Sleuthhounds Production Update – Designing for Two Characters
January 29, 2016
Recently, I’ve been doing design work for the next Sleuthhounds computer game. It’ll be the first such game to feature together both Pureluck Homes and Jane Ampson, the playable characters from the first two games. I’ve discovered that having two playable characters in an adventure game leads to some interesting design challenges as well as some interesting design opportunities.
When I was developing the second Sleuthhounds game, The Cursed Cannon, I quite intentionally included a sequence where the player would have control of two characters. I knew that in future games I would have Pureluck Homes and Jane Ampson together and so I wanted a chance to work out the technical side of a game that supports two playable characters (PCs) at the same time.
For The Cursed Cannon I kept the sequence with multiple PCs—in that case, Jane Ampson and her friend Amelia Deerhart—quite limited. The two characters were confined to a single room with no other characters to interact with, only a couple of inventory items between the two of them, and a very small number of interactive “hotspots”. Basically, I had wanted to keep the scope of the sequence very small so I could focus on the technical implementation rather than the complexities of story and game design.
In turning my attention to the next Sleuthhounds game I was confronted with having to design a game with two PCs completely in mind. This lead to some design challenges and opportunities that simply didn’t exist in the first two games.
Division of Labour
The most obvious design challenge with having two PCs is giving them both something to do. More specifically, it’s about giving them roughly the same amount to do. Since I setup both characters as main characters in the first two games I knew that I’d have to treat them equally in any games that involved both of them at the same time.
Fortunately, at about the time I really started working on the design of the first game to incorporate both PCs I also stumbled across an old blog post on puzzle dependency charts by Ron Gilbert, one of the designers of such classic LucasArts adventure games as Maniac Mansion, Zak McKraken and the Alien Mind Benders, and The Secret of Monkey Island. In his post, Ron depicts constructing a puzzle dependency chart, which basically is just like it sounds. It’s a diagram that shows the dependencies between puzzles (i.e. which puzzles must be completed before other puzzles, which puzzles can be done simultaneously, etc.).
Ron uses these puzzle dependency charts to make sure his games don’t become too linear. He watches for opportunities where the player could be working on several separate puzzles at the same time. Each puzzle is given its own box with lines connecting it to the puzzles that must be solved before it and more lines connecting to the puzzles that can only be solved after it. It occurred to me that this approach would also work really well with multiple characters if I color coded the boxes to match the characters. For example, any puzzles that could only be solved with the Homes character could be coloured blue and any that could only be solved with the Ampson character could be coloured purple. Puzzles that could be solved by either could be left white. Then, I could simply count up the number of blue and purple boxes to determine if the characters were roughly balanced.
This approach to the puzzle design also worked out really well from a writing/character point of view. In the first Sleuthhounds game I tried to establish Pureluck Homes as more of a logical, deductive detective (a la Sherlock Holmes). In the second game I tried to establish Jane Ampson as a more intuitive, “people-oriented” investigator (more like Hercule Poirot, but not as obsessive compulsive). I’ve tried to design the next Sleuthhounds game to take advantage of these character traits to match puzzles to characters. It’s a different way to design from dealing with only a single PC so I’m still practicing and refining this.
A while back, I wrote a blog post on what I saw as the deficiencies of dialog trees in computer games. Dialog trees are presented in games of all sorts—adventure games, RPGs, even some flight simulators—when talking to another character. In these trees, the player picks from several options what they want to say to a character. As I detailed in that other blog post, from a straight design perspective there are several issues with this particular game mechanic.
When I started designing and writing for the next Sleuthhounds game I very quickly ran into an issue with having multiple PCs being able to talk to NPCs (non-playable characters) in the game. If the player could switch between Homes and Ampson and talk to a suspect, then traditionally separate dialog trees would be needed for both characters. That doesn’t seem too bad until you consider the question of how does the player know which character they need to use to get key information from a suspect?
Do I force the player to try one dialog tree from one character and then try the second dialog tree from the other character? That seemed like it would lead to a lot of unimportant dialog for the player to slog through.
Do I setup the dialog trees for both Homes and Ampson so that the player can elicit the same information from the suspect regardless of which detective they’re currently controlling? That seemed like it was simply doubling the amount of implementation work without really giving a lot of value. It already takes a long time to build these games so doubling the work for no value wasn’t an attractive proposition.
Then I got to thinking about the natures of the two characters in my games. As I mentioned above, Ampson is the more people oriented character and Homes less so. It dawned on me that I could both simplify the dialog trees and make them more effective as a game mechanic if, when talking to a suspect, I only presented two options, one for what Ampson would say and one for what Homes would say. Depending on the suspect, I could then have either Ampson’s approach or Homes’s approach be the path of least resistance. And if the player picked the other character? Well, then maybe the suspect would still give the information, but only after the player solves an alternative puzzle.
I got really excited by this idea of having different outcomes based on the character options chosen. It means that the choices a player makes in a dialog actually do have some impact on the play experience in the game, rather than simply being a grocery list of items to chug through. It also means that I can introduce a bit of replayability into the game by having spots where the player has to do different things based on the dialog choices they make.
It’s a new way of approaching dialog and means having to design so-called “alternate paths” into the game. As such, it’s something I’m still experimenting with and need to get a little more practice at. But as I said, it’s quite an interesting and different take to dialog than I can recall ever seeing before.
While I’m quite pleased with how the first two Sleuthhounds games came out, one of the things I wished I could have done better was to include more humour in both games. One of the reasons I couldn’t was because both games, at their heart, have a mystery to solve. As such, when the player is examining and interacting with their environment, they need to be given the facts about what they’re seeing and what’s going on. It’s very hard to do that in dialog and make it funny without the humour obscuring the facts. Most of the dialog has to be played straight so that the player can glean the information they need to have to be able to work their way through the game.
As I started writing dialog for the next Sleuthhounds game I realized I had a wonderful opportunity. If the player was playing as Pureluck Homes and they decided to investigate a bookcase, for example, then Homes could give a factual indication of the contents of the bookcase. As Homes is describing the bookcase, Ampson can then pipe up with humourous asides on the contents. In essence, I can have my cake (providing key facts to the player) and eat it too (incorporating more humour into the dialog).
Having multiple playable characters in the game provides a lot of interesting design opportunities. Opportunities that I believe will allow future Sleuthhounds games to have an even more unique identity and feel from other adventure games that are available.
Beyond the writing and game design, multiple playable characters also has a very real impact on certain production choices. But that’s a tale for another time. Check back next week to see the more practical side of dealing with two characters.