As with many of my ideas, this one sparked because I’ve been playing a videogame. Borderlands 4 came out on September 12th and I’ve been having a really good time with it, and in particular the boss fights have been a massive improvement over past installments.
I am not the type of gamer to relish in boss fights — at least not the type we’re discussing here today. However, as part of an extended study on dungeon design, I played through Elden Ring. I dislike games of skill-based mastery — I find it more rewarding to play games where mastery is attained linearly through time investment or studying the system. To put it another way: I’m not a soulslike gamer. Despite this, forcing myself through the entirety of Elden Ring, including all of the optional bosses, has permanently rewired my brain when it comes to those fights. There’s a particular type of pleasure that you can only get from throwing yourself against a boss and learning the patterns until they’re engraved in your heart.
We don’t really have that in ttrpgs.
The closest thing I’ve personally experienced is the FFXIV ttrpg, which does have boss telegraphs and changing behavior based on player positioning. Additionally, I’ve been informed that Rowan, Rook and Decard are currently working on a game called Hollows, which seems inspired by soulslike influences and features “many-v-1 tactical boss fights”. I started drafting this post a week before I found this out, and I was shocked to see a “circular grid” on their backerkit — one of my ideas I thought was going to be entirely original. In any case: I’m excited that other designers are thinking about this problem and I’d love to know about any other projects along the same trajectory.
radial positioning
The foundation of this post is that you frame all questions of position relative to the boss. Rather than use a grid to represent a specific place, the radial wheel is used to represent where you are in relation to the boss. My system in this post is opinionated based on the level of granularity I enjoy and tailored to something I could use in The Serket Hack, but you can add or remove layers to suit your own needs.
To start with: the cardinal directions — north, south, east, west. This system defines directions such that the direction the boss is facing is always north. We use compass directions as a shorthand, but you could just as easily use something like numpad notation if you desired. By default, these directions are cone-shaped, but there are some advanced cases I’ll address a little later.
Layered above that are the inter-cardinal directions — northeast, southeast, southwest, and northwest. These are four more cones rotated 45 degrees, or perhaps quadrants now in the traditional Cartesian style.
Finally, there is a distinction between “in melee” or “out of melee”, which I abbreviate to simply being “in” or “out”. This corresponds to the frontline / backline in The Serket Hack or Ryuutama.
Now that we’ve established our cool radial positioning system, what does that actually give us? Even as a system agnostic concept, this does nothing for us unless we have attacks and abilities based on directional positions. Even then, changing the way we think about space doesn’t suddenly recreate the feeling of fighting difficult bosses in action games. For that, we need the other half of the equation.
telegraphs
A telegraph is an indicator that the boss is about to do something, and it should include hints or even outright delineation of the details. There are three components to a telegraph, though more experienced GMs should try experimenting with the format and adjusting to taste.
First: the name
Announce that the boss is telegraphing. Then state the name of the attack. Names can also be a hint, and function best when all of the attack names are interrelated. Some quick examples:
- Strong Rush – the boss attacks the north cone
- Strong Swing – the boss attacks all north zones (NW, N, NE)
- Tail Sweep – the boss attacks the south cone
Second: the hint
Include a description of anything the players can see related to the upcoming attack. These can be diegetic hints (stuff the characters can see) or they can be non-diegetic (stuff only the players can see). Hints can be auditory, like a specific voiceline, or they might be environmental changes in the arena. More examples:
- The boss holds up a spear, ready to thrust it forward – attack hits the north cone
- The boss shouts “Face my wrath and suffer the consequences!” – attack hits all north zones
- The GM places an AoE (area of effect) marker on the south cone – attack hits the south cone
Third: the timing
Different games use different methods of handling initiative. This post assumes that the game takes turns and that the players have some way of knowing what order those turns happen. As part of the telegraph, the GM should communicate when the attack happens:
- The attack happens after two turns.
- The attack happens at the end of the round.
- The attack happens the next time players damage the boss.
how it works
Whenever the GM announces a telegraph, each player writes down a position for that attack. This information is not represented openly.
Players may talk amongst themselves as is common in many other games, but there is no centralized grid with miniatures to reference, so planning must be done vocally — perhaps even in-character. If a player wishes to devote time and attention to making sure everyone is in place, they can take the Coordinate action. This represents taking their eyes off the boss and checking the location of their party members. Doing so allows them to view everyone’s positions (both current and planned), and allows everyone to edit their positions if necessary.
In essence, positioning becomes something of a memory game.
In order to be good at a boss fight, you need to understand a boss’s telegraphs and communicate with your party about your location and movements. The end-point of a system like this is the development of pre-planned tactics and formations.
the juice
I’ve been talking my way around it to make sure that the concept is clear, but I want to walk through some ideas about what the combat puzzle looks like in action.
Unordered telegraphs, saturated:
This is sortof the basic scenario of telegraphs in action. Nothing can really go wrong here unless the players misinterpret the safe zones for the telegraphs.
- Player 1 turn
- Telegraph 1, happens at the end of the round
- Player 2 turn
- Telegraph 2, happens two turns later
- Player 3 turn
- Telegraph 3, happens the next time the boss takes damage
- <Attack 2 happens>
- Player 4 turn, damages the boss
- <Attack 3 happens>
- End of round: <Attack 1 happens>
Tank required:
A player must stand in front of the boss and weather Attack 2 or the boss will rotate to find a target for that attack. The positions written for Telegraph 1 do not “move”, effectively changing their relative orientation to the boss’s new direction. If a player does not tank Attack 2, someone must take the Coordinate action to update any positions for Attack 1 if necessary.
Stack and spread:
A “stack” attack does not target a specific zone relative to the boss. Rather, it requires that multiple players be in the same zone when the attack happens. Typically, a single player will get a “stack marker” and the damage from that attack will be split among all players in the same zone as the marked player. Failing to stack usually results in a lethal amount of damage.
Similarly, a “spread” attack marks one or more players and deals damage to the entire zone that the marked player is inside. If two or more players stand in the same zone, they each take all of the damage inside that zone.
Stop and go:
A “stop” attack deals damage to characters if their position for the attack is different from their previous position. As you might guess, the “go” attack deals damage to characters if their position is the same as their previous location.
These attacks are fun to layer into a round because they can force players to move out of position for a later attack or keep them from moving into a stack.
There are…rough edges here. I think the foundation is solid, but there are too many moving parts to clearly see whether this would hold up in a real game. In short, it needs playtesting.
In an ideal world, the players have some kind of token to keep track of their position stack, but I haven’t quite figured out what that would look like in practice. I think if you had a deck of cards with positions it would help, but I also think you’d need a lot of cards to represent the amount of attacks per round. Writing them out is definitely faster, but that’s also a pain if the boss rotates or if you need to add a new attack in-between two that you’ve already written.
Let me know if this inspires you or if you have any ideas on how to improve it!