Why & How I Make Games (Live Action Berzerk Edition)

I like to know how things work. My curiosity about the world is a significant inspiration or catalyst for why I make games.

The seed of a game comes from a feeling. Sometimes, these sensations come while playing existing games, and sometimes, they are completely unrelated to games at all. It’s like a chef trying to capture the flavor of Autumn—I try to capture the essence of a moment.

This is a tricky hobby, though, as games need context. They cannot simply exist in an imagination or a notebook. They must be expressed, experienced, observed, and iterated.

--

My goal as a young adult was to own a camp. I needed a laboratory to test my ambitious experiential creativity. Camps are the perfect venue for experimentation. There was no expectation of fidelity, nor was anyone telling me “No.” Bryan and I would run into the woods and create something new to see how it would play out - whether that was forging a new trail, leaving weeks worth of clues, or propping up a coconut that could predict the weather. I introduced full-camp experiences to the venue, from starting a morning announcements period that bulged with songs, surprises, and epic drama to weekly events that often mixed elaborate spectacle and storytelling with engagement mechanics. These camp memories and a grounding as an outdoor experiential educator strengthened my knack for sculpting a big game idea into reality.?

Big games are challenging. They often lack the vigorous playtesting of other games simply due to the desired context. Additionally, these games are more likely to occur during one-time events, which means there is no second chance to get something right. There is a thrill about testing an idea with a group the first time, anticipating success. I’ve encountered this for almost thirty years, luckily with no major failures.

This isn’t to say that my predictive powers are perfect. Not at all! I use the one significant advantage big games have over most games - facilitation. Almost all of my games were facilitated by me. I was even the original ESC Games facilitator (MC), and created the role to ensure the games worked, even when they didn’t. I wrote the handbook on it! Facilitation is the superpower that lets you control the narrative in real time. It allows you to tweak and adjust to a fluid context. A facilitator can always change rules, and experts can do it invisibly, making plan b seem like it was the intention.

Planning is the key. I always plan, even if it’s just a list in my head of backup options. There is a primary plan and fail-safes along the way. This is another gift from the hectic life of a Camp Director. Things change, props break, rain falls. A facilitator has to be ready for everything, often to their detriment. In late 2023, I ran a group and arrived with so many games that we barely scratched the surface before the event was done. It wasn’t that I miscalculated timing; the group arrived late and segmented and then happened to lean in firmly to the first couple of experiences. These were just the appetizers, and it was a bummer that I spent days prepping a main course that never materialized.

One of my nightmares is being somewhere and having someone call on me to run a game immediately. I have games for exactly this ask, so that’s not an issue. It’s more that I don’t know I’m giving everyone the best experience without proper consideration. Give me an hour or so to plan or prepare props! Let me make a backup or three. Let me properly understand the goals and desires of the encounter.

I consider every game that I create deeply. It’s why I have so many games languishing in my notes or the back of my head. The game needs to manifest fully within my imagination. Sometimes, that’s easy, like with Steal the Bacon: Professional Edition. I knew the game would work within seconds of contemplating it. The trick with that game was thinking through the nuances, such as the total number of bacon pieces and players.

Other games, however, take a long time. This book (coming soon!) contains examples of me fighting with mechanics to get them right. For instance, Live Action Pocket Monster Snap was a struggle for a long time. So was Mine Craft Adventure. It’s why notebooks have so many pages.

One such game is Live Action Berzerk. Berzerk is a classic arcade machine that caused the first video game-related death. The player had a heart condition, and it’s unlikely the game itself had much to do with the accident, but the story made headlines, and the headlines made a legend. Regardless of the history, the game is excellent.

The player travails a series of rooms. Each room is maze-like and filled with some number of robots. The robots advance toward the player or shoot at set intervals. The player is capable of shooting the robots. Bullets do not go through walls. The player dies if they touch a wall or a robot or get hit by a robot. Robots die if they touch a wall or get hit by the player. The player receives a bonus if all of the robots are killed. On leaving the room, a new room and a new set of robots appears. There is no goal other than survival, and there is one last twist. Instead of a timer to motivate the players to move from room to room, there is Evil Otto, who flies through walls and kills the player on impact. Evil Otto is a giant, smiley, invulnerable face. It’s awesome.

I have thought about Berzerk for a long time. It’s a game that I wanted to make with stanchions, such as those at airports or banks. Those are expensive, however, and even when I had some on hand, it wasn’t enough. Additionally, stanchions would need to be reconfigured for each room - that’s a pain. The slow pacing would kill the game.

The robots were another issue. They would have to be performer players, like in Field Frogger or Do Move Say, but how would they shoot? Any projectile would go through the barriers, and dodging bullets in the arcade machine was possible. I didn’t want to get caught up in arguments over the walls. The game felt hopeless for a long time.

Then, I started writing this book (coming soon, I promise!). It’s been a journey, and as I write these sentences, I’m nearing the finish line. That’s a strange thing to write so early in a book (likely chapter 2), and I know it’s taboo to get too meta in one's writing, but I started with the game chapters and moved backward. That felt correct, and even if it wasn’t, I’m here now. And I had a revelation.

I try to consider every game that I create deeply. Berzerk was one of my longest considerations, but I was never pressured. Often, when I was stuck with a particular set of mechanics, I would make a deadline. That is, I would promise someone to bring the game to a festival in a short period. Field TD and Field Q*Bert were like this. I knew they could work, so I just promised Adam at Obscure Games that they did work. Field TD went well, but Field Q*Bert didn’t, at least in my recollection.

I never promised anyone Field Berzerk. Perhaps this is because I always have another option, or maybe because I thought no one knew the source material. It’s more exciting when folks are chomping at the bit to play your game. The madness over Mine Craft Adventure is a great example of that phenomenon, or the virality of PacManhattan.

There was no pressure to make Field Berzerk, so it was a seed in my mind. But writing this book inspired me to consider the challenges once again. What if I could use this to explain how I make games?

--

How I Make Games: Field Berzerk Edition

Let’s break down the critical mechanics of Berzerk.

• The player travails a series of maze-like rooms.
• Each room is filled with some number of robots.
• The robots advance toward the player or shoot on set intervals.
• The player is capable of shooting the robots.
• Bullets do not go through walls.
• The player dies if they touch a wall or a robot or get shot by a robot.
• Robots die if they touch a wall or get shot by the player.
• The player gets a bonus if all of the robots are killed.
• On leaving the room, a new room and a new set of robots appears.?
• After a set amount of time, smiley-faced Evil Otto flies through walls and will kill the player on impact.

Now, let’s organize this list (shockingly, one can't make multiple bullet levels in LinkedIn).

It is essential to break a game down into its key components to understand exactly how it works. This list isn’t comprehensive, as I haven’t specified details like the actual scoring, the speed of the bullets on screen, the player's movement speed, or any of the opponents. Once some of the main mechanics are solved, those details will be part of balancing.

The first challenge is figuring out the maze. The game needs to have a relatively quick pace, so room creation has to be already done when the game starts. Here’s where context comes into play. The solution for this opportunity will be different if I create a game inside a student union versus on the street. Many game festivals, such as Come Out and Play and IndieCade, have streets as part of the big game venues, so I’m going to solve for that context. The other quality of the map is that robots move through it. I want to keep that in mind while coming up with a solution. There is likely a way to satisfy both requirements at once.

The initial solution is to create a grid and use multiple colors on the grid to indicate different rooms.?

Each map is superimposed on the same 15 x 15 grid. Each square is 2’ - 3’ large (to be determined with space and testing). This map showcases the walls of the Red Room and the Blue Room. Once a player exits a room, they are told which color room they enter.

This is a practical solution. I can draw out a grid and add the colored walls using chalk or colored duct tape (more likely duct tape, as they will get trampled on). It will take a while to prep, but the grid will come into play in other ways. There is an obvious flaw in this first pass when creating maps - the rooms are too easy to leave. There are two solutions to this issue. One is that a player may only leave the room once all the robots are defeated. This seems like a good rule, but I’m not convinced it’s the best idea. Instead, each room will have two X’s, one for entry and one for exit. The player can choose when starting a level which X is a good starting spot.

The overlaid rooms add another factor to the game. Players can play ahead, studying the different colors before playing a particular setup. This also ups the difficulty, as the walls can get confusing - especially when the player does well enough to have two different colors constitute a level (this is an idea in progress, but it’s worth testing!). Another consideration is to have only four or so rooms, as difficulty can be ramped up in other ways, and as a single-player game, the total game time should be relatively tight.

Note—Another consideration when using overlaid colors to differentiate walls is accessibility. High-contrast colors or tape with a pattern on top would be optimal to allow the greatest level of participation.

The robots are the next consideration. Tag can be tedious, as it leaves out players. Some are faster and more agile than others - that’s not the kind of game I want to create. Also, I would have to cycle robots if the game was simply Tag constantly, so I won’t use that as a primary mechanic. The robots can either shoot or move, but not both simultaneously, and they do so at a regular interval. This is solved by creating a soundtrack for the game with a kind of integrated metronome. Every time a particular sound is made, the robots take an action, whether a movement or a shot. The grid helps, as it confines and standardizes the movement - robots can only move to eight adjacent squares (technically a nod to the original game) and must move toward the player.

The movement makes sense, and the difficulty for both robots and the player can be dialed up by increasing the metronome frequency. The bullets are still an issue, however. How should I solve that? The bullets in Berzerk are merely an annoyance. They are not hard to anticipate or judge, and they are stopped by walls, which makes it even easier for the player to evade trouble. My current strategy is to ditch shooting bullets entirely and borrow a mechanic from the game Ninja. The robots can either take a move or make a tag motion toward the player. This motion has to be a single complete movement to tag with a hand or dodge (more on this later).

The Ninja mechanic is fun and easily understood. This will create a challenging arena for the player - every moment, the robots get closer, and once they are close enough, they begin to attack. These tags won’t work through walls, and facilitating the edge cases will be much easier than tracking who threw which projectile. My first impulse was to use two different sounds and force the Robots to attack every other turn. Adding a layer of choice to the NPCs makes it more of a game for those participants and adds challenge for the player. A couple of additional constraints for the Robots is that they must have two feet inside of a square after every movement and, during an attack, they must plant their feet and keep balance. This is mainly a safety consideration, as some players can get a bit enthusiastic!

Note - Do Move Say and Field Frogger rely on humans acting as NPCs (non-playing characters). Giving these characters a tiny bit of agency in the game makes all the difference - they don’t mind making up the fabric of the game board as long as they can somehow contribute.

The player’s offensive move must draw them close to the robots. Killing robots in Berzerk is relatively easy, as the bullets fly straight and accurate, but shooting them often puts a player in the way of danger. This is recreated by putting flags around each robot, similar to Flag Football. This forces the player to get close to disable a robot, adding an element of risk vs. reward. The player will have to plan each grab precisely to avoid a robot’s reach or compensate for a dodge and evade other approaching enemies.

That is the main game. Robots move and attack in a regimented style. The player moves through the maze from one point to another, avoiding walls and attacking robots. The origin spots for the robots could be defined per room, but it might be better to allow them to start wherever they want, with a few exceptions. Robots cannot begin within five grid squares of the player’s starting position, nor may they start within two spaces of one another. This adds another layer of strategy for the robots, making the role more fun.

Two last game details need to be solidified. The first is the win condition. The arcade machine is about getting the highest score. This is tricky in a game where there is a line of people waiting to play—an entire game session should be fun, satisfying, and short. Keeping an experience tight time-wise is always good practice. We want the players to want more, not get bored.

Let’s review the feelings of the original machine. The primary sensation is anxiety. The game kills you with a single misstep, and the walls as a hazard amplifies the tension for the player. We want to recreate that as closely as possible.

The player only gets one life in Field Berzerk. This choice keeps the game rounds moving quickly and should terrify the player. The goal, therefore, is not just to get the highest score; it’s to get through a set amount of rooms. Three rooms seem to be the optimal number. This presents several vectors to adjust with scoring: number of robots killed, room complete bonus, all robots killed bonus, and total completion time. The goal is to attain first place on the scoreboard, and assuming that players kill at least one robot or complete a single room, they will have a score greater than zero.

Tension is the goal of the game. If I can get players to shake while playing, I have won. This feeling is essential not just because the arcade machine evoked that in players but also because of the legend: this game killed a person. That’s how intense it is (whether true or not). Ratcheting up that feeling makes the experience wholly unique and more akin to a haunted house than a field game.

The final piece of design is Evil Otto. The main idea is that the player is attacked by Evil Otto after a set amount of time in a single room. A sound effect and likely end the game will trigger this. The sound should be an unwelcome surprise and rare enough that players almost forget that it could happen (so soundtrack timing will require testing).

There are several ways to present this enemy. One is to have the robots throw projectiles on hearing the sound. This would effectively kill the player, but it doesn’t feel as terrifying as it could be. Another idea I held onto for a long time was to put a giant ball on a rope and run it across the maze. While this would be hilarious, someone would get hurt. Introducing Evil Otto should change the game from a slower-paced strategy game to an all-out game of Tag. The player should sprint to the exit point while avoiding robots, walls, and Otto.

Evil Otto should be another NPC, perhaps one of several standing around the outside of the playspace, that dons a mask and immediately attacks (runs at and tags) the player. Ideally, there are four NPCs, each with a mask at each of the four corners facing away. One of them is the chosen Otto, unbeknownst to the player, and on hearing the sound, they turn around and chase. This execution adds to the mystery - which of the four will it be? When will the sound trigger? How much time is left?

The introduction of Otto will likely cause the player to hit a wall. The outer borders of the grid are walls, and the natural impulse of someone in a Tag situation is to find space. Balancing the wall placement, attacking robots and Evil Otto, and still getting to the exit point will be all but impossible, and that’s okay. It should be hard!

The following steps are to finalize the small details, such as the cadence of the metronome for robot movement and the amount of time per room. These particulars are easy to test, but they will depend on the final space constraints (for instance, more movement opportunities for a larger space). The challenging aspects of the game are done, though, and it will be fun with a little bit of playtesting.?

So what now? Well, run it and let me know how it goes. Make some changes, and make it work. My guess is that 3' per square is correct, meaning the field of play should be 45' x 45' (and then another 2.5' around the perimeter as a gutter). This can be shifted and the space doesn't need to be square. Also, be aware of the surface you play on - I'm always nervous about too much physicality on streets. The danger should be noted, especially to the robots, who might opt to tag a bit harder than necessary.

You made it this far in an article devoid of imagery(!). Find a gathering and space and run the game. Send me your thoughts, your choices, your recommendations. Take some pictures! Do this, and I'll update my book to include your story. Good luck!