorney and Danganronpa. In this games, you're not a real detective (more so in the latter), but you're still forced to try and solve crimes using your own logic and whatever tools you have at your disposal, but just because of that, they don't need to be overly complicated, just easy enough that it can be solved by anyone (anyone that knows how to use their brain), and at the same time not be blatantly obvious or easily guessed. I think both franchises manage this very well, despite the obvious ups and downs. They are still challenging because, sometimes the answers are clearly evident, but sometimes they're not, and that's when you need to, to quote Mia,"turn your thinking around" or "think outside the box" in order to progress.

[Autumnotopia]

I'm a fan of most of the big murder mystery games (Ace Attorney, Dangan Ronpa, Umineko, etc), but weirdly I think that the game that really captures that 'aha', detective-y, feeling most for me is actually something more like Scribblenauts. It's not a visual novel, but it does a very good job of navigating the problem of alternative solutions. For me, one of the most frustrating experiences in murder mystery games is having a line of thought that you can't translate into the gameplay. Like in Danganronpa I'll have a theory, but obviously since I'm not Naegi or whoever I just have to accept whatever line of thinking the game presents.

That said, I don't feel like always having an actual literal puzzle solving mechanic is that important in a mystery game (probably because of the above mentioned issue). I'm always going to come up with theories of varying accurateness that won't be accepted or acknowledged in game. Therefore, I appreciate just being able to put things together on my own without having to be mediated by the system. Umineko does this, though it's definitely more in the realms of 'literally just a linear book'.

That said, if you have to gamify it, obviously that's really hard to master. I run a Dangan Ronpa tabletop game, and trying to guess what the players will be able to figure out is always difficult. I appreciate the thought of leaving things open, but in my experience, giving players a ton of choices usually just ends in frustration. Adding in random red herrings is something that I've noticed that particularly frustrates players, because usually they expect every item to be relevant in some way. Sort of the Chekov's Gun thing.

It's not always possible, but a good way to check the difficulty on your puzzles/mysteries is to see if someone's comfortable play testing it for you locally. Sometimes it makes them more nervous and thus makes it harder for them to solve the mystery, but it definitely shows the rough spots that could use a little more foreshadowing.