Developers Need to Find the Balance Between Procedural Generation and Their Own Personal Touch
As procedural generation becomes more and more prominent in gaming, titles like Chasm and Black Future '88 show how a more personal touch to the mechanic can help.
Procedural generation has been around since the early days of gaming, with one of its earlier and most notable interpretations being in Rogue, the ASCII title and namesake of the wildly popular “roguelike” genre. Since then, procedural generation has been applied to tons of games in various ways. Some games like Dead Cells or Spelunky do it well, while poor implementation of procedural generation could hamper an otherwise solid game, like 20XX.
I am usually a huge fan of roguelikes, including those that take advantage of procedural generation in unique ways. That being said, I’m curious to see if there are better or more interesting ways to implement it into future titles, and games like Chasm and the upcoming Black Future ‘88 have found an interesting way to go about it, with a ton of potential.
Let’s start by looking at some of the flaws of procedural generation. While it can allow for a near infinite amount of content, lackluster implementation of procedural generation or messing up in other areas can severely hurt the experience. For example, No Man’s Sky was bombarded with negative reception at its launch even though it featured procedural generation on an unprecedented and impressive scale, mainly because it forgot to make the game fun. Even if you have an infinite amount of content to explore, the lack of a personal touch when it comes to gameplay can make things problematic.
“While it can allow for a near infinite amount of content, lackluster implementation of procedural generation or messing up in other areas can severely hurt the experience.”
Meanwhile, there’s a game like 20XX that almost has the opposite problem where its gameplay is solid, but the procedural generation is lacking. This Mega Man X-inspired roguelike has extremely tight controls and fun weapons that do a great job at emulating the gameplay of the titles it is based on. Unfortunately, 20XX’s procedural generation can become quite random, resulting in some situations that are overtly obtuse and seemingly impossible to overcome. Several other lesser procedurally generated indie games suffer from this as well. While No Man’s Sky lacked a personal touch (before its recent relaunch with the “NEXT” update) on the gameplay side of things, games like 20XX suffer from a lack of care put into level design.
Procedural generation shouldn’t just be a tool thrown around to create more content: the greatest benefits will be yielded when it’s balanced out with more traditional, hand-crafted level design. Chasm, a game I reviewed recently, and Black Future ‘88, a title that I spent some time playing at E3, both go about procedural generation in a different way: handcrafted rooms, but placed procedurally.
Chasm from Bit Kid Inc. isn’t a traditional roguelike game, but a Metroidvania-style platformer that randomizes its layout each time through. All of the rooms were designed by the developers, which gives them a more practical and well-designed feel that slowly builds upon the game mechanics that you are learning. Black Future ‘88 from SUPERSCARYSNAKES on the other hand is a roguelike, but does the same during its seventeen-minute runs, keeping things fresh with unique and creative enemies and weapons.
“Chasm…and Black Future ‘88…both go about procedural generation in a different way: handcrafted rooms, but placed procedurally.”
This mix of handcrafted design and procedural generation works well in both games. Chasm has been just as fun during my second playthrough, and I’m itching to revisit Black Future ‘88 at launch after trying it for under an hour at E3. Meanwhile, other titles like Dead Cells also balance the two to great results. While this may not be the best solution for all developers, it’s definitely an interesting way to approach it, and shows how a more personal touch to game design can go a long way for players.
When it comes to games that utilize procedural generation and handcrafted design, I think the best solution for most developers is likely somewhere in the middle. I have played plenty of well-designed games that were procedurally-generated, and even more that were not. Finding a balance between adding one’s personal touch to the game during the design process and using procedural generation in one way or another to make more content should maximize both quality and quantity, which will only benefit the player.