Posted April 05, 2021 by 3 Halves Games
Another month, another Sovereign Tea devlog. It's been quite a lot of fun since we started these, hasn't it? Let's talk about where we are with the development of the title and what's left to do. We're almost done!
Our goal last month was to reach 100% completion of all levels. We didn't quite hit that goal because the design for late-game units got complicated, forcing us to spend more time on paper prior to jumping to implementation. We'll talk about each and every hurdle, why they came up, and how we plan to go forward.
The first hurdle we came to face was the enemy AI itself, particularly in battles against the boss Coffee Regents. There are many types of enemy behaviors that exist in the title, including:
While these work for general units in the game, they don't quite work for the human boss characters, as it causes them to play in an unexpected manner. Our goal is for the enemy to make decisions that would be reasonable to make if a human were playing them. This is, of course, a massive undertaking. We're still iterating ideas and concepts for how to really juice the boss levels in order to make them feel truly special. One of our goals for early April is to nail this; in fact, we are spending this coming weekend (April 3 - 4) explicitly focused on this front due to its importance. The most likely end result is that each Coffee Regent may have its own unique AI behavior type to accommodate each unit's unique style.
The levels aren't finished yet because units aren't finished yet. This makes sense, as you can't reasonably finish iterating on a level without unit mechanics in place. The real question is why the units aren't yet finished. The answer is... Counting.
Since the dawn of development of this title, we wanted to build towards a unique mechanic called Counting. The idea was simple: units can trigger specific abilities to give themselves a stack of a status effect called Count. By itself, Count does nothing. However, when a certain number of Count is accumulated, the unit will upgrade and transform into a much stronger unit. From a design perspective, the goal was for you to carefully leverage your units in specific ways to get a specific end result. Due to this complexity, we saved these units for the endgame; all Milk units utilize this mechanic in some capacity.
Those of you who played the beta and even the alpha of this game may remember the Level 03 boss, Kopi Cham. Kopi Cham behaves in a very similar way: when it gets one kill it transforms, and when it gets an additional three kills it transforms again. This is, in fact, a counting mechanic! However, the implementation of Kopi Cham for both the beta and alpha didn't actually use the Count status effect, as it didn't exist in the game. It instead used a different, but similar, status effect called Boss Power. The underlying implementation of Boss Power was very hacky and tuned specifically to Kopi Cham, so it couldn't be reused for any other unit.
We ultimately had to redo the entire way in which Kopi Cham works to use the new Count status effect. This was a bit of a significant undertaking as our goal was to develop the counting mechanics in a generic way so as to maximize our ability to iterate on unit designs. We reached that goal in late March, but unfortunately one week is not enough time to design and play through every single remaining Black Tea level. Thus, the progress is only at 40%; we have all the layouts for the levels ready to go, but we haven't been able to test them out fully as units are still being created on a daily basis.
We expect to reach 100% on all levels, and thus reach gameplay complete, within the next 1-2 weeks.
We don't want to showcase too many of these new Milk-based Count units, as they represent fun endgame content. We will, however, introduce you to just one. You already know about the aforementioned Kopi Cham, but we'd like to introduce you to the basic Milk unit you can brew, quite simply named Milk Tea.
Milk Tea has two forms: its initial form and its post-Count awakened form. The initial form is your standard small unit that has 5 HP, 2 Attack, and 3 Move. Its ability for how it gains Count is simple, as it's meant to onboard the player to the concept of units that using counting mechanics. At the end of the turn, it will gain 1 stack of Count. When it reaches Count 3, it will undergo a massive transformation and instantly gain 10 HP, 2 Attack, and 2 Move. That leaves its awakened form at 15 HP, 4 Attack, and 5 Move; an incredibly powerful unit.
This powerful unit is just the elementary Milk unit. The stronger units you can brew, the ones that require Milk and other special resources, are a lot more powerful.
Gameplay balance is something we need to talk about. On paper, a unit with 15 HP, 4 Attack, and 5 Move that only takes 1 Tea Leaf and 2 Milk to brew is insanely powerful. We even just noted that this is one of the weaker Milk units. Why are we making such overpowered units? Let's discuss:
We had a long conversation internally among the team about this direction. The first step was to engineer the tools needed to design these units, which as stated above took some time. The next step was to design the units. The third step finally is to try them out in real combats and see if they're fun and compelling.
The end result? They're extremely fun to use! These are some of our favorite units in the game and we hope you like them too!
We didn't just spend the entire month thinking about the above, of course. There are more units to share with you that don't use the above mechanics! Let's take a look at the three you see above:
As you can imagine, communicating all this information to the player is a challenge. We spent a lot of effort recently revisiting our tooltip system to ensure all information is conveyed. For example, in the beta units like Stoma (seen above) would have tooltips explaining that they would deal stacks of Poison and Regen, but we had no tooltip explaining what those were on the unit itself. We now support the concept of referenced status effects, for when a unit's tooltips mentions a status effect that it does not have. These are visually displayed as smaller than other tooltips to make it clear that the unit itself does not have that status.
We also now showcase the range of certain effects on tooltips as needed. This is also useful for Dragon Well in particular, as the Dragon Power status effect grants her temporary buffs to her attack range. The highlighting in green for augmented, buffed status effects is also there for improved communication.
The last noticeable tooltip improvement is for units with Multi-Move and Multi-Attack. In the above example, you can see Ballet Tea now shows its Move action more clearly as 1x6. Previously, it was displayed in a manner that was difficult to read.
There also is another fun optimization fact regarding tooltips. Did you know rendering tooltips is the most CPU-intensive operation in the game currently? Yes, it is more expensive to show tooltips than it is for the enemy AI to make a decision. Since this discovery, we have made massive optimizations to the game; tooltips now are 65% more performant than they were before! Even after this, tooltips are still the most expensive operation in the game, so there is a lot more to go on our end.
Lastly, those of you who have played the beta may also notice that the tooltips shown above are a lot larger and thus more legible. We still have a bit more work to do on this front, but we're really happy with all the clear improvements made to this system over the past month!
We are continuing to hammer out the massive amount of writing required for this title. We haven't talked much about this in devlogs as we'd like to keep the progression of the game's plot a surprise for when you play the game. That said, we want to talk about Level 08, the last level of the available beta. Spoilers below if you haven't completed the beta!
The original design for Level 08 was very different than what ended up in the beta. Initially, at the end of Level 07, Rooibos and Chamomile were driven out of their lands and forced to sea. Level 08 was supposed to be a naval combat level, where neither side would have direct access to resource tiles, as they would both be on ships. Ultimately, that idea was scrapped as it was difficult to visually implement naval combat. The general idea of a level without resource tiles still persisted and is what ultimately ended up in Level 08.
In terms of narrative, Level 08's post-battle dialogue was going to be very different. We wanted to introduce Masala Chai at this point. The idea was that after fending off the Coffee Empire's naval attacks, Rooibos and Chamomile would briefly land on Masala Chai's shores and meet her. This served as the first opportunity to introduce the otherwise seclusive, pacifist Tea Princess. Unfortunately, this plan for Level 08 was cut for the beta as it was out of scope, particularly in its CG needs to fully realize.
We would like to go back to the original plan, to an extent.
We aren't going to change the gameplay or scenario for Level 08; we like the way the fight with the Soluble Admiral plays out. Instead, we want to expand upon the post-battle dialogues to include meeting Masala Chai. With how the rest of the writing has played out so far, Masala Chai doesn't get many opportunities to show up in the game. As we don't want to start you on Level 19 with her without you first knowing who she is or why she's important to the story, it is imperative to include her early and often.
In general, this is the biggest narrative challenge of the title: we have 12 playable characters who all have very different backstories and motivations for their participation (or lack of participation) in the war. Atop of that, we have 5 non-playable characters from the other side of the war. Internally, we are keeping close track of how much screen time any single character gets. Some characters have more relevance in the game's overall narrative than others, which is unavoidable, but we want to ensure that each character has their chance to mold the story.
To keep it brief, we felt that Masala Chai wasn't getting enough love; after all, she is the only Tea Princess not present in the beta of the game, though she is mentioned. One solution to this is to introduce her earlier and thus afford her more opportunities to appear earlier. We want to do this by editing the post-battle scenario for Level 08. When the full game comes out, we recommend replaying the level if you have already beaten it so you can see the new story content.
We want to apologize for the inconvenience, but we also hope we've provided justification as to why we wish to make this change. It will ultimately make the game's narrative a lot better, so we feel we must make this change. We'll remind you to replay Level 08 when the full title launches; after all, we don't want you to miss out on exciting new content!
Wow, this post sure has gotten long, hasn't it? Congrats on making it this far if you're still reading! We still have a bit more to cover. Let's start with our remaining updates:
Priorities for this month of development:
As you have seen in this unbelievably long devlog, we are nearly complete with the development of Sovereign Tea. We'd really like to thank you all for sticking with us through this surprisingly transparent development cycle. It hasn't been the easiest thing to produce this game during the trying times we find ourselves in, far from it, but we're all very proud to have gotten as far as we have with how things have been. Please consider following us on our social media, on Twitter, Facebook, and Discord. Also feel free to comment on this devlog with questions and feedback if you have anything you would like to share with us. Thank you!