Recently I have been replaying Borderlands. Since I started playing it, I've been reflecting on how it's affected my game development philosophy. It's not a game that I would have thought of as being particularly influential in my thought process, but at the same time it has some very interesting lessons to teach based on its mechanics.

The game that I am most proud of is Hammercalled, my multi-genre modular role-playing game. I have not been good at keeping up on recent releases for it, but I've been working on it for almost 2 years now. Borderlands was a critical influence in Hammercalled because it served as a lesson about developing deep and meaningful systems that exist on a single game layer.

The clearest example of this, I believe, can be found in the way that I decided to have Hammercalled have a modular parts-based system, similar to how Borderlands generates its weapons.

Borderlands wasn't the first game to use procedurally generated content, but it did so in a way that was much different than previous games. If one thinks about games like Diablo, which in many ways is a spiritual precursor to Borderlands, there is a level of random modification apply to items. However, the core of those items remains generally similar. You might have swords of various types, but they are the same sword with adjectives applied. This is true of many games, like Dungeons and Dragons online, one of my favorite games. With games like roguelikes, a large focus was creating levels using procedural generation, which often created issues rather than solving them: random levels are often boring, and any storytelling elements need to be interspersed within those spaces.

Where Borderlands is different is that it creates a system to generate new entities, most notably weapons, from parts. In Hammercalled, I took the system to another conclusion. Borderlands uses this approach to enable the spontaneous production of equipment without player input, but as a tabletop role-playing game designer I wanted to have more player stake in the creation of content. The more I looked into this the more I saw that there is something immediately compelling about this whole design principle.

A Joy for Newcomers and Veterans Alike

Parts-based design permits a fusion between simplistic easily logically assessed parts, and more advanced and complicated final elements to allow player choice and optimization.

A good contrast to this can be found in the role playing game Dungeons & Dragons, where players create characters primarily by choosing a class for them. Rather than being parts-based in its design, the game is oriented around one large decision. Depending on the edition, they may also choose races, skills, feats, backgrounds, or other options at character creation and during character advancement, but these are largely supplementary. The weakness of this, however, is that leaves players with an unwanted level of complexity. Players who would like the opportunity to fine-tune their characters find themselves disappointed by the lack of options, and players who simply wish to play a game find themselves overwhelmed by the level of detail that comes with having many elements of a character determined by a single choice.

Before starting work on Hammercalled, I was involved as a playtester and a very small contributor to a game called Open Legend. While playtesting the game, I was immediately drawn to its deep, high-quality characters. They were made of very simple parts, almost too simple (I would opt to use a more complex, granular, system in Hammercalled), but they were accessible to newcomers in a way that Dungeons & Dragons can struggle to be and satisfying for more competent gamers who often feel that Dungeons & Dragons lacks deep customization options.

This is where Borderlands presented a great inspiration for me. Being prone to tinker with, and perhaps maybe even cheat at, games I found myself using a piece of software that allowed players to create custom weapons for Borderlands. This was long before Hammercalled had even been conceived, but it was cool enough it stuck in my memory for many years. It was a way to take all the parts of a weapon, including legendary qualities which are very rare in the game itself, and combine them into a sort of dream weapon. Needless to say, this had a negative effect the play experience because of obvious balance issues. However, just one glance at how dozens of parts across a few distinct categories could be combined into millions of options, of which a significant portion are meaningfully distinct, was breathtaking.

In Hammercalled, my philosophy is that characters should be built from a bunch of different parts. The core character themselves often functions more like a traditional role playing game character, with a handful of attributes and skills. Each is bought independently, but there is an arbitrary differentiation between each. However, other systems like gear were intentionally designed to be built from parts.

The effect of that is that adding new parts allowed for constant evolution of the game, without requiring the addition of dozens of pages dedicated simply to adding content. It also proved invaluable during play testing: every rule that was to be added could quickly be tested in a variety of circumstances. Of course, this isn't a unique element in the Hammercalled system and it is not always difficult to test new changes in other games, but in Hammercalled there was an incentive for players to synergize things, which meant that parts would be tested and refined very quickly, and any meaningful voids could be filled.

Simplified Designs, Complex Layers

Another upside of this modular design is that many things that would otherwise be possible only through using additional gameplay layers can be done through one universal method.

I cannot claim too much credit for this. In fact, it was Open Legend that inspired this design philosophy.

One of the revolutionary decisions in Open Legend is that instead of having distinctive spells, spell casters would just have an attribute that reflects their ability to use magic. This is a strikingly elegant solution to a major balance problem in many other games. It enabled spellcasters to maintain a distinctive identity, but also kept their powers constrained within very clear game design limits.

The focus on such a design may be limiting in aspects, something that Open Legend got around by adding long lists of abilities unlocked by having certain attribute combinations, but it's also liberating in other ways. It means the players can come up with their own rules for the game, then fit them seamlessly into your system. If a particular narrative device requires a player to do something unexpected within the original ruleset, the Game Master needs simply ask them if they have any specialty that would permit them to take a bonus on it and determine which attribute best fits the situation.

In this sense, Hammercalled works like any other game, the things that fall outside the rules of the setting cannot be done. However, there's no arbitrary distinction drawn between different types of action.

The reason why this is so liberating is that players do not need to remember different systems, which also raises a key concern with other games. In Dungeons & Dragons, for instance, if a wizard wishes to stop themselves from falling to their death, they might cast the featherfall spell. However, there are strict limitations defining what a wizard can do in the setting. They may cast a certain number of predefined spells, but filling the gaps is left to the imagination of players and Game Master. For balance purposes, wizards are not generally permitted to use any magic beyond their spell list. The consequence of this is that the game can unintentionally lure them into traps, where they may have poorly prepared for a particular situation and have no recourse. A good player will figure out creative solutions, or plan ahead for a broad range of situations, but this requires skill that a novice player may lack.

This is not necessarily such a terrible problem, since it is likely that another player can provide a solution, but it does lead to potential frustration for newcomers who find themselves perpetually incapable of doing things that would be trivial for fantasy wizards in other settings. Various solutions exist for this, but they are often inelegant. One example is the Prestidigitation spell (and an assorted family of similar spells for other classes of spellcasters), which covers a broad variety of situations. This can create as many problems as it solves. Because spells come from lists and have very clear predefined effects, and Prestidigitation and similar spells are basically wildcards, it can be difficult to distinguish where the line should be drawn with regards to these broad utility spells' capabilities.

One player in a Dungeons & Dragons game that I am running continuously tries to use the spell Shape Water as an offensive spell. I have no personal problem with this, except for the fact that this is a violation of the rules the have been laid out. Since the other spellcasters in the party are generally stuck with spells that have very clearly defined rules for dealing damage, to permit one exception would be to do potential injustice to other characters. This creates an odious situation in which a rules judgment can lead the game to destabilize. By contrast, in games like Open Legend and Hammercalled, it is possible to simply ignore spell lists and permit the GM and players to come to a conclusion about what is acceptable in the moment.

At first, compared to the very clear parts base design laid out in games like Borderlands, this seemed risky. After all, I had been building a game that was largely built out of parts parts. Many of my previous efforts in game design had involved modular spell-casting system, often inspired by the video game series Two Worlds, but the difficulty of creating such a system is not trivial. As a consequence, I was never satisfied with any of these systems. Even when they managed to work as well in effect as a list they failed to deliver on speed and intelligibility. Despite being interesting, they were impractical.

This is also a good place to mention the game Sryth, which I have played for many years. Unlike Borderlands, I did not consciously consider it as I made Hammercalled. However, it has skill checks for powers that function, despite a radically different underlying system, exactly like Open Legend. Hammercalled takes some more minimalist approaches to magic, but seeing these two good examples of games that did away with a spell list helped to solidify my decision making process.

In this sense, it was better to get rid of using a parts-based design for a particular system, and instead placing it entirely into the context of an existing game layer.

Balancing Parts-Based Systems: Choice and Power

To get back to Borderlands, the initial focus of this long-winded, there is something to be said for the quality building a design that can work beyond its initial purpose, which can extend upon itself with the addition of content even if no manual adjustment is done.

While such systems often require complex balance, it is particularly useful for the sorts of games that Borderlands and Hammercalled represent, which are primarily cooperate experiences.

In these cases, mechanically tight balance is superseded by the need for players to feel like they have choices and an impact in the game.

Dungeons & Dragons Online, based heavily on the Dungeons & Dragons system mechanically speaking, differentiates itself from its namesake by adding complicated series of enhancement trees that actually reflect a move toward a parts-based design. At certain levels of experience, characters gain points which they can spend between these trees, and a character typically has access to at least four trees. While not strictly speaking parts-based design this permits much more individuality in characters and also ameliorates some of the balance concerns, since players can tune their characters for specific situations. This does not mean, of course, that all options are created equal. However, the presence of options is enough to permit choice, and so long as players are able to make informed decisions they can usually find a choice that has a desired level of impact in-game.

Where parts-based design excels is that it permits choice. While Borderlands has many distinctive and interesting weapons, its sequel was often criticized for not being as interesting as the first game in its weapons designs. This is not, strictly speaking, because there were fewer interesting weapons in the second game. The real source of this concern stemmed from the fact that there were so many more uninteresting weapons. One of the important design passes that Hammercalled underwent was a conscious act of going through and removing useless elements, while also streamlining many of the existing elements.

The reason for this is that there were choices that have served literally no purpose to boost a character's power: their storytelling impact was not significantly altered by choices they made. Despite this, it's worth noting that Hammercalled still has several highly-specialized and rarely used parts.

Because Borderlands has randomly generated content–or more accurately procedurally generated content, since there are combinations that specifically are more likely to be found together or which incompatible with each other–Hammercalled relies entirely on deliberate human choice. As a result, it gets away with more "dull" options: there are parts available for players to choose when building characters that are intended for very specific purposes, for players who want to excel in very specific situations. If randomly given to players, there is a good chance that this would run contrary to a particular player's style, but when chosen they are much more meaningful. Hammercalled is also built around letting players change many of their decisions, especially with regards to the most clearly parts-based systems, so that there is less room for aggravating missteps to tarnish a newcomer's experience with the game.

The challenge as a game designer in Hammercalled goes from making sure that every individual part is interesting, the problem Borderlands could not overcome, to instead making sure that each of the elements generated is interesting. In Borderlands, the solution was impossible.

Because weapons were randomly generated in Borderlands, and many were designed to be inferior to others, making every part good would have required significant changes to the balance of the game. This would have sabotaged the natural looter-shooter progression loop. The designers must instead do their best to ensure that interesting options are always viable, even if they are rare, which is where the second game fell apart. This required them to find good options, figure out how rare they should be, and then build a system that combines good options frequently enough that they can be kept relevant to the challenges that the player encounters.

In Hammercalled, on the other hand, there is no such need for this. Players find the options that work best for them through experimentation, and continue to refine them as their characters grow stronger by investing character development points into their gear. The result is that the designer only needs to find good options and make them available.

The thing that fills me with the most excitement about Hammercalled is the notion that a future designer can add content which stacks neatly with that which I have created. This would be a magical moment for me, because it's a testament to the system, which I consider to be the most important part of a design.