Around the time I decided to join Beregon's jam, I was working on a heavily-edited version of Yanfly's discontinued Conditional Turn-Based (CTB) battle system. This plugin, I planned to showcase for a Harold Jam 2024 (if it were to happen). However, my edit needed used in a real-use case where any unresolved bugs are easily caught and smashed.

And so, I was in need of a stepping-stone project where my goals are to practice my core dev skills. My big goals for this game at the start was to the effect of this:

• Showcase my CTB battle system which is to be mechanically comparable to Final Fantasy X.
• Showcase my platformer plugin and use it for the purpose of "spicing up" the game loop.
• Try a more traditional Dragon-Quest style damage formula of 4 atk - 2 def. I've previously used basePower * atk / def which came with massive problems.
• Experiment with a buff system that adds a flat number to battler stats and stacks infinitely
• Create a Fred game. For those who do not know Fred is a RTP character similar to Harold but is hidden in unused files of RPGmaker MV under \NewData_FantasyEN\.
• Characterize Fred as the "that guy" and practice writing 'friendly antagonist' characters who don't play well with your party.
• Create a battle system with five party members to familiarize with how that dynamic plays.
• A more traditional, straight-forward save-the-world plot built around a theme around thinking for yourself vs thinking about others.
• Get practice with digital painting with RTP edits
• 90 minute to 2-hour playtime
• Dev time of three weeks. I will not use the full 5 weeks due to school.

For the most part, I consider this game jam a huge success.

The good

• Transitioning from a Default Turn-Based system to CTB system make battles much smoother and more enjoyable, that I've found for the first time I'm beginning to enjoy my own combat.
• I've applied lessons from past jams on what a balanced difficult looks like. Early game content* I've gotten absolutely no complaints about.
• Giving tank skills to enemies I've seen do very good at encouraging players to learn defensive play

Mid-production lessons

• However my original plan for just switching to a CTB system wasn't enough, I've found it necessary to use Yanfly's Action Sequences to further streamline the battle system. So, I tacked on another additional goal of re-learning to use those series of plugins. Now, I have an even more controlled feel for my battle systems.
  
• The castle originally had to place the five macguffins into different locations but lack of dev time meant I now had to pool them all into the same location and be found all at once. This now meant I had much more limited capabilities to characterize the antagonist.
• I also originally planned a much shorter Act 2 but realized it may be a good idea to pad that out to flatten the learning curve. This definitely played a part in a shorter Act 3.
• I did not get any opportunity to get that tiny practice with digital painting. Replicating RTP line quality is a very tall order for someone not accustomed to those motions.
• Because this game also debuts my platforming plugin, seeing it in a real-world use made it immediately clear it's still a very bare-bones plugin with minimal mechanics. One step at a time.
• Flat buffs don't synergize nearly as well as I initially expected and I needed to bump up flat buff values to adding as much as +10 per use. A +1 stat buff is insignificant late-game when support skills are supposed to matter the most, I may need to try an exponential defense formula with a polynomial or exponential formula for attack.
• Designing a five party member game is more challenging than it appears. Character balance wise, I've had no issues with, but in terms of writing I feel I did not put enough mental resources into executing a five-man-band type of cast and focused on a couple of "showcase" characters. 

The not-so-good

• This is probably my first game jam where I have encountered major issues of execution as the theme I had in my head did not convey very well in practice. In hindsight, the idea I had was a very complex one. What's not yet clear to me is if the execution issues are due to the complexity of my point or there are more foundational issues.
  • If I had to attempt this theme again, I may need a much longer Act 1 to firmly establish the protagonist's core motivations, and a much longer Act 3 to firmly establish antagonist motivations as well. If this is a problem of my point being too complex for the alloted storytelling time, that is dev time and play time I cannot afford for this jam game.
• This game is also a rude reminder that it's not enough for a game to be easy if me, the dev, knows all the strategies would cheese through the game. I've added a lot of mechanics for this game, and seeing some people play made it very clear I cannot expect players to utilize and maximize on every strategic option.
  • For next time, I should not just playtest with lazy attack spam but also playtest with an offensive style that does use skills but with no buffs, and expect that to be the default playstyle.

For next time

This whole game is for the purpose of staging for the next Harold Jam (should it come). Is Harold Jam going to be the next time I design a short RPG? I wouldn't know, but the takeaways from this game tells me this is what I should try:

• Use exponential attack and defense formulae which should be the easiest to convey information to the player. (e.g. Increase attack by 7%. Stacks infinitely until the end of battle).
• Use SRD's Battle Popup Customizer to add better feedback for buff skills
• Come up with better battle designs to encourage more defensive and supportive play
• Rework my platforming plugin to be controller friendly. (I may or may not showcase it again for Harold Jam)
• Add in nudges to encourage players to use haste and slow skills and realize the importance of controlling the tempo of battle.
• Design final bosses that have ebbs and flows from doing absurd damage to them playing completely defensively. This means I'll need to refine my current approach to coding battle AI.

Many more lessons can be taken, but these are the immediate ones I think about.