Originally posted on Steam
These screenshots show Midori, the protagonist of Broken Spirits, exploring her wealthy family's enormous mansion.
August 5th, 2004.
Returning home after a suspicious accident leaves her broken and burned, Midori Kariyama finds herself pulled into the chaos of her estranged family, who are being targeted by a serial killer.
My goal with Midori was to create a stark contrast with the protagonist of the first game, Noa. Midori is athletic, brash, loud, acts without thinking, and embraces being a loner. She has a swagger to her that attracts people's attention when she walks into a room, as well as her arresting scarred appearance.
Her cat Naughty Frump always rides on her shoulder. She tends to wear military-style camo pants and jackets. She is actually kind of dumb, and isn't too good at piecing together logic.
Katsura is Midori's mother. She is the first female CEO of Japan, and is ridiculously wealthy.
For her character and design, I wanted to make sure she didn't echo Noriko too much — she's more cold than outright mean.
Ryou is a (terrible) rapper who is hoping to take his "talent" overseas someday! He's Midori's cousin.
His design is sort of the usual "gangster" anime stereotype, turned up to eleven with the gold face-studs and full body tattoo. He's actually pretty nice and sort of a yakuza poser.
Midori's father. He's been a lot quieter since the divorce and has taken up hunting.
I always pictured Gunpei as this big, friendly guy. Of course, certain events in his life have made him a little more gloomy.
Yoshinori works as a stock broker. His #1 award-winning stock tip is to invest in the Urane family company. For his design, I wanted to make sure you could tell that he's on the Urane side of the family, so I copied Katsura's eye shape. Gray hairs are sort of creeping slowly over him.
Midori's cousin, from Katsura's side of the family. He aspires to work in film and television. As soon as I thought of Mokei as a director, I knew he should probably be in a very flashy outfit with a monochromatic color scheme. I chose yellow to reflect his energy in a friendly way. It's kind of a similar coat as the Ghost Trick character — big hair and shoulder pads suggest a lack of substance. I added the buttons for a unique touch.
Ume is the only character in the game from Broken Minds. Ume claims to be trying to steer the YPDA in a less criminal direction under her leadership, which is partially thanks to Takuma and Yuzuki not being part of the organization anymore.
Ume was really hard to age up, because my original design of "29-year-old" Ume already had lines under her eyes. It was difficult to design her with longer hair, because it just didn't look like Ume unless her hair was short. But I knew that, being a whole decade older, she would probably have changed her hair style. Then I figured out the "razor line" and finally, something clicked and she looked like her snarky self again.
The reason I chose to focus on Ume for Broken Spirits and not Takuma and Yuzuki is because I felt that Broken Minds had more of its focus on Takuma — he's the one making all the deductions and most of the decisions. Yuzuki had the least focus of any of them, but she has more relevance in the third game. Ume's particular brand of detective work is at the forefront of Broken Spirits — so it was a natural choice. The reason I left Takuma and Yuzuki out entirely is because I wanted to shake up the dynamic of the detectives' interactions. You'll notice that in Broken Minds, Ume often follows Takuma's lead. Ume looks up to Takuma, and often imitates his mannerisms — a lot of her snarkyness and rude remarks come from wanting to impress Takuma. I thought it would be interesting to see Ume in charge of the YPDA without Takuma's influence, and without Yuzuki's extravagant spending and criminality to worry about — what would Ume be like? Anyway, you will hopefully see more depth to her in this game.
Note: the characters listed here are not the entire cast.
THE STYLE OF THE GAME
The aesthetic of Broken Minds leaned more towards Western architecture with slight Japanese elements, but I wanted to flip that dynamic for this game. For the UI, I'm drawing on influences from obscure Japanese horror games, cyberpunk, and websites from the early 2000s.
The game design philosophy from Broken Minds was sometimes deliberately confusing, and sometimes just plain confusing, and I think it was too concentrated in the second half of the game. The gameplay in Broken Spirits is being approached from an entirely different angle.
It revolves around rewarding two things: exploration and deduction, and both of those things lead into each other: exploring allows you to make deductions, and deducing allows you to explore more.
A simple example is exploring a room and finding a key, deducing the key unlocks a door in the room, and unlocking the door and exploring the next room.
Remember the "Social Link" gameplay I mentioned in my last post? In Broken Spirits, they're called Social Trees. The idea is that by exploring, you can find various objects relating to the characters. Combining these objects in the right ways lets you unlock conversations that deepen your relationship, "growing the tree". When you reach the end of these conversations, you'll gain access to a room.
Explore and find objects > make a deduction about objects > unlock new area to explore.
Having played both Persona 3 and the Danganronpa series, which both use similar gameplay, I hope to alleviate some of the frustrations of those systems while retaining the positives.
For instance, with Danganronpa, there's very little incentive to use the Social Link system, from a gameplay point of view. What's the gameplay reward for spending many hours choosing the right gift and answering the right questions? Abilities in the Class Trial sections that I never used? It didn't feel worth it. I was discouraged, not only because I often got the multiple choice questions or gifts wrong and ended up making zero progress with characters, but because the character I was attempting to progress with would die!
Persona 3's system is better in my opinion. I actually ended up completing many Social Links because doing so gave me a useful gameplay reward, plus what I had to do to advance was always very clear. However, like Danganronpa, you have to kind of "main" certain characters because it's very hard to max all of them in one play-through, unless you look up a guide.
In Broken Spirits, you aren't constrained by time, but you are constrained by sections. Let me explain a bit.
Broken Spirits borrows the concept of "Freetime" — which is defined as the time between major story events (aka MURDERS!!), where you can just relax and do whatever.
If you would like to skip right to a main story event (IT'S A MURDER, FOLKS!!), you are able to do so.
But otherwise, you can spend Freetime exploring and growing Social Trees. Each Social Tree is split into three horizontal sections: Roots, Trunk, Branches. In one day of Freetime, it's possible to unlock the Roots of every character's Social Tree. You have to wait until the next day of Freetime to unlock the Trunk, and then the next to unlock the Branches.
The three horizontal sections are split into three vertical sections: three parts to the Trunk, three parts to the Roots, three parts to the Branches. So in total there are nine conversations to have with each character.
Hope that makes sense!
Once again I'm drawing on many different styles for the music, but this time more geared towards a "2000s" sound. My influences for the soundtrack are the Witch Hunter Robin, Persona 3, Beyond Good & Evil, and Ghost in the Shell Standalone Complex OSTs. I'm also incorporating waaay more traditional Japanese elements like koto and shamisen, and I've written many tracks in the Japanese scale.
The OST will be a mishmash of traditional Japanese music, orchestral, techno, house, neo-jazz and alt-rock.
How the heck do all these things come together, you might ask?
Well — good news! I'm going to be releasing a preview of Broken Spirit's OST once every month from now until the game's release. Right now I've composed over 30 tracks, but most of them I want to keep a surprise, so that's why there's a slow roll out. Here's the first track I've released:
I decided to release "Crimefightingers" first because I think it effectively illustrates a large sample of the different styles I'm meshing together in this soundtrack.
Unlike Broken Minds, this time I have several leitmotifs that will pop up throughout, and tracks specifically for certain characters that play during their Social Tree interactions.
I've also invested a little in hiring some real musicians to record melody lines for certain tracks — it's thanks to your support of Broken Minds that I'm able to do things like that which improve the experience.
REVEALING TOO MUCH?
If you're worried about this post spoiling too much of the game, please don't be. This game will be very long, and there are very significant plot elements and characters that I'm leaving out. How long will it be? Well, I miscalculated how long it would take other people to play through Broken Minds. I can complete all of Broken Minds (every route, plus Hard Mode and Detective medals) in roughly four to five hours. However, I read quickly, plus I know the answers to the puzzles. It turns out it takes most people that long to finish only one route. Broken Spirits does not have multiple routes (though I have decided it will have two endings). I'm adding support for a maximum of two play-throughs with a New Game Plus mode.
However, as it currently stands there are six chapters outlined. I don't know how long each of them will take to play, however it's reasonable to assume they'll take at least as long as the deduction section in Broken Minds, so about two hours. I'm going to estimate one play-through taking 12 to 15 hours WITHOUT Freetime, but I will update this estimate as I develop the game.
I wanted to at least describe the Social Tree gameplay in order to get some feedback — does this sound like a fun system to you, or a boring system? The time to tell me is now!
WHEN WILL I FINISH IT?
Well, seeing as Broken Minds took two years and this game is like 3-4x the length... just kidding. It's a bit more complicated than that — there are a lot of factors influencing the amount of time it will take to finish this game.
SHORTENING THE DEVELOPMENT TIME
1. Unlike Broken Minds, I'll have a complete outline for the plot and the mysteries. This tripped me up last time because I kept having to revise stuff or squish things I'd forgotten into other puzzles.
2. I don't have to invent the universe. Broken Minds involved a lot of taxing creative decisions that slowed down development. However, now I know what the YPDA is all about, the creative goals and rules of the series, and how to portray the world.
3. Less complicated mini-games. Did you know that creating the eight Logic Trains in Broken Minds took about six to eight months? I was probably even more confused creating them than you were playing them! While coming up with, designing and coding the mini-games rarely took more than three or four days, implementing all the puzzles took a huge amount of time. Broken Spirits' mini-games are less complex, due to the fact that I don't have to account for two versions of each mystery.
4. I don't have to render and code exploration. For Broken Minds, each time I created a new environment I needed to render about forty to sixty different images. Each of those images had a lot of soft lighting, which increases render time. Each render probably took at least two hours. Each one also needed to be brought into Photoshop for post-processing, which was very tedious. And every time I added a new angle to an environment, I had to code it in and test it. Developing Broken Spirits in Unity, I don't have to re-render a background if I want to change something. I do have to re-bake the lighting in the scene, but that takes less time than rendering one of Broken Minds' backgrounds, and I don't have to do that every time I change something.
5. I don't have to draw CGs. I'm the first to admit that the CGs in Broken Minds look pretty bad. It's not my favorite thing to do — I'd much rather just design characters. In Broken Spirits, there won't be CGs. Rather, there will be cinematic angles during conversations on the 3D character models. This takes literally seconds to implement, whereas drawing CGs takes hours. Because of this, I will be able to have wordless, visual-based sequences, which cuts down on how much I need to write.
6. Writing multiple parallel routes slowed everything down. Every time I wrote a sequence, I would have to write six different things for Noa to think about it. Now, on one hand, it was a fun writing exercise. But on the other, those were five extra lines that could've been contributing to linear instead of parallel story progression that the majority of players wouldn't even see. And it could also be very tedious to essentially rewrite what I'd already written. I was too exhausted by the end of it to do much revision.
7. Adventure Creator. Broken Minds was created using Ren'Py, which is incredibly easy to use and super fast. But Adventure Creator is even easier to use, and it uses blocks to code instead of a special brand of Python. As long as I use it for what it's intended for, a lot of the infrastructure is set up already for me, and I just need to make unique graphics, which is something I enjoy more than trying to figure out how to code.
LENGTHENING THE DEVELOPMENT TIME
1. Unity is slow. Running Unity uses up memory on my computer. I've seen disk space slowly drop while I have it open, and if I keep it open over night, that becomes a big problem. Plus, I often need to have it open alongside other memory-heavy programs like Cheetah3D, Photoshop and Animate CC. If I have a huge scene with a high polygon count, dragging script blocks around in Adventure Creator gets way slower. After some bad experiences with this, I decided to split the Kariyama Mansion environment into three sections, and now it's much faster.
2. Obviously, the story is longer. As I mentioned before, this game will be at least three times the length of Broken Minds, if not more. The outline is still not entirely finished — I'm taking my time with it to make sure it's really solid. And I've been working on that outline for a year. So yeah.
3. More minigames. Even though the mini-games will be less complex, there will be more of them, simply because there are more mysteries. Developing them could take longer than the Logic Trains, or it could take the same amount of time. But it probably won't be shorter.
4. Freetime. Broken Minds didn't have Freetime, which will require that I create many different collectables, a shop interface, and of course the Social Trees. There are over twenty-one characters in the game that I plan to give Social Trees. YIKES !
5. Coding is harder. I know I just said that Adventure Creator is easier to use than Ren'Py. It's true, but only for things that Adventure Creator was already built to do. Some of the ideas I have could get very cumbersome to code.
6. Character models. As you see in the screenshot, as well as sprites each character gets a 3D model. These take around four to six hours to create and animate. So far I've made eleven that are sort of "in-progress" and will need to be tweaked later. Each model needs to have idle, walking, running, and talking animations just to start with, with more animations added to enhance dialogue.
7. The environments are larger. The sets in my games are extremely important to me, so I spend a lot of time on them. And some of the sets in Broken Spirits are 5 to 10 times larger than the ones in Broken Minds. The Kariyama Mansion alone has more rooms than every location in Broken Minds. With more space comes the need for more props. Fortunately, making props and environments is something I enjoy, but it does take a long time — I've been working on the Kariyama Mansion set for three or four months now and I would say its only 50% complete. I have a new system where I try to complete a room a day. So far I'm only three rooms into that schedule, but it works, for the most part.
With Broken Spirits, I'm hoping to address many things that I feel were poor choices that I made with Broken Minds/Egg/Spiral.
One of those things is stigmatizing mental illness.
Most of the characters in the Kowareta Kokoro series are terrible people, and some of them are murderers, and right now most of those murderers have mental illnesses that caused them to become detached from valuing human life.
Though their motives for murder are more than just that, I regret relying on mental illness as the underlying justification for them to be able to murder people, not just because it's lazy and overdone, but because the majority of murderers in the real world are sane people (the revenge-seeking boyfriend, the desperate robber, or the friendly neighborhood police officer who shot someone who was unarmed "in defense", for example). Going forward, mental illness will not be a scapegoat in any shape or form for the murderers' violent or morally questionable actions.
To be clear: Noa's mental illness was not an underlying reason for her to murder her parents and Uzumaki. However, calling Noa's murderous persona "Psychopath" was, I think, in poor taste. Uzumaki was meant to exhibit narcissistic traits that also link him to psychopathy, and Hato's savant-style delusions also hinted at unspecified mental divergence.
I think that there is a reason why psychopathy has that stigma, but it's not entirely justified, and of course, sweeping generalizations should be frowned upon, generally speaking. It's a nuanced issue, especially once you start talking about psychopathy in terms of ASPD, and I think that's worth exploring.
Broken Spirits will tackle the stigma around mental illness and ASPD head-on in what is hopefully a productive, interesting way that won't be preachy. With Broken Spirits' larger cast, I will be able to branch out a bit and have characters who are not murderers, or stalkers, or victims, who are good people, and who happen to have a mental illness. The actual change here is that I'm letting some characters be good, in other words. But don't worry. It won't be too many.
Another improvement will be the release of the game.
Broken Minds was first released on itch.io. The only marketing I did was two posts on visual novel forums.
Then I released it on GameJolt, where nobody bought it. I'm probably not going to bother with GameJolt next time.
Several months later, after I'd repeatedly updated the game with new backgrounds and features like the Flowchart, I released it on Steam. It hasn't done amazingly well, but it's done better than itch.io, mostly thanks to being included in a few bundles.
Then I reluctantly wrote a Reddit post about it, which didn't result in any further sales — and now, here we are!
Looking back, it was a pretty terrible rollout. I have no enthusiasm for marketing my game, I kind of just wanted it to sink into obscurity, where maybe it would be found by a tiny group of people and become a cult thing. But Broken Spirits will be a little different.
For one, I'm going to prioritize Steam as the release platform, and instead of a full release, I'm going to make Broken Spirits an "Early Access" title first. This allows it to have two releases, in a way. As far as marketing goes, I'm attempting to be more public about my development process — in fact, you're reading marketing for Broken Spirits right at this moment! You fell for my trap!! Now you'll have to buy Broken Spirits and the entire line of limited edition Yuzuki dolls and the Utsu Neko t-shirt I haven't yet made!!
The final improvement is that I am attempting to make some of the characters attractive this time. I know that Takuma is incredibly sexy, but his shining abs did not rake in the big bucks, as I had hoped. This is priority one for me, because all I care about is money, just like the YPDA. It will be extremely difficult to draw, but one... maybe two characters in Broken Spirits will not have broken-looking fingers. I will be putting their attractive hands in all of my clickbait thumbnails — so be prepared for this series to really take off.