GameClub 2.5 - What Cavalier told us about The Sexy Brutale (Patreon)

Hello! So, when we chatted about The Sexy Brutale on Monday, we were lucky enough to have all three members of Cavalier join us on the Discord. Thanks to Patron Dave for hooking me up with them.

So, we spoke to narrative director James Griffiths, design director Charles Griffiths, and technical director Tom Lansdale.

Because most of this happened in a very European time zone, and because there was so much good stuff said, I thought I'd archive some of the key quotes from the discussion for everyone to read.

If you want to read everything, the Discord channel has a complete archive of the discussion.

The inspiration behind The Sexy Brutale

Charles: The game idea was a whole cluster of things, really. In the background, we had always liked the idea of some kind of "clockwork house" game. But it was only a vague ambition.

Then we actually started building something quite different. A game with much more simulated elements, but also about characters and their schedules.

Then over time, we started to realise that the best bits were the parts that were a little more characterful and orchestrated. And so we drifted back to the idea of "what if we went all-in on a clockwork scenario, and nothing at all was simulated?"

Tom: Even then, what we had at first was a far cry from the Sexy Brutale as you know it now.   We had guests you could talk to, hadn't settled onto the idea of murders or the Marquis, and all sorts of things that gradually took shape over the course of development.

Charles: And then that ambition started to get informed by a lot of other stuff we liked - interactive theatre, Japanese games like Chulip, Gregory Horror, Moon RPG Remix, killer7....

And once it was focused down, it got a lot better. And then James wrote up a setting, complete with the title!

Speaking of that title...

James: It came about right at the start. As Charles said, the games that I personally loved and inspired me were quite offbeat, strange ones. I'm such a Grasshopper fan, for example. I love the huge brushstrokes, larger-than-life approach.

[Did it hurt] sales? POSSIBLY. But I think for as many people who might be turned off by it, it's caught attention from others.The point kind of is, whether or not people LIKE it, it's not the wrong title.

It's the name of the mansion, as named by the person who built the mansion. It could have in that way been called anything. It's another opportunity to tell you something about the world and the characters. What kind of person calls their casino "the Sexy Brutale"?

You can have your own opinion on that. It's not right or wrong.But that to me was a truthful representation of the attitude of the central character.

So yes. We knew it would be provocative. But weighing up the pros and cons, I kind of thought "if I heard about this game, I'd be interested to hear more"

When they knew the idea "clicked"

Charles: Hmm...Maybe it was different for everyone else, but for me it was probably - weirdly - the building of some of things before it was even the playing

When we all started coming up with puzzles and making them ourselves with the tools as we had them

Tom: Lots of points, I think.  Although a big one WAS getting the setting and overall theme down (it wasn't a masquerade ball immediately, either).  One important decision was that the content of the game was going to be exclusively murders. 

Before then it was going to be clockwork-based puzzles... some of which would have been "prevent the murder", but others were going to be less violent.  It was lacking that focus and when we just agreed - yes, it's all murders, all the time, it's a masquerade ball, everything became a lot clearer and easier to work with.

The scare-away system

(There's a system in the game where if you go into a room that an NPC is currently stood in, time stops and you're forced to leave. I asked Cavalier how they came to this decision)

Charles: That is a good one. That is a perfect example of a basic problem - a rule that we knew we had to have, but could be implemented in many different ways.

We tried many, many things with that. That mechanic probably took more individual iterations than anything else.

Players had to know that they could not interact with anything in the room very clearly, they had to feel a strong incentive to leave, the presentation of it in look and sound had to be true to the fiction, and they had to be actively punished for refusing to leave

Those were some of the hard requirements The softer requirements were things like: characters must stop entirely for the duration, player must be able to die, etc.

Tom: (Very) early on, the guests would spot you and tell you to leave (or call an NPC to kick you out).  This didn't last long because it raised all sorts of questions about why you couldn't talk back, why you were forbidden to be in that room, what they thought of you, etc.

Charles: In the end,it turned out that "characters must stop entirely" was not a soft thing at all. They had to stop 100% because otherwise it wouldn't at all with keeping the whole thing in sync.

The slightest change of sound or atmosphere was important too, because the first time you trigger the event it becomes a big tell about the story and world in a way.

One last thing on those Masks - one of the reasons that became the design solution rather than anything else was because it communicated that the "problem" was that you were sharing a room with an NPC, while ALSO communicating to the player that it was not the NPCs themselves that were out to get you.

Which was very important, obviously. We didn't want people thinking everyone is a bad guy.

Tom: By keeping the NPC oblivious to the player, it also meant that they wouldn't change their schedule at all - they would literally continue on with their action, conversation, etc none the wiser.  Without that level of "it's the mask, not the NPC" we'd have all sorts of weird cases to pick up and deal with.  So there was that too.

Charles: The hardest part of any design process, I think - is that constant back and forth between: "Are the rules that I am trying to enforce broken, or have I just not found the right way of enforcing them?"

Designing the map

(A Patron asked: I really liked the map feature where you could see the movements of all the characters throughout the day, but I felt I didn't have much need for it in my playthrough. Are you guys happy with how this particular feature turned out and how it was utilized?)

Tom: I am yes - it does do an important job for people who get stuck, or who are planning their puzzle completion (but haven't quite gotten there yet).  It tells you where the loose threads are, and where you can learn more.  But if you're able to complete the puzzle without needing the map time-scrubber, then yes you'll probably not get much out of it after the fact.

It means that if you are totally bamboozled if nothing else, you can look at where someone in the area you are in disappears off the map and follow their story to see something you haven't seen yet.

Puzzles and "temporal reasoning"

(We had a discussion about whether the puzzles were really puzzles at all, and how Cavalier thought of the game as a temporal reasoning puzzler (in relation to my Boss Keys thoughts on Zelda as a spatial reasoning puzzler)).

James: The intention was, it would be an almost impossible puzzle without context. Then building context is the game. And yes, that's not strictly a puzzle in itself. But then once you have the context, the puzzles weren't intended to be too brutal.

Tom: We definitely thought of the game in terms of a 4D world.  The map is part of that, but also the light changing plays into that theme.  The ballroom at night is a different place, almost, to the ballroom at daytime.

Charles: Yes, I think that's very true. It's part of why I think in the end we didn't go for enormously "hard" puzzles. Because the puzzle is the story and vice versa. It is just about building a physical and temporal map and then finding your way in.

Tom: And also, the nature of the game means that we really have very little control over how the player experiences these puzzles.  One person might see Clay at the moment of his death, and that is their first experience of the casino.  Another might miss it entirely, but wander in on his corpse, and that will be a different first impression.  A third will see him drinking away and move on, not yet aware of what happens next.

So we knew this was the case and built our game around observing and understanding, and building a full picture (at which point the puzzles are ideally not too complicated at all, because the fun was the journey in getting there)

Charles: One thing that automatically makes all these puzzles considerably harder is the exponential increase in possibility space that results because of having a time loop.

Quantifying the puzzles

(That brought us onto a conversation about how Cavalier used different factors to objectively quantify the complexity of each puzzle, and get a good difficulty curve. As you can imagine, this all made me very hot under the collar).

Charles: A long time ago, I tried to come up with a language so that we could talk about the constituent parts of a puzzle with one another consistently and also be able to measure the difficulty of each one.

There were 3 main things as I recall: possibility space, execution space, and overall complexity (number of steps)

And within number of steps, there were both "green" and "white" steps. Green ones were permanent actions, and white ones were temporary (where your action would reset on a day loop)

Tom: A good example of a "green step" would be learning the code in the casino.  You need to do it once, but once you've done it you don't need to do it ever again.

So we realised that we could have complicated telescoping puzzles - as long as we "banked" it with a green step.  A good example of that is the Willow puzzle, where getting the recipe for the Hungry Charm is a green step, but you had to get the key before it (a white step).  You might get the key and do nothing (white step lost, do it again).  But then you might get the key and unlock the desk, and get the recipe (green step done!  You no longer need the key)

Charles: The possibility space was the total amount of schedule time that existed. So if there were 3 characters present in the observable area for 8 hours, there would be 24 hours.

And this actually doesn't include the sheer physical space available to the player at the time, which was also a factor. So the possibility space might be 24 hours + 16 rooms.

The execution space was the time window that existed for you to execute a solution (putting the bullet in the gun, for example)

So - you can make a really shit (but very "hard") puzzle by putting in 5 characters, moving around all day, and making there be 1 thing that you do to complete it, that is only possible in a 5 minute window

You can make it even harder by having you do 5 steps to complete it, and having them all be things that reset (the white steps)

It was really useful for us [to think of it in this way] because it gave us a way to quantify our puzzles

Chapel is about 8 rooms, two characters, five hours of possibility space... so we can quantify how complicated that is

And it also highlights how much more complicated things get when you move to the casino!  Now there are about 15 rooms (+ others in library and ballroom you can get to), 3 major characters (+3 or 4 others you can see), and the full twelve hours

The player doesn't necessarily realise that Thanos and Aurum have nothing to do with their next puzzle, so they're "live" and count towards the player's mental burden unless we can focus them away

Tom: we needed to ramp up gradually.  Chapel was two white steps, Casino was a simple green step followed by a white step.  Willow is white step -> green step -> several more white steps.

Giving characters back-up plans

(We asked about the possibility of killers having back-up plans, giving you two puzzles to solve instead of one).

James: It's about giving people a situation and making it logical enough to be solved with the information at hand.

You can get a full picture of everything that happens and all the tools that are available to you quite quickly with that. One 5 hour day to follow the victim. Another to follow the murderer. Then you have seen everything in that area, pretty much, and it's recorded on your map.

So you formulate a plan: what can stop him being shot? Plant the blank. The problem you have seen is a shooting murder. The solution you mind-map out is to prevent the shot being fired.

But if you switched the bullet and then the staff member says "curses! Foiled! Time to get my stab on!" and knifes sixpence to death, you had no way of knowing that until that moment. It would feel like an unfair, unseeable bait and switch

No final plan

(We asked why Cavalier decided against having a final challenge where the player had to foil all of the murders in one go)

James: One was that we honestly didn't (and still don't) believe that a time-based "speed-run" would be fun. You've already solved the problems. So why would having to do them all again under a strict time limit be more fun?

Why make players do that to see the ending?

The other really big consideration was that we didn't particularly care if all the murders could be solved in one day. It wasn't that important to us. We were more than happy if the only way was for Boone to be in 2 or more places at once.

Hope you enjoyed! The next GameClub will be... well, who knows. I'm also considering an episode on TSB, because we learned so much interesting stuff in this chat!

Thanks

Mark