Designing Incremental Roguelikes

In this post, I’ll delve into the game design decisions behind Strata and share insights from lessons I’ve learned. This blog is for anyone with an interest in (incremental) game design or those curious about the behind-the-scenes of Strata’s creation.

Design Pillars of the Game

Embarking on a new project often starts with identifying its genre and deciding how you (as a game designer) intend to meet or challenge the expectations associated with that genre. These foundational concepts are typically referred to as design pillars and serve as the backbone of your development process.

It’s common for people to confuse design pillars with “hooks” though I consider them distinctly separate concepts. While a game’s feature might serve as both a hook and a design pillar (for example, in Stray, the player navigates a sci-fi underground city as a cat, an engaging hook that also significantly influences gameplay mechanics, world-building, etc.), this isn’t always the case.

An example of a distinction can be found in an insightful piece on Game Design decisions behind Slay The Spire here - note that these pillars are very fundamental to the game (example: “Provide more than one problem, or more than one aspirational goal.”), and in sum they add up to the whole experience, while a single one of those probably couldn’t be advertised as “hook” easily. For a more in-depth discussion on the importance of design pillars check this Game Developer article.

For Strata, the design pillars that guided its development are as follows:

1. Everything scales
2. Items are one-time use only
3. K.I.S.S. (Keep It Simple, Stupid!)

Some of these have been both a blessing and a curse and let’s go through each one of them to see why.

1. Everything scales

I initially planned to create the game in a few days and after some brainstorming I’ve come up with the idea on how to achieve this: the player will go through floors linearly. This meant that both the challenges players face and the rewards/items they receive would scale accordingly based on the floor they are on. However - while this made gameplay intuitive and predictable after playing for some time and realizing what’s going on, it made it a lot harder to balance and provide fair experience to larger audience.

There was a fine line between making the game too easy, which could quickly become boring, and making it too hard, which could lead to frustration if you didn’t get a random item you wanted. Scaling curves behind both enemy and player power levels couldn’t be adjusted to balance it out - since power of the player directly correlated to the items they were getting (and this is all to noticeable in the early game).

If you’ve read through previous 2 paragraphs you’ll notice how a more minor decision (items with randomly generated effects that easily scale ) basically undermines whole design pillar from which it originated. From my experience this is completely normal for an indie game. You try things; if they don’t work, you iterate and keep the fun parts until whole system is fun. Reason why I referred to this game often as prototype was because I realized very early on this design flaw - but wanted to see how far I can take these core pillars without changing them. If you’re taking your game seriously - don’t do what I did, iterate over the ideas and play test them until they fit your vision.

Everything I mentioned along with some randomness of items you get made this game not for everyone: to stand a chance of advancing far you have to pick “meta” picks that skew these scaling power curves in your favor which only comes with playtime.

2. Items are one-time use only

Implementing one-time-use items was a decision aimed at defying the expectation of a roguelike. This design choice is supposed to force players to think carefully about when and how to use their items, adding a layer of tactical decision-making that is crucial for success. However, due to imbalance in scaling, items go “stale” pretty fast after acquiring them, making the whole idea behind this kind of fall apart.

While 1st design pillar had fundamental flaw, this pillar is salvageable. What game needs is stronger hand-crafted items placed on specific floors along with incentive to keep items (upgrading them when not used, stacking bonuses etc). I tried to somewhat address this in v0.3 with introduction of powerful relics - but it’s only part of the solution. Game really needs more complex mechanics that will incentivize this kind of decision making: “Should I use this now, or keep it so it gets even stronger?”

Another thing is that this pillar kind of requires more complex item system. Think about Slay The Spire - when you build decks, you most often pick things for their synergies even tho they might not be strongest. So you bend everything game throws at you, in order to advance your current deck/play-style as much as possible. My game fiercely lacks this as items are basic buffs with no synergies at all. Synergies help emphasize importance of both individual items and item as a part of synergy. Current items feel somewhat flat with “one-and-done” deal.

This is the pillar that I’ll be very thoroughly refining for the full game I’m planning.

3. K.I.S.S. (Keep It Simple, Stupid!)

The K.I.S.S. principle guided me to streamline the game’s mechanics, making it very accessible. This simplicity allowed players to quickly grasp the core gameplay and immerse themselves. This is the reason why items only work with base stats (most of the time) and why there is only one action to use at the end of the turn (attacking).

I think this pillar in itself is fine, however it definitely needs to change to “Keep It Simple, Unless it Is Fun!”. To me it seems like I’ve missed on some fun additional concepts that I threw away for the case of simplicity (additional actions, weapons, enemy decisions etc). Again: iteration is key - especially in gameplay heavy game in roguelike genre.

Speaking of which… what about roguelikes?

Core Of Roguelike Design

Roguelikes are characterized by their challenging gameplay, permadeath, and procedural generation, aspects that ensure each playthrough is unique. In Strata, I tried to embrace these elements wholeheartedly, but with an unique twist of not keeping the items you get, reflecting design pillars I’ve chosen.

Gameplay is probably the most interesting part of all roguelikes. My game had a relatively simple gameplay loop built upon aforementioned design pillars. However despite this, it was challenging to categorize this game for marketing and pitching purposes - as there are very little established things similar to this game. Thinking about the game itself I later realized that I basically built a deckbuilding roguelike but without the core parts: cards and interactions that come from their combinations. Lesson to be learned here: it’s easier to establish your game in known genre or sub-genre if it can be - you’ll thank me when you need to do elevator pitches and win over potential players on social media. With that being said I’m proceeding to build this game as deckbuilding roguelike since it has always been that: I’m just letting game be what it needs to be to reach it’s core audience.

Permadeath, combined with one-time-use items, underscores the importance of each decision. There are many discussions nowadays about whether meta progression ruins the roguelike experience since it kind of removes “perma” from permadeath - since you almost always come back stronger. Proceeding with my game - I’m also deciding to keep it meta progress free, knowledge and experience will be all you need to play and have fun.

Procedural generation is also very interesting part of roguelikes that my game kind of misses. There’s no variations in the enemies you face and it’s fully deterministic based on what floor you are (only thing different are basically actions they take). This is also part of the genre I need to incorporate better and I’ll be playing around with as well for full game.

New Frontiers

Looking ahead, I’m going to try my best to expand upon what Strata proposed with it’s design pillars - without sacrificing and stripping fun parts. This includes adding new layers of narrative and thematic depth, introducing more varied and complex enemy types, items, synergies, proc gen and many other things that are clearly missing.

As I continue to refine and expand Strata, the feedback from community remains invaluable. Your insights help me navigate the challenges of game development and steer the game towards new and exciting horizons.

Thank you for reading. I hope this post has shed light on the complexities and joys of game development, and perhaps inspired you to create something yourself :)

If you’re interested in keeping up with my work consider following me here on Itch.io or Twitter!