Designing for Discovery: Actuated Diversification in Classic Shooter Game Design
Learning is Fun. DOOM is Fun. Learning while DOOM is Funner. In this article we'll take a look at how classic shooters teach their players through experience, and how you can use these practices too.
Quick note before we start! This article is a recreation of a talk I gave at the official UnityNYC event on September 18th. If you’d like to watch this article instead of read it, you can do so by clicking here
We Constantly Learn Through Play
We love to explore. It’s playful. As designers, it is our job to outline and define rules of play which allow players to consistently discover and mess with new game mechanics. Both in and out of games, humans are naturally curious. It’s the way we learn about the world. As babies we’re chewing on everything. The more we learn, the more we realize we don’t know. From a young age we start asking questions, and as we grow older we are met with a natural curiosity to understand the universe around us.
We muck about, we find out.
Players are Self-Guided Learners
Players of all ages are self-guided learners. They are self-motivated and directed in the space of play. Leaving a player in a room with a button, cube, and door will always result in them opening the door.
It is this basic principle which allows us to design puzzles and environments which teach players through experience. In the world of education the two major approaches to teaching are Pedagogy and Andragogy.
Pedagogy being the more lecture-directed approach. It is the methodology we are all familiar with from grade-school. The teacher dictates knowledge to a classroom, and the students learn from highly directed learning.
Andragogy is the opposite approach. It is self-guided learning in which students are allowed to drive and direct their own learning. This is vastly more common in adult learners. As we grow older we build more critical thinking skills, which creates a stronger desire to seek out and answer questions on our own.
The majority of players, even from a young age, are andrological learners. No one likes tutorials which simply ask us to walk through a level with a series of tooltips and popups. Those are boring, they stop us from playing. But, no one cares if they are left alone in a room with a button and a cube.
While these two styles of learning are not mutually exclusive, when players are forced to use pedagogical forms of learning in play, it can result in a far less engaging experience.
To learn more about Andragogy and Pedagogy and how they relate to game design, I highly recommend checking out this book by famous games educator, James Paul Gee
Just how far does this go?
Throughout this article we’ll be exploring how classic shooters (specifically DOOM (1993), DOOM (2016), and DOOM: Eternal (2020)) use methods of experiential learning to teach their players how to play.
DOOM (1993) is a fantastic example of this. The design of its levels go from short corridors with only a handful of enemies and paths, all the way to massive environments with snaking tunnels, multiple levels, and loads of secrets to discover.
We can see that Level E1M1 is a significantly smaller and less-complex level, compared to E4M6 which appears to be a confusing and vast web of interconnected tunnels, levels, rooms, and secrets.
But how do they do this? What methods do they use?
It’s all Self-Guided Learning
Strangely enough they simply continue to use self-guided learning. The first level of the game presents a simple rule: In order to succeed you must reach the end of the level without dying.
The next level introduces an evolution of that concept: In order to reach the end of the level you need a key to unlock this door.
While the game never dictates this, the player can attempt to open the locked door. After they fail the only other option they have is to explore the space. You can see in the map above that our very first key is a big deal. It’s surrounded by uniform and heavily designed geometry.
More Tools, More Obstacles, More Complexity.
Throughout the course of DOOM (1993) the design continues to place new obstacles between the player and their exit. This means new enemies, more keys, and hazards such as lava and acid.
Parallel to the introduction of new obstacles, the player also sees quite a few new tools at their disposal. Exploration grants access to new weapons, and allows the discovery of new powerups. By placing health packs, armor, and ammo around the map, players are encouraged to explore. It reinforces the idea that exploration is not just an essential element, but an abundance of it creates an easier combat experience.
This Builds Scaffolding
As the player discovers new enemies, weapons, and powerups, they quickly begin to understand how each of these enemies function. This creates a level of understanding through play that can be designed around.
If players understand that Imps fire slow moving projectiles, Barons are slow moving and have lots of health, and that enemies with mechanical weapons require more precise movement, then we don’t have to reteach these concepts to the player.
When they are confronted with a boss who both resembles these components and utilizes them mechanically, the player is able to confront it without needing to review every aspect of how it works.
If you’re curious about Scaffolding both in Educational Psychology and Game Design, check out this free resource
How are these elements organized?
Orthogonal Unit Differentiation
Orthogonal Unit Differentiation is a term coined by Will Wright during his design of the original SimCity. It states that units in a game can be described by actions, abilities, and characteristics that are orthogonal (numerically different) to each other regarding functionality. This allows us to build elements that can be numerically understood and tracked.
Organizing enemies in DOOM (1993)
Take for example the chart above, where Endurance is a combination of each enemy’s total Health and (lower) Stun Chance, and their Aggression is their Movement Speed, Attack Rate, and how much they Damage the player with each hit.
We can quickly see how enemies in this game are diversified by these values. In the bottom left we can see the easiest enemies in the game. They have low health, and a high chance to be stunned when hit, making their endurance quite low. Their aggression is also very low, making them less threatening to the player.
Then we have the polar opposite - in the top right we have an enemy known as the Pain Elemental, which is quite apt. It has a huge pool of health, is rarely stunned by the player, rapidly shoots high-damaging projectiles, and moves quickly around the space. Needless to say this is one of the most difficult enemies in the game.
But then we have a few anomalies. In the bottom right we have the Skull demons, who have nearly no health, are almost always stunned, but will always attack you on sight and can potentially kill you in one hit. They are a prime example of how Orthogonal Unit Differentiation can be used to design a more dynamic cast of enemies by simply modifying rules.
Using this methodology we can begin to outline different simple types of encounters that the player might see in their playthroughs. Each of them utilizes a different kind of play, and combines enemies of similar quadrants.
In the bottom left we see a combination of Zombie-men and an Imp, which allows us to teach the player the basics of movement, enemies, and projectiles. It’s not particularly difficult due to its low Endurance and Aggression, but it is a helpful tool to teach and fill space. The combat here is going to be short and easy.
The bottom right is a group of low endurance and high aggression enemies. It’ll more than likely be over quickly, but it will require the player to move more dynamically and thoughtfully around their enemies. Due to the high aggression, the Pinky demons will charge the player at the same time as the Skulls, forcing the player to move around. Getting hit here is risky, but this combat only lasts a short time.
Focusing on the top left, these enemies will have high endurance but lower aggression. All of these enemies can be stunned quite easily, allowing the player to pick and choose their targets more strategically. While getting hit here doesn’t nearly yield as much damage as the bottom right scenario, this combat lasts significantly longer, making it just as dangerous in a different way.
Finally in the top right, we have what can only be described as an encounter that is extremely difficult. The high endurance means it’ll be an extremely long combat, but the high aggression means it’ll be extremely easy to fail. This should be used extremely rarely, as its difficult is rooted in the player building scaffolding around the enemies they are facing. In my opinion, games should have one, maybe two of these depending on the length. They are walls which may not feel fun to overcome, due to their intense sponge-like and quick-to-lose nature.
Linear Difficulty
Now you’ve probably already noticed this, but there is a clear and traceable linear increase in difficulty on this chart. Using this isn’t necessarily bad, but relying on it too much is a mistake that early designers frequently make.
Planning Encounters
Breaking up the chart, we can more easily read these four quadrants as High Risk, Hard, Easy, and Movement. Using these, we can plot our encounters for a level out onto them.
Using our linear difficulty placement, we can quickly see that our encounters are all quite similar in combat style with each of their attributes increasing at a steady rate overtime. Let’s put this onto a level and think about what it is like to play it.
Here every single encounter is roughly the same, with enemies all rising in Endurance and Aggression as the player progresses. Functionally the play the player is performing in encounter 1 is the same as encounter 6, they are just doing it for longer periods of time in more frustrating scenarios as they continue.
Encounter 1 is quite easy, 2 is a bit more difficult but no sweat. 3 begins to bring on more enemies and difficulty but isn’t necessarily difficult. 4 is the first one which is actually hard. 5 is pushing the limit of what the player wants to do, and at this point they may have already had to die and reload their save a few times. At the very end of the level, 6 is more than likely too difficult for a new player to beat. It is a long fight, is extremely difficult to stay alive in, and requires a significantly larger pool of experience to draw from.
If the encounters increase linearly on our orthogonal scale, they simply become more difficult without adding nuance. Throughout this level the player is not learning anything new, they are simply repeating the same task over and over for longer periods of time. This is the easiest trap to fall in when designing combat encounters, and can result in boredom.
Let’s take another look at our encounters, and see what happens if we distribute them differently on our scale, but leave the first and last encounter where they are…
Even with a more random placement we begin to see some variation in the encounters. Let’s walk through what this looks like on a map to see how this might fit into play.
Orthogonally random distribution of encounters creates a more interesting environment of play where the player is constantly doing something different.
Immediately we can see clear and exciting differentiation of play here. The first encounter remains easy, teaching the player how to go up against enemies. Encounter 2 presents them with a far more movement-based combat, introducing a new area of play to improve. Encounter 3 presents a higher risk, with larger demons which attack less frequently. Moving to encounter 4, the player draws on their knowledge from encounter 2 but in a more intense environment. The enemies here don’t necessarily have to be any different - if the player fought Skulls in 2, then maybe encounter 4 simply sees a jump-scare involving Skulls directly behind a locked door, or more of them in the same space. We then reach encounter 5, which is a riskier combat than 3 and begins to move into the Hard quadrant, but is just on the border. This combat takes a little while, but isn’t something so insurmountable that the player gets stuck. Finally we reach encounter 6, where the player is pushed to their limit. New players will more than likely die multiple times here, and this encounter will have to be properly tuned to match the level of difficulty throughout the level, but it makes for a far more engaging and satisfying path of play.
Pushing players to overcome a variety of obstacles, each with a variety of solutions and and tools, allows them to learn new things on their own. Learning new things is fun when we can figure them out on our own.
Motivation & Reward
Linearly increasing combat does not make a game ‘bad’, it simply means the reward & motivation for completing combat must be external from the combat itself. For example, the Looter Shooter genre is designed in which the reward for completing repetitive combat is delivered through rare and scarce items.
In Classic Shooters the Reward for Completing Combat is New Combat. This is why we must rely on the differentiation of play throughout when designing combat rooted in this principle. Building a shooter in which the combat is too similar throughout, without rewards or sub-systems to reward the player, the fantasy of play will quickly be lost, and players will lose interest.
Compare DOOM (2016) and DOOM: Eternal
Next I’ll be discussing the difference of application of diversification in DOOM (2016) and DOOM: Eternal. This section is extremely subjective, and represents my thoughts on what truly separates the 2016 game from the newest entry in the franchise. Remember: The successes and failures which I outline here do not make these games better or worse, but are measurements of what I value and look for in these kinds of games.
First I’ll summarize what we’re going to explore, then I’ll outline each element in-depth.
DOOM (2016) Diversification Successes
In DOOM (2016) you are consistently introduced to new mechanics. Individual elements of play are fun in their own right, and don’t need to rely on other systems. Mechanical elements are introduced through player exploration, there is rarely hand-holding, and exploration is both encouraged and safe.
DOOM: Eternal Diversification Failures
In DOOM: Eternal mechanics are explained to you directly via popups rather than allowing them to be discovered throughout play. Many of the individual elements are not fun or interesting, and in many cases actually impede play. Exploration often results in punishment or impediment, and enemies often have weaknesses to specific weapons, limiting and restricting valid forms of play.
Comparing and Contrasting the Two:
Experiential Learning
In DOOM (2016) you are forced to experiment with each of the different weapons to discover new forms of play and engagement. Rarely does the game stop play to inform you of actions you can perform. The vast majority of the time the player has no option but to try out new weapons and abilities to seek out new forms of engaging enemies. This experiential learning allows the player to dynamically and quickly change their style of play depending on the needs of the enemies and environment. Out of ammo with the assault rifle, but need to break off a Mancubus arm? Flip to the Shotgun’s grenade launcher attachment. There are avenues to success that the player can discover by experimenting.
In Eternal weaknesses and combat solutions are dictated to the player. Rather than allow the player to connect the dots themselves, popups appear which direct the player to play and engage in specific ways. This ultimately results in a removal of discovery and learning, creating a far drier experience of play. It removes the fundamental aspect of exploration which I believe makes games like this so interesting, and replaces it with tutorial-ization.
Scaffolding
DOOM (2016) sees the player encounters increasingly complex enemies and challenges throughout play which continuously build on top of one another. Enemies with metallic weak points allow players to quickly identify weak points on new enemies. Challenges with weapon rewards directly reward player skill with those weapons. Each of these parts of play works off of either previous experiential learning, or scaffolding constructed by it. Never in play will you find something completely foreign to the rules at hand.
DOOM: Eternal abandons this principle. With the previously mentioned addition of popup tooltips, players no longer need to figure things out on their own. This means that new high-level mechanics are introduced randomly and suddenly, resulting in a drastically rapid stalling of play. Upon release many players were drastically upset at the addition of the Marauder to the game, and with the abandonment of scaffolding we can begin to see why. In DOOM (2016) all enemies are built off of one another. The skills and abilities of one enemy can be quickly applied to another. In Eternal the Marauder is randomly and suddenly the first enemy to include a counter mechanic. He is invulnerable until his eyes glow red, just before his axe swings at you. There are no other enemies in the game which employ this mechanic before this, making it naturally something frustrating that players have to beat. If their only prior experience to fighting the Marauder is one popup tooltip which says “shoot him when his eyes are red” then naturally you’ll be annoyed when you get stuck on his fight.
The same can be said the purple ooze placed all over maps. Whenever the player is standing in the ooze they are slowed down to one fourth their movement speed and they cannot jump. This new mechanic simply goes against what the player has been learning this entire time: Move Fast, Kill Stuff. It places a new constraint that blocks the forming of scaffolding, and ultimately creates a small space of play.
It is important to remember that neither of these ideas are “bad.” Many games use counter mechanics or sticky surfaces - the only difference here is how these elements are introduced. In my opinion, it is a failure of the game’s design. If there were a series of smaller enemies leading up to the Marauder which required the player to counter, then no one would care about the Marauder.
Encouragement of Exploration
DOOM (2016) encourages exploration. Rarely does exploration actually involve any hazards which can ruin and entire playthrough a level. Exploring off the beaten path is a low risk and high reward endeavor which mainly focuses on the player’s skill of discovery and observation rather than their ability to dodge obstacles. Much like the older DOOM (1993) this form of exploration invites the player to explore more.
Eternal gives us the opposite impression. With the addition of the parkour system, exploration is now more obvious to the player: Jump through these obstacles and you’ll be rewarded on the other side. Fundamentally there are two things I have gripes with in these design decisions.
Firstly, making exploration now an obvious target by painting parkour elements into the map removes the novelty of discovering a secret. While there are still some natural secrets to be found, the vast majority are now pointed out with symbols that might as well say “jump here.”
Secondly, exploration now features significantly more obstacles and dangers in the way, resulting in players quickly taking damage when attempting exploration. No matter the difficulty setting of the game, the player takes huge amounts of damage from their health pool whenever they fail a platforming challenge. This rapidly discourages any form of exploration, turning it into an aspect of replay-ability at best. Why would I risk my health and armor here when I could just ignore that platforming challenge and come back later?
Actuated vs Forced Diversification of Play
These terms are entirely made up by me as I couldn’t find any good documentation of them online.
In DOOM (2016) all forms of play are valid with minimal constraints. Want to use the shotgun for the entire game? Go for it until you run out of ammo. Want to only melee and use the rocket launcher? Great, go for it. Enemies have multiple weaknesses and don’t force you to use any kind of specific weapon. This is what I am calling Actuated Diversification of Play.
DOOM: Eternal sees the opposite form of design. Enemies have singular specific weaknesses which can only be fulfilled by specific weapons and abilities. Take for example the GIF above. The player fires a grenade into the mouth of a Cacodemon, the demon then pops comedically, and is now ready to be killed. If you attempt to kill this enemy with any other weapon, you’ll quickly learn that their health pool is dramatically larger than you realize. Almost every enemy in the game features some kind of mechanic like this in which the player is forced to diversify their playstyle rather than explore and build their own forms of engagement. Because of this, I’ve called it Forced Diversification of Play.
Application to Design
Now that we’ve done quite a bit of analysis, how can you apply these principles to your own project? What are the steps to doing so?
Outline Units, Rules, and Variables
Before you do anything else, take the time to outline the Units of your game. If you’re building a game about racing, what kinds of terrain do you have? Is weather an element here? In DOOM units comes in many forms of weapons, enemies, maps, and much more. Before you dive into your diversification, analyze these aspects of your game. Assuming you’ve been designing up to this point, this process should be pretty straightforward.
After this take note of their rules. Your tires grip the track, do they have friction? Can they burn out? What about the aerodynamics of your car, does that come into play?
From there ask yourself, what are the variables behind these Units? In the example I built around DOOM (1993) I used enemies as a Unit type, and then graphed them based off of variables built out of rules. Endurance is a variable built out of the rules “enemies have health” and “enemies can be stunned.” Ultimately it is up to you to find the variables to graph which serve your game the most effectively.
Once you can view these clearly, review them and ask: “Are my variables diverse enough to serve my gameplay?” More often than not, the answer is no. If enemies in DOOM (1993) did not have a stun chance then the only stat to measure endurance by would be Health, and that’s pretty boring. By creating more rules you can expand your variables, making them more interesting to engage with and play with.
Intensify Diversification
If you are hesitant about adding new rules, or you have a lot of rules and want to mess with them, a great thing to try is intensifying diversification to see what happens. This means pushing your variables to the limit. In Mario Kart the contrast between a Green Shell and a Blue Shell is as intense as it gets, with the Red Shell sitting in the middle of the two. More direct examples from DOOM can be seen here too:
If we intensify only the rule of firerate, then the two most intensely diverse weapons are the Pistol and the Minigun. Both weapons deal the same amount of damage per shot, with the only difference being how fast they shoot. By doing this with every rule we are able to quickly and effectively create a series of extremes within our ruleset.
Once you’ve created these extremes it should be pretty clear whether or not you have enough rules to create an explorable space of play. If you find yourself only creating a few weapons or new elements in your game because you’ve run out of room, or your extremes simply aren’t extreme enough, consider adding new rules. Can your weapons reload? What if bullets could bounce? How slowly can bullets move?
Prototype & Test
After making it this far it’s time to get creative. Prototype and test your ideas. Start with what you think will work best, and go from there. Keep in mind that not all of these ideas will necessarily work, and more often than not you’ll result in failure. However, every once and a while you’ll find something that really hits the nail on the head, and is just wildly fun to engage with.
There are entire books on how to prototype new game ideas so I won’t be giving you a process on it here, but if you are new to game design and interested in prototyping, I recommend reading A Playful Production Process: For Game Designers (and Everyone) by Richard Lemarchand and Rules of Play: Game Design Fundamentals by Katie Salen Tekinbas and Eric Zimmerman
Examples From My Own Work
To round off this article, the final section will feature how I’ve applied these design methodologies in my own practice and where they led me.
Praetor Prototype
All the way back in September of 2020 I was still building prototypes to figure out what I wanted my graduate thesis game to be. After a few months of experimentation I had begun building a metroidvania third person shooter out of gameplay elements and ideas which I enjoyed. I named this project Praetor, after my love for the DOOM games and how much they had inspired so many elements. It featured a lot of great aspects such as a Borderlands-style weapon generator, a fully procedurally animated player character, obstacles, enemies, and music. However, there were a few challenges I kept on running into.
Although there were hundreds of generated weapons, they all felt the same. This made the system feel flat. I felt similar about the combat, it was pretty awkwardly paced and quite flat. There was no nuance or interest in how the player engaged with enemies. Nothing forced you to move in different ways. And after all of this, dying simultaneously felt meaningless and too easy. So I took a step back and outlined my variables, and decided to push things in a more extreme direction.
Godwalker Prototype
At the start of January 2021 I built the first prototype of what would become Godwalker utilizing the tools I had built for Praetor. I removed the weapon generation system and used it to build a set of three intensely diverse weapons. The shotgun, rifle, and sniper. I then pushed the player to have to play quickly by prototyping the God-meter. This was the first test I had ever performed of the mechanic “The faster you play the stronger you get,” and because I had pushed everything else to their own extremes, I was able to extrapolate a much larger design from it.
I made weapons widely different from one another. Combat now had curated waves to control the pacing, which helped to remove non-linear variation in combat. Enemies and the environment forced you to move differently, and healing was tied to how quickly you played.
Godwalker Vertical Slice
Now in 2023 the current version of Godwalker as many know it features loads of different weapons, enemies, and maps. Functionally it is a vertical slice of a much larger game, but it employs all of the elements I had built in the original prototype to much more intense extremes. Everything centered around the refinement and testing of new mechanics.
Fundamental Rules and Results of Playtesting
By far the most difficult aspect to expand upon was the diversification of weapons themselves. I had to build a ruleset for weapons which allowed for specializations of each one. There were hundreds of avenues to go through, but I ended up finding a few simple fundamentals that I was able to expand upon widely.
There were only a few rules which I added that created more interesting forms of play. First, weapons only reload when they are in your backpack, but you have unlimited ammo. Second, weapons become less efficient at filling your god meter overtime but their damage remains the same.
Immediately the results of this presented a new challenge. Playtesting revealed three major components of play which directly informed my future design decisions.
The best players are those who can play fast with every single weapon
These rules incentivize picking up and learning new weapons constantly
Build making quickly becomes an important aspect of play
After these rules were added, I was then tasked with building more rules into the weapons to allow them to be diversified even further. I also had to overcome a design challenge in which the best weapons were simply the highest DPS.
To increase diversification I added new projectile and weapon behaviours. Explosive bullets, homing rockets, bouncing grenades, slow moving plasma, and close range shotgun blasts all yielded enough space to create a series of interesting new weapons. I also added in behaviours of weapon inaccuracy and maximum range to help the arcade-like feel of them come through.
Many players who were far too good would pickup a sniper-rifle, get into God-mode where you have unlimited ammo, and then kill everything without a hiccup. They would be able to snipe every single enemy in the space without breaking a sweat. To combat this I added in a new rule which players never picked up on, allowed good snipers to continue try-harding, but didn’t disrupt player. I made Godwalking more difficult with higher DPS weapons.
This meant that players could still play just as fast and aggressively with the high end weapons, but it now required them to do so. Previously having a rocket launcher with unlimited ammo was a super easy strategy, but now players have to ensure they maintain a constant barrage of rockets or they’ll lose their God-mode.
On top of all of this, this mechanic remains secret (until now). It is definitely something players can figure out on their own, but there are so many dynamic effects for how the God-meter is filled that it is hard to discern. Shotguns are more effective up close, Explosives are extra effective when you hit multiple targets at once, and Plasma weapons are most effective when hitting the same target in succession.
Results After Changes
After designing all of these rules and building 34 different weapons of extremes, ranging from Plasma snipers to Explosive rapid-fire guns, I found some very promising results from playtesting.
Orthogonally extreme weapons became cult-fan favorites. What one group of players loved another group despised. This was especially true with weapons that catered to a specific playstyle.
For instance, this is the Fleet. It is a close range weapon that fires three projectiles at once. It has an extremely low damage per second, a tiny magazine, but has the highest effectiveness of any weapon in the game. To an inexperienced player, if one were to pickup and attempt to use it you would fire the shots off for a few seconds, do very little damage, and gain a bit of your God-meter. To the more experienced player; Getting into God-mode, gaining unlimited ammo, and then spamming this weapon is a master strategy. To players who focused on making builds with their weapons rather than focusing on play, intensely diversified weapons were the most used.
On top of this, there was no “best” playstyle. Players were able to beat the game using a myriad of different styles. Some focused on long-range play, while others relied on short range. Players used what they were most comfortable with to beat the game, and relied on systems differently based on their level of skill. Highly skilled players were able to ignore the God-meter entirely, they simply did not get hit and managed to kill enemies easily with any weapon they had, while low skill players relied on the God-meter intensely, constantly swapping to more effective weapons in order to stay alive.
To Conclude…
Games are at their best when we are constantly learning. The more we explore both as designers and players, the more we discover which we have to learn. By creating rulesets with diverse ends we can create spaces of play in which players can explore and discover new elements of play rapidly.
Through games, we are constantly learning and growing. The constraints we as designers create are inherently those of educators for our players. Exploration and Discovery are fundamental to self guided learning.
Thank you for reading!