Everyone has been crunching hard for the last several weeks as we rush to finish off Reina & Jericho. The game has been very close to feature complete for a while, but there is still content that needs to be designed or improved. And bugs. Lots of bugs. Lately the focus has been on two things: finishing off the boss fights, and testing the level design throughout the game for both platforming difficulty and routing issues. Routing in Metroidvania games is a challenge, and in the case of Reina & Jericho it has a few added layers of complexity because time travel always adds layers of complexity. Thus far the bug rate has been very normal, but the actual traversal and routing issues have been quite minimal. All of this is either good news because the levels were designed well, or bad news because no one found any of the mistakes. Time will tell! The other big achievement from the last few weeks has been wrapping up most of the boss fights. The final battle of the game remains outstanding but beyond that they are ready for play testing. Designing the boss fights has been interesting. I feel like the traditional game design advice – the sorts that would come up in a YouTube video – would be to come up with a mechanic for the fight and then build around it. Having tried that approach a couple times, I think I dislike it both as a designer and a player. There's an almost ideological rigidity that comes with it, and it feels like design decisions are being made in the wrong order. To use scientific terminology: Game mechanics should be the responding variable, not the manipulated variable. The other approach I have heard of is to focus on the player’s behavior and ask, “What should the player do?” For example: “The player should have to duck underneath a big attack” would lead to to attack that covers a lot of vertical space but has a gap at the bottom the player can duck under. While I found this approach useful as a thought experiment, in practice it created battles that felt scripted. My preference has always been games where there were several acceptable behaviors to choose at any given time, and execution of general game mechanics was still the most important thing. The boss fights that have come the most easily are the ones that were started by creating a clone of an existing boss fight, stripping out most of the attacks and AI behaviors, and then asking "What would make this fight feel different?" If all of the other fights are fast can this one be less about speed and more about timing? Sometimes this has led to us using the exact same mechanic more than once but in a different way, or in concert with other boss abilities that force the player to use different strategies. Other times it has led to novel one-off ideas. In many cases this led us to migrate attacks from one boss to another to keep them more thematically consistent, and I think that’s been a good thing too. In most cases there was still an initial loose idea – the boss of the Laboratory area is based on the idea of “Bullet Hell” – but there was never a specific mechanic in mind. Reina & Jericho doesn't feature a boss fight that is all about throwing boulders or hitting switches or bouncing energy balls back at an enemy. Some of those things happen, but quite frankly: in Reina & Jericho a *lot* of things happen in the boss fights and the player just has to keep up and be ready to pounce when an opening appears. Sometimes this approach resulted in a game mechanic that became prevalent over the course of the development process, but other times it was nothing so concrete and more resulted in a feel that was unique to the fight. I think this encapsulates my own prejudices well. Most of the games that have boss fights I’ve wanted to emulate have just focused on the players ability to respond to what they are seeing. In Reina & Jericho almost every attack has numerous potential counters all tied to the core game mechanics, and your success or failure in the fight comes down to execution. I'm excited to see what people think.