After-Jam Postmortem


Hello, everyone! This is going to be a long post, so make sure you have some refreshments, just in case.

This is my first time doing something like this, so my thoughts may be a little bit here-and-there. Postmortems are sort of grisly things; you have to tear apart something in order to diagnose the story of its life. It's something which is necessary in order to improve on your skills and get better at something, so of course, it might be a bit messy or otherwise meander. 

Every Jam period, I hope to try and learn a few things-- intentional or not-- and try some new angles I think could bring a unique feeling to a project. In this Jam, I immediately wanted to do a handful of things:

  • Make choices matter more
  • Create a lived-in environment
  • Have a full voice cast
  • Find a better writing process

So I'll go through these four attempted goals first, then talk about more as it comes to mind. 

Make choices matter more

When Pedestal and I were first starting to formulate this idea back in March, there were two games which immediately came to mind: "Coffee Talk" and "VA-11 Hall-a:. This was something I had always intended to create at some point, and Pride Jam's longer window and theme lent a great opportunity to explore the themes present in both of those titles. 

(Coincidentally, "Kokoro's Doki-Doki Sabbatical" was initially inspired by "Coffee Talk"'s conversation system, and meant to be more of a static speed-dating game before it ultimately morphed into more of a walking simulator. Suffice to say I'd been toying with this idea awhile!)

Something which had also inspired me to go down this path was Megapig9001's "Heart of the Tiger" from Pride Jam 4; the locale of a bar is an inspired one because of the bar's place in society. The bar is a meeting place where communities often gather to blow off steam and get to know each other better in a way that coffeeshops can't.

Gensokyo is particularly unique in that respect; both humans and youkai congregate in these places and can intermingle with one another.  "Heart of the Tiger" was a thoroughly solid concept and story. But I had a nagging feeling in the back of my head that there was unexplored territory to be discovered in such an environment.

The singular focus on romantic relationships was ultimately what bothered me. 

Of course, it being a Pride event, romantic love is impossible to avoid, and trying to avoid it too long is a bit silly. But at a bar, all sort of relationships come to the fore, and I wanted to navigate as many of those possible relationships as could fit.

This lead to the research phase of the Jam; I immediately played "Coffee Talk" and "VA-11 Hall-a" and watched various let's plays to see how people reacted to choices or how they tried to think things out. The more I played through and watched, the more I noted that NPCs were incredibly patient. Like, really incredibly patient. And the main characters always know just the perfect thing to say. In the world of bartending games, the main character seems to have such charm and charisma, they can stop time and think through the perfect response as if it's nothing.

But real life isn't like that. Real life bartending requires you to be on your feet both literally and figuratively. Customers can be demanding and impatient. They might not consider how busy you are or that you're not into a certain topic. They'll notice when you're spacing off. Choices (or the lack therein) affects their experience. So our main goal was to create a game which made the urgency of the situation front and center.

We introduced the idea of timed questions during conversations to try and help with that. I figured that enforcing a time limit would test the player's ability to retain conversation pieces and consider their actions while still being attentive. Typically, a choice made would grant the player a certain rank depending on their answer and open up the next bit of dialogue. However, there was also a mechanic where if time ran out, the character would instead chastise you for your indecision, and you would lose ranking as a result.  Unfortunately, I never saw this happen in any playthroughs on choices of consequence.  I blame my making them too long; each defaulted to about 10 seconds so you could read the choices and deliberate a little bit, but in execution, it ended up being almost too much time on most decisions. Players often clicked a choice within three seconds, something I did not account for. If I had another chance at it, I would make the choices considerably shorter in order to up the challenge of the game. I also did not account for Ren'py's ability to go back and re-choose a choice or for the history log. It effectively neutered the idea basically on launch as a result.

Another design which was implemented to make choices matter was the drinks menu. Much like "VA-11 Hall-a", we give the player a lot of leeway in the drink they create for the characters. This allows for more personality and ownership for the player, and the opportunity for unique dialogues with the characters. However, we were unable to execute this exactly how we would have liked due to some unforeseen circumstances in programming. We had to make a very tough choice to switch engines with just three days to go before the deadline, and it cost us the ability to reasonably implement this feature. In the original intention, if the player made a mistake in the blending process, or served the wrong drink, there would have been a "mistake" counter which went against the player's overall score and would have ultimately  affected the ending if enough were triggered.  In this version, it merely meant a lower "commission" score, which while still somewhat of a penalty, is much more forgiving. (More on this later.)

There was a lot of great work we did on this front, but some of the tougher choices had to be sacrificed by necessity due to the engine change. If we ever have a chance to revisit the game, it would be in an engine which supports something closer to that real vision.

Create a lived-in environment

One of the best compliments I every received for a jam project was the game felt "lived-in." Ever since reading that remark, I have made it a goal in every game for it to get to that feeling. A grounded, realistic setting can add so much to a story, and that was most certainly something I wanted to get across as the game's writer and director. 

This was exclusively a one-scene story, however, which makes the definition of "lived-in" a little more difficult.  We planned to have a handful of things in order to make up for this deficit, primarily on the audio side. The biggest thing was ambient noise; even when the bar was closed, there needed to be noise in the background at all times. The idea was to create a true bar atmosphere, from being behind the counter to hearing the music and din of the crowd. 

When mayama and I were conceiving the music, the original plan was for each day to have a different tone. Day 1 is meant to be cold and strange, Day 2 is meant to warm up a bit but still show apprehension and Day 3 is meant to completely open up and have a hint of dreaminess to it.  We very much achieved this, and I am proud that the songs managed to get those messages across. mayama did a good job finding the emotions of each song, delivering an ambiance which served the narratives happening in front of it very well.  A combination of my lack of ability to offer proper feedback and changes to mayama's schedule ultimately doomed the idea of a "concert" style setting I was hoping to achieve. The original concept was that each day would have a handful of musical "sketches" interlaced with bar noise in order to emulate a house band.  We did achieve this somewhat, with one musical sketch and a plot of bar noise looping in every chapter, but I do wish we could have done more there. 

The choice to go away from "Touhou" music was also intentional, and I do wonder how people felt about it, as we didn't get a single comment on it. It would have been very easy to just say "we'll make VA-11 Hall-a music, but make it Touhou" and go to town for higher scores, but I opted to stick to the principle of a true American bar scene, and said as much to mayama in the creation process. We used no themes of any of the 13 characters in the game, and used mostly Western-style music in the concepting process. I do wonder if we should have at least done Japanese-inspired original music, but I stand by the initial decision.

Quickly back to the visual choices: the bar itself is based on a bar from my college days, which was famous for its all-wooden interior and exterior, as well as the fact the owners actively encouraged its patrons to write all over the walls.  I felt this was something which would help make the game feel more home-y. It certainly worked; I saw several people pause to take the time and read all the graffiti, and that felt really nice to see in real time. If you're curious about some of the references, here are a few of them:

  • The rose in the center is a reference to Roz Arai, a variety VTuber who often participates in Touhou Jams with her crew. (twitch.tv/roz_arai)
  • The "ticket to Pen Shinma" is a reference to Nonatsu Kazari, a bartender VTuber who is an acquaintance of mine. They also voiced Rinnosuke and created the drinks menu for the game. (youtube.com/kazarinonatsu or twitch.tv/kazarinonatsu)
  • The rat who says "I am in your walls" is a reference to AspreyFM, a famous Touhou Project streamer on Twitch. (twitch.tv/aspreyFM)
  • The "dirty ninjas" line is from "The Iron Yin and Yang", a crossover game which combined Touhou and Final Fantasy 11.
  • The Teddiursa portrait is a reference to Pikasprey, a famous variety Twitch streamer who also does variety short-form YouTubing. (twitch.tv/pikasprey)
  • The math equation on the right is a reference to Eight Ghosts in Old Genso, a  BabaCi project currently in hiatus.
  • The drink order on the left is a reference to a secret drink in "VA-11 Hall-a".

The shrine gate in-lay was an idea of Pedestal's in order to give the bar scene a bit of Gensokyan flair and not make it feel so foreign. I was initially resistant to the idea, but I'm glad Pedestal insisted on it. The shrine gate is such a lovely centerpiece and adds a richness to the scene. 

There was also one other way to make the game more lived-in: creating an economy. 

A realistic economy was ultimately the crux of the whole game. "VA-11 Hall-a" did an excellent job in making you feel the pressure of paying bills and having to be perfect in order to afford to eat and the like. I wanted to emulate this urgency by making the player work for their new home. In the original vision, M would have to make enough in "commission" and wages to pay for three days at an average, run-of-the-mill hotel (about 17,000 yen.) The commission is about 30 percent in this game (but would likely be closer to 15-20 in real life), and I think at the end of the day, my weakness in math truly showed, as I erred in the final amount and made the game much too easy to win. I based the "commission" win rate on the yen necessary divided by 30 percent, and then went down from there so I could ensure the game wasn't too merciless since there was no longer brewing mistakes to worry about. 

I hoped the weight of those decisions would affect the feeling of the game, especially where money was concerned. Mistakes would mean reduced commissions AND markers; bad conversations would mean less tips, which meant less take-home. At the end of the day, with the engine change to consider and the sheer number of menus we had to create, there simply was not time to effectively create the economy we wanted to create. If we ever have a chance to revisit the game, this will be first and foremost on the list. 

Have a full voice cast

The other two headings above are filled with some bittersweet notions or "what could have beens?", but I am happy to say this section is full of nothing but positives. When coming up with this project, I felt what needed to be in a barkeeper game like this was, ya know, real life people. As much as I loved "VA-11 Hall-a", the one thing which kept me up at night about it was there was never any aural reaction. I had to imagine the laughter, the frustrations, the barking, etc. That can be fun in its own right, but bars are meant to be lively, and filled with noise and humanity. 

Getting a voice cast which could express every character with passion and vigor was a vitally important part of making the game work. Thankfully, everyone who ended up on this project was a huge believer in it, and did what they were asked to and then some. We reached this goal with aplomb and for that, I cannot thank our wonderful crew of VA's enough for bringing this game to life.

I'll go down the list in order of appearance and write a little blurb about each:

  • Saishoo was a hero for us (and many, many, many other Jam entries.) His work was very quick, and he brought a bit of a gap to the character of Hana which I thought was extremely necessary. The delicate frame, prim fashion and laid-back stance combined with the gruff, boisterous and masculine voice of Hana really drives home the mystery of the character. 
  • Nonatsu was someone who was targeted for their bartending expertise, but their voice was also a perfect fit for Rinnosuke, as it dances with a type of "thoughts on tongue" nature I feel Kourin would possess. The big brother nature of Kourin was evident in the way Nonatusu emoted-- as was Kourin's natural goofiness. I was very happy they accepted the assignment, and I think they did an excellent job. 
  • Julie  volunteered to join the project at launch, but was very unassuming. They had never done a serious VA role before, and felt they couldn't hold up. But even despite their nerves, Julie knocked their two characters out of the park. Marisa's mischievous nature really comes through in her bratty and knowing tone, while Mamizou's grandiose and pragmatic nature is beautifully accented by Julie's slower, more dramatic interpretation.  Their ability to grasp tones based off of samples is second-to-none.
  • Prior to this project, Trin had never done any VA before. At the start of the project, Trin approached me out of curiosity and sent an audition for a role, but I initially rejected it. However, I later revisited the tape and realized there was something there with some training. They accepted the role of Aya, and I had to train them pretty hard to get to where I wanted Aya to go. Trin accomplished that and gave their character a nuanced and unique interpretation; I am very proud of the effort they put in to get there. 
  • Zy was an absolute charm to work with from start-to-finish. She always looked for feedback and was prompt with her lines. Her peppiness in group chats always kept us going during lulls as well, she's a saint. The way she brought such an innocence and natural wonder to Kosuzu's character was genuinely wonderful. It was a pleasure to listen back to her reads when placing them in the game. Her interpretation really added to the conversations between Kosuzu and Aya, given Aya's more reserved voicing. You really got the sense Kosuzu was full of heart and vigor. 
  • I cannot begin to praise Emilie enough. Emilie auditioned for several different characters. but I caught them off-guard by assigning them to an off-book option. Emilie rolled with the punches, knocked out the assigned lines quickly, and then stepped up for Doremy  (a VERY different character!) when we were in dire need and did the exact same thing and then some. Emilie's range is amazing; their read for Megumu really brought out the intensity of the scenario and the character, while their read of Doremy was  as enchanting as the lines she uttered. Their innate understanding of both characters was unabashedly awesome, and really saved the dream of a full cast.
  • Rose was the first VA to submit her name for every available role when I posted for VAs on a voice-acting site. Her infectious excitement and submitted tape told me immediately she was going to be a force to work with, and she was. Rose told me she always wanted to do a Touhou fangame, and it was evident. Her interpretation of Hatate was pitch-perfect; she caught all of the myriad emotions of that little bird and brought them to the fore in a way many people appreciated. Rose truly brought Hatate to life and made the tengu scene that much better.
  • Meg was another voice actress who was a gem to work with. It was actually Meg's line reads which inspired me as a director to try and go back to get more line reads for the other characters. It was a happy accident; a misunderstanding resulted in her submitting her entire script as a single file, which allowed some some great voice barks to give her interactions as Byakuren more charm.  Meg brought a saintly, soft-spoken timbre to Byakuren which I greatly appreciated. You can really feel the light coming off her character, and it's thanks to Meg's read.
  • TigerJoleto was the last VA to join the crew, but they knocked Alice out of the park. That's no exaggeration; Tiger joined with about two days remaining, and sent their lines in within 12 hours to help ensure the lines could get into the initial game. Tiger's interpretation of Alice was befitting of the character we created; shy, nervous, but optimistic in the connections between people. Alice's position as an outsider was perfectly accented by Tiger's read of the character. It was a pity I didn;t get to meet them until well along in to the process; maybe there's another chance someday.
  • Mistona was a perfect fit for Keiki as soon as I heard the tape. They had an earthy tone which I felt was appropriate for the clay ease god, and their voice was unlike a lot of the other voices we had on staff. It was something I valued for such a strange character as Keiki, and Mistona did their best to submit their lines and capture the right tones. I was especially happy with their laughs, it felt genuine and nice. One moment which will stick with me in this production was Mistona DM'ing me after reading the script to ask if Keiki was queer. When I told them yes, Mistona expressed such a happiness to voice a character they could identify with. It gave me such heart to keep going, knowing the people I wanted to reach would be heard.

Find a better writing process

The other goal which I aimed for and succeed in finding was a better writing process. In all my prior projects, I had great difficulty in going out-of-order or being able to get out of ruts when it came to figuring out plot details. So one goal I established for myself in particular was to not worry about the order of conversations and just let the characters approach me naturally. 

if one scenario wanted to be written, I'd write it. I wouldn't pressure myself into doing all of Day 1 in a day, or all of Day 2 in a day, etc. Instead, I broke it up into several smaller scenarios, and let the voices talk as they may, not considering a certain "order" unless it was necessary for plot purposes. This presented a few issues at times; I accidentally wrote myself into a corner in Day 2 due to writing Mamizou and the tengu scenarios on separate sessions without consideration for each other. There were also issues with lengths of scenarios which I had to go back and fix. Despite all that, I managed to persevere and eventually settled on an order from all the various sessions which made the most sense.

Doing it that way helped me feel a lot less pressure in general, and freed me to help with other aspects of the game, such as the music, art and voice direction. It is a great balance which really gave me a new perspective on writing in general. I hope I can keep that up for the future.

Bits and bobs

Now that the big four things are out of the way, I'll talk about a few things which were cut from the game in the concept phase, mostly as it pertains to  plot:

  • The game was initially meant to be five "days", with the good end being enjoying a day off at a festival with the new friends you made.
  • In the "five day" set-up, there would have been several more "scenarios", including Sanae/Kanako/Suwako, Yuuka/Letty and Saki/Urumi. There was also the potential of a Sanae/Reimu/Byakuren scenario, but I liked the mystery of the shrine maidens being morally ambiguous and seemingly feared instead of being accessible.
  • Parts of the Saki/Urumi plot were written into Keiki/Alice scenario, while Sanae/Reimu/Byakuren plot and part of the Yuuka/letty plot were grafted into the Miko/Byakuren scenario.
  • The Prismrivers were originally supposed to be in the game, and would have made appearances in all three days. That was scrapped for a number of reasons, the biggest of which was the fact Kokoro was scrapped as a character. Kokoro was originally an anchor scenario, with Miko encountering her at the bar watching the Prismrivers in a continuation of a storyline from KDDS. It would have involved an argument about Miko abandoning Kokoro before she runs out of the bar, much to the consternation of the Primsrivers (who would threaten to stop playing music until Kokoro returns) and Miko. Byakuren would then arrive on the scene, and the family would have a heart-to-heart about their situation. (Any more conversation about this would ultimately spoil a route ending of KDDS.)
  • Much like in VA-11 Hall-a, we wanted to set up a system where you could buy items to keep M satisfied and prevent them from getting confused. This was very hard to implement in either version of the game, and had to be scrapped for simplicity.
  • Another economy thing we had to cut right away was the limits on alcohol. In the original vision, you would have to manage resources when serving drinks. If you ran out of an ingredient, you would have to buy it against your daily wage and hope to make it up with commissions down the road. This was meant to bring more of a challenge to the game, but ended up being basically impossible to implement fairly, considering the daily wage was equivalent to $48 USD (6657 yen) before commissions.
  • There is probably some confusion about the whole "add optional gin" thing. Well, the simple reason goes back to the original engine. In the original vision, you would mix the drinks yourself, so adding the gin would be necessary. We kept the lines in for simplicity's sake and reflected when it was added in the script since that was the most efficient way to take care of it.
  • This also resulted in a lot of deleted secret dialogues; every character had unique drunken dialogues if you served them enough alcohol at some point. However, with the switch to a different engine and no time to implement such things, we had to cut the vast majority of these. A few still exist, but as regular dialogue (Doremy and Mamizou.)\

Conclusion

Every project comes with its own unique lessons, and this one is no different. 

This was the first time I had to go through the process of auditioning voice actors. That entire process was a wild one; having to find perfect elevator speeches for every character, figuring out the right lines to get an idea of the talent of each actor and trying to pair actors with characters was a process I'd never been part of before. I've been on the other side in a past life, so I gained a new-found appreciation having to be the decision-maker this time around. I think, in general, I found a lovely crew who was always willing to take feedback and contribute more than necessary in order to make the best project we could.

Given everything we went through just to get any version of this game out, we did all we could to show the love and care we put into the game. We received so many lovely comments about how much fun people had learning about this world, or how much people identified with the characters and stories we had to tell. 

At the end of the day, that's ultimately what drives me to continue participating in these Jams: to tell the stories which need to be told, and show the people who need to see those stories that they are seen, understood and accepted exactly as they are. If we can make even one person read a project of ours and see themselves in it, we have done our job, no matter what the scores ultimately end up being.

I cannot wait to create a new project and continue to live up to that ethos. This project was dedicated to the love which needs time to bloom, and I hope those who read this story found their love grew just that much bigger.


With love, 

junejijo, on behalf of BakaBaku Circle

Files

Room of a Thousand Petals Happy Hour 1.3 (PC Version) 107 MB
Jul 02, 2023
Room of a Thousand Petals Happy Hour 1.3 (Mac Version) 72 MB
Jul 02, 2023

Get Room of a Thousand Petals

Leave a comment

Log in with itch.io to leave a comment.