Design: Brainstorming the Sleuthhounds Story Board
May 1, 2015
Last week I discussed analyzing dialog trees in adventure games to see what worked about them and what didn’t. I did this analysis because many adventure games, fun though they are, only use dialog trees as extended cutscenes that need to be un paused every so often. I felt that dialog trees could be more than that with a bit of work.
Sleuthhounds, being a mystery game, involves meeting with suspects and interrogating them to get valuable information. I had this implicit feeling that dialog trees could be used to do this interrogation in a more meaningful way than simply exhausting all possible dialog options.
I also had an initial grain of inspiration from Tex Murphy: Martian Memorandum. This game is also a mystery game and you also get to interrogate suspects. An interesting aspect of this is that each suspect is shown in close up when talking to them so you can gauge their reactions to tell if they’re lying or not. However, even if you believe a suspect is lying about a particular topic, there’s no way to push them further to try to make them crack.
When I came to work on the interface design for the second Sleuthhounds demo, I wanted to have an “interrogation board” of sorts. I thought that perhaps a suspect close up could be shown, Martian Memorandum style, and that you could ask questions. As the suspect answered you could observe their responses looking for tells that would help you proceed further. That idea stalled out early in the design phase. It felt too much like the dragging bits in The Secret of Monkey Island where you have to learn a suitable set of sword fighting insults as I discussed last time. Even worse, you would have to do that for every suspect you met.
And so started the design journey to the interface that now adorns the second Sleuthhounds game.
After I threw out the initial idea, I tried to think of other ways an interrogation system could work. One of my first thoughts was to go a little RPG-ish and introduce stats into the game. Each suspect you met would have a level of annoyance, a level of resistance, and an attitude towards the player. As you asked questions, the more personal they were the more annoyed the suspect would get but the closer to real answers you would come. A character would be less annoyed if they had a good attitude towards the player.
The stats idea would have been nice from a programming standpoint since it would be all about numbers and levels and so on. However, I couldn’t think how to make it work from a story standpoint. How would a player gauge which questions would be too annoying? How would they know which lines of inquiry to follow? So another possible idea hit the floor.
For some reason, my thoughts turned to the use of playing cards. Specifically to the use of Tarot cards. The cards of the Tarot are essentially used to tell stories. I thought that perhaps you could observe a character from afar or learn about them from other characters and that would give you different Tarot cards representing that character.
You could then confront the character and play different Tarot cards. For each card you played, the other character would play another. You would then respond with another card in turn until finally you had a complete picture of the character. The more I thought about this idea though, the more abstract it seemed to be. There was a large gulf between the archetype nature of Tarot cards and how they would apply to specific characters in a game. And the learning curve for the system seemed like it would be really high. Another idea down, though one I would revisit a bit later.
Ok, next attempt. How about the venerable children’s game Hot and Cold? Maybe you could ask a character questions and “somehow” get an indicator of if you were on the right track to the truth or not? I never got far enough to figure out the “somehow” because I quickly realized that people would just grind through such interrogations by watching for the “hots” and “colds” and not paying attention to the logic of the story. I really wanted to develop a system that would reward players for thinking through the actual story of a character.
At this point I took a step back to think about what other games had done with other character-to-character interactions. Douglas Adams’ Starship Titanic had little dials showing the attitudes of characters that you could mess with (the characters not the dials) to affect how helpful they were. Hm. The Legend of Kyrandia Book Three: Malcolm’s Revenge let you set the attitude your character took when approaching others, being truthful, lying, or somewhere in between. Hm hm. Blade Runner provided a similar character attitude control and also let you administer a lengthy “Voight-Kampf” test to individuals to determine if they were replicants or not. Hm hm hm.
I thought about all these games and others. I thought about implementing some sort of visual node system a la Pipe Dream. Too mini game-ish. I thought about having a visual length of chain where you had to pick specific icons to fill each link with in order to follow the chain all the way through to a successful answer. Too abstract.
Ultimately what got me unstuck was not another computer game but a television show. I’ve been a big fan of the series Castle, starring Nathan Fillion and Stana Katic. Fillion plays the title role of Richard Castle, a mystery writer who helps Detective Kate Beckett (Katic) of the NYPD to solve crimes. Invariably, Beckett hauls some suspect into the precinct’s interrogation room. Based on the evidence gathered to that point, she then outlines a scenario or story that points the finger of guilt at the suspect.
Light bulbs went on. Light bulbs that cast their glow on the much earlier Tarot card idea I had.
I had discarded the Tarot idea because I felt that the cards were just too abstract. Then I got to thinking, what if instead of having abstract Tarot cards players had actual story sentences to work with? The sentences could be gathered by talking to the main suspect, other characters, or even examining other evidence or locations in the game. Once all the sentences were collected the player could confront the suspect. The player would then be tasked with putting the half dozen or fewer sentences in the correct order to tell a story.
Once I had this idea it occurred to me that I had seen something very similar in another game from years back. I think the game was Dragonsphere although I didn’t actually replay it to check. My hazy recollection is that you had to construct a song or a poem from different phrases. Although the details are sketchy in my mind I do remember thinking it was a rather clever puzzle at the time.
I also recently replayed Myst IV: Revelation, which technically has a similar sentence organizing puzzle. However, that one is extremely bogged down and muddied in the use of abstract shapes to represent the sentences and a sort of slider puzzle type interface. Those extra layers of abstraction confirmed for me that I was right in abandoning the abstract nature of the Tarot cards as the puzzle in Myst IV feels like a lot of churn with very little progress. But I digress.
After struggling through many different ideas and design sketches it seemed like I’d finally found one that would fit nicely into Sleuthhounds. The question was, would that actually be the case or not?
And the answer? That will have to wait for another blog post. Come back in three weeks for my wrap up on Sleuthhounds’ story board interface. Why three weeks? Well, next week I have to finish off my series on eBook publishers and the week after is the next installment in my series on the benefits of writing recurring comic strips. So three weeks. See you then.